A Review of Image-Based Simulation Applications in High-Value Manufacturing

Abstract

Image-Based Simulation (IBSim) is the process by which a digital representation of a real geometry is generated from image data for the purpose of performing a simulation with greater accuracy than with idealised Computer Aided Design (CAD) based simulations. Whilst IBSim originates in the biomedical field, the wider adoption of imaging for non-destructive testing and evaluation (NDT/NDE) within the High-Value Manufacturing (HVM) sector has allowed wider use of IBSim in recent years. IBSim is invaluable in scenarios where there exists a non-negligible variation between the ‘as designed’ and ‘as manufactured’ state of parts. It has also been used for characterisation of geometries too complex to accurately draw with CAD. IBSim simulations are unique to the geometry being imaged, therefore it is possible to perform part-specific virtual testing within batches of manufactured parts. This novel review presents the applications of IBSim within HVM, whereby HVM is the value provided by a manufactured part (or conversely the potential cost should the part fail) rather than the actual cost of manufacturing the part itself. Examples include fibre and aggregate composite materials, additive manufacturing, foams, and interface bonding such as welding. This review is divided into the following sections: Material Characterisation; Characterisation of Manufacturing Techniques; Impact of Deviations from Idealised Design Geometry on Product Design and Performance; Customisation and Personalisation of Products; IBSim in Biomimicry. Finally, conclusions are drawn, and observations made on future trends based on the current state of the literature.

Full text

Vol.:(0123456789)
1 3
Archives of Computational Methods in Engineering
https://doi.org/10.1007/s11831-022-09836-2
REVIEW ARTICLE
A Review ofImage‑Based Simulation Applications inHigh‑Value
Manufacturing
LlionMarcEvans1,2 · EmrahSözümert1 · BethanyE.Keenan3 · CharlesE.Wood4 · AntonduPlessis5,6
Received: 8 April 2022 / Accepted: 15 October 2022
© The Author(s) 2023
Abstract
Image-Based Simulation (IBSim) is the process by which a digital representation of a real geometry is generated from image
data for the purpose of performing a simulation with greater accuracy than with idealised Computer Aided Design (CAD)
based simulations. Whilst IBSim originates in the biomedical field, the wider adoption of imaging for non-destructive test-
ing and evaluation (NDT/NDE) within the High-Value Manufacturing (HVM) sector has allowed wider use of IBSim in
recent years. IBSim is invaluable in scenarios where there exists a non-negligible variation between the ‘as designed’ and
‘as manufactured’ state of parts. It has also been used for characterisation of geometries too complex to accurately draw with
CAD. IBSim simulations are unique to the geometry being imaged, therefore it is possible to perform part-specific virtual
testing within batches of manufactured parts. This novel review presents the applications of IBSim within HVM, whereby
HVM is the value provided by a manufactured part (or conversely the potential cost should the part fail) rather than the actual
cost of manufacturing the part itself. Examples include fibre and aggregate composite materials, additive manufacturing,
foams, and interface bonding such as welding. This review is divided into the following sections: Material Characterisation;
Characterisation of Manufacturing Techniques; Impact of Deviations from Idealised Design Geometry on Product Design
and Performance; Customisation and Personalisation of Products; IBSim in Biomimicry. Finally, conclusions are drawn,
and observations made on future trends based on the current state of the literature.
Abbreviations
AM Additive manufacturing
AR Augmented reality
CAD Computer aided design
CAM Computer aided manufacturing
CBCT Cone-beam computed tomography
CDAM Cyber design and additive manufacturing
CDM Continuum damage models
CFD Computational fluid dynamics
CFRP Carbon fibre reinforced plastics
CMD Custom-made device
CPU Central processing unit
CT Computed tomography
DIC Digital image correlation
EBM Electron beam melting
FEA Finite element analysis
FIB Focussed ion beam
FVM Finite volume method
GDL Gas diffusion layers
GPU Graphical processing unit
GTN Gurson–Tvergaard–Needleman
HVM High-value manufacturing
IBSim Image-based simulation
LBM Lattice Boltzmann methods
L-PBF Laser powder bed fusion
MDCG Medical device coordination group
MDPS Medical device production systems
ML Machine learning
MMPS Modified-moving particle semi-implicit
MPL Microporous layer
MRI Magnetic resonance imaging
* Llion Marc Evans
llion.evans@swansea.ac.uk
1 Faculty ofScience andEngineering, Swansea University,
SwanseaSA18EN, UK
2 United Kingdom Atomic Energy Authority, Culham Science
Centre, Abingdon, OxfordshireOX143DB, UK
3 Cardiff School ofEngineering, Cardiff University,
CardiffCF243AA, UK
4 School ofMechanical & Design Engineering, University
ofPortsmouth, PortsmouthPO13DJ, UK
5 Object Research Systems, MontrealH3B1A7, Canada
6 Research Group 3DInnovation, Stellenbosch University,
Stellenbosch7602, SouthAfrica

Ll.M.Evans et al.
1 3
NDE Non-destructive evaluation
NDT Non-destructive testing
OSSM Optical serial sectioning microscopy
PEM Polymer electrolyte membrane
PMD Personalised medical device
R&D Research and development
RAM Random access memory
RANS Reynolds-averaged Navier–Stokes
RVE Representative volume element
SaMD Software as a medical device
SEM Scanning electron microscopy
SLS Selective laser sintering
TEM Transmission electron microscopy
TRL Technology readiness level
XCT X-ray computed tomography
XFEM Extended finite-element method
μCT Micro computed tomography
STL Standard tessellation language
1 Introduction
Image-Based Simulation (IBSim) or modelling can have dif-
fering meanings depending on the context. In the case of this
review, we define IBSim as “engineering simulations based on
3D geometry captured by some form of imaging technique”.
This review focuses on applications of IBSim within high-
value manufacturing (HVM), where the simulation techniques
typically used are Finite Element Analysis (FEA)1 or Compu-
tational Fluid Dynamics (CFD), but IBSim can include use of
any geometrically based numerical method. That is, improv-
ing the accuracy of engineering simulations with the use of
ultra-high resolution non-idealised model geometries which
estimate the performance of components ‘as manufactured’
rather than ‘as designed’. In this context, IBSim does not mean
1D modelling (or systems modelling) based on measurements
obtained by imaging as input parameters, e.g., performing
image analysis of a video monitoring the flow of raw material
stock to provide measurement data for use in an algorithm
which estimates product yield during processing. IBSim is
considered an aspect of ‘digital twin’ technology being devel-
oped for the smart manufacturing methods of Industry 4.0.
The IBSim workflow can be broadly divided into four
main stages as shown in Fig.1:
1. Digitisation of parts through a volumetric or surface
imaging technique.
2. Conversion of the image into virtual geometry.
3. Preparation of the geometry into a simulation ready for-
mat.
4. Image-Based Simulation, visualisation, and post-pro-
cessing.
Due to these different stages, it is a highly multidiscipli-
nary process involving the fields of microscopic imaging,
image analysis, high performance computing and data sci-
ence, engineering simulations, material science and increas-
ingly machine learning. The combination of such a broad
field of disciplines can in itself be a challenge and barrier
to adopting IBSim. The initial stage, i.e., 3D scanning tech-
niques for producing a volumetric or surface image, can
range from a topological scan using methods such as:
– Laser scanning
– Structured light scanning
– Ultrasound
– Photogrammetry
– CMM
To full 3D mapping with techniques like:
– Computed Tomography (CT)
– Magnetic Resonance Imaging (MRI)
– Confocal Microscopy
– Optical Serial Sectioning Microscopy (OSSM)
– Focussed Ion Beam-Scanning Electron Microscopy (FIB-
SEM)
– Serial Block-Face Scanning Electron Microscopy (SBF-
SEM)
– Transmission Electron Microscopy (TEM)
Each technique has its own strengths and limitations and
usually the size and material of the item being imaged, and
the context will dictate which method is most appropriate.
The resolution of the IBSim geometry will inevitably depend
upon the resolvability of features within the image on which
it is based. It is, therefore, essential to select the most appro-
priate technique and acquire the best resolution feasible for
the given circumstances. Although some corrections may
be applied with image-processing methods, there are no
replacements for following best practices for the imag-
ing technique of choice. The most widely used acquisition
method is CT, due to its non-destructive nature, combined
with high resolution capabilities matching well the typical
requirements for IBSim investigations.
Once the 3D data has been acquired it must be converted
into a virtual representation of the geometry to allow simula-
tions to be run. Topological scans, e.g., from laser scanning,
are the simplest forms of imaging data to create IBSim mod-
els. Since they only capture the external geometry with no
internal features (e.g., micro-pores or inclusions), they are
1 The term FEA is often used interchangeably with Finite Element
Method (FEM) although they differ slightly in meaning. For consist-
ency FEA will be used throughout.

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
relatively small datasets, but still significantly larger than
Computer Aided Design (CAD)-based geometries.
These will often be formed of point clouds, which are
a collection of cartesian coordinates representing the sam-
ple surface. Techniques like photogrammetry can provide
additional information such as colour, which facilitates
distinguishing between materials in multi-phase samples.
Post-processing methods are used to interpolate between the
points and define surfaces. Smoothing algorithms are often
used to ‘clean’ data and remove spurious points or fill in
voids in the data.
Full 3D volumetric images are data-rich and, depending on
the imaging method, can include features of interest that are less
than 1/1000th the size of the parent sample. The images typi-
cally consist of a discretised voxel domain, with each voxel (3D
pixel) providing some information about that location in space.
For example, in conventional X-ray CT (XCT), a voxel provides
information about signal attenuation at that location that can be
used to infer material density [1]. When the data is rendered as
an image, rather than a three-dimensional matrix, the attenuation
is visualised by being assigned a given colour or grey scale. This
can be visualised with volume rendering or with 2D images as
cross-sectional slices through the part, as shown in Fig.2.
A segmented volumetric image will still consist of a vox-
elised domain, but with each voxel having a phase number
rather than greyscale value. Fig.3 shows examples for an
image of a lemon fruit, segmented to increasing level of
detail. In addition to the examples shown in Fig.3b−d it
may be possible to segment many more phases up to what is
resolvable with the available image resolution (e.g., separat-
ing the albedo and flavedo).
Many segmentation approaches and software solutions
exist, from fully manual voxel ‘painting’ on a slice-by-slice
basis to semi and fully automated methods assisted by image
processing algorithms [2, 3]. In practice, more complex
images (i.e., sample geometry, number of phases, level of
noise, and artefacts) tend to require more manual interaction.
Once a voxel geometry has been defined, it is possible
to perform simulation analysis directly on this data. This is
often the approach of ‘first pass’ or ‘rapid turnaround’ mesh-
based methods by using the voxels as hexahedral elements,
for example with the Finite Difference Method. For more
in-depth analysis with mesh-based methods, such as FEA
and CFD, it is usually desirable to perform some preparatory
steps such as smoothing, mesh validity and quality checks,
and mesh refinement and/or partitioning.
If the preparation of the IBSim mesh has been carried
out effectively, the running of the actual FEA/CFD analysis
should not differ significantly from a CAD-based simulation.
There are still some considerations worth noting. The main
Fig. 1 Schematic showing the broad stages for an IBSim workflow
which, in this instance, converts X-ray Computed Tomography data
into an FEA simulation. This example is a metallic component from
a heat exchanger, where the geometry and quality of bonding at the
interfaces are integral to the part’s thermal performance

Ll.M.Evans et al.
1 3
additional consideration should be how to work with data
volumes which are orders of magnitude greater than conven-
tional CAD-based models. This includes use of computing
hardware of sufficiently high enough specifications (CPU
cores, RAM, GPU) and software workflows (usually paral-
lelised) that make efficient use of this hardware for simula-
tion, data analysis and visualisation.
As previously stated, the focus of this review is IBSim
applications within HVM, and at this juncture it is worth
providing a disambiguation for the commonly misused
term ‘HVM’. A report by the Institute for Manufacturing,
Cambridge [4] states that a simple definition of HVM is
not possible and, rather, sets out a framework by which to
contextualise it. Their framework broadly defines four types
of manufacturing which can be considered ‘high-value’: ser-
vice led producers; product manufacturers; service manu-
facturers; and system integrators. They explicitly make a
distinction between manufacturing and production “A key
point in defining HVM is that manufacturing is not produc-
tion and vice versa.”. That is, if used for its true meaning,
HVM also includes stages from research and development
(R&D) to ongoing post-production services (in both the
physical and digital realms).
For that reason, HVM is predominantly used in this
review to mean the value provided by a manufactured part
(or conversely the potential cost should the part fail) rather
than the actual cost of manufacturing the part itself. For
example, a critical part in a satellite, whose mission value
is estimated at hundreds of millions of dollars, might only
cost a few hundreds of dollars to manufacture, but the impact
of failure could lead to a catastrophic loss. In such circum-
stances, it is prudent to spend substantial effort in performing
non-destructive testing and evaluation (NDT/NDE) to build
confidence that the particular part in question will perform
as expected. The cost of this effort might be greater than the
cost of manufacture. No manufactured parts are ever defect
free, if investigated at sufficiently small scales defects are
always found to be present. The important outcome of NDE,
therefore, is to quantify the ‘effect of the defect’ to build a
better understanding of what limits a given part should be
operated under. IBSim models are data rich, giving unprec-
edented insight into localised fluctuation in behaviour due to
micro-features as well as their global impact. High resolution
visualisation allows researchers to investigate these in detail.
It is worth noting that IBSim’s roots lie in the biomedi-
cal field, which can primarily be attributed to the fact that
this is also the field that has been a substantial driver for the
development of volumetric imaging, such as XCT and MRI.
It is difficult to identify the first instances of IBSim, however,
early work used external measurements of patients to amend
Fig. 2 Visualisations of a carbon fibre-carbon matrix composite: a photograph, b X-ray radiograph, c volume rendering of XCT data, d 3 ortho-
slices in the xy, xz and yz planes from XCT data

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
CAD-based models and thus make them patient-specific [5].
This progressed to using internal measurements from volu-
metric images [6] as XCT and MRI became more prevalent,
it subsequently led to full conversions of volumetric images
directly into simulation geometries [7]. As could be expected,
the use of IBSim within the industrial sector coincided with
the increased usage of imaging methods, such as micro com-
puted tomography (μCT), which has seen a growth of greater
than 10% year-on-year over the past decade [8]. IBSim is also
used in other fields of research, such as geology (largely in
relation to the oil & gas industries) [9], archaeology [10], and
palaeontology [11]. This review will restrict itself to applica-
tions within HVM other than select examples from biome-
chanics, where a manufactured part is used in the medical
field. However, the review should also be useful to readers
interested in the other aforementioned fields because there is
much in common between the methodologies. This review
is the first of its kind for HVM and aims to give a thorough
overview of the literature to date rather than an update on
recent publications alone.
2 Review ofHVM Applications ofIBSim
IBSim is already being used within research and develop-
ment (R&D) cycles to accelerate development by provid-
ing additional insight at various stages [12–14]. The pro-
gress along the R&D cycle of producing a new concept is
described by its ‘technology readiness level’ (TRL), which
is a method of categorising its maturity stage. These levels
range from the conceptual stage (TRL 1) to full production
with a proven in-service track record (TRL 9).
To increase efficiency in R&D cycles it is desirable to
accelerate progress through the TRLs. Within manufactur-
ing, identifying optimal products (their design, material
selection, usage parameters etc.) is achieved by iteratively
down-selecting candidates through testing. Much of this
development process is constrained by available resources.
That is, the number of candidates which may be consid-
ered are limited by costs and time. Virtual testing through
computational simulation techniques have increasingly been
facilitating the R&D process [15]. With simulations it is
Fig. 3 Schematic demonstrating various levels of detail possible when segmenting a complex object. a Photograph of a lemon cut in half and
b–d image segmented into increasing number of phases: b background, fruit; c background, peel, interior; d background, peel, flesh, seed

Ll.M.Evans et al.
1 3
possible to iterate through many more designs quickly and
cheaply without needing to prototype concepts.
However, there exists a gap between observations dur-
ing simulation and experiments [16]. As such, simulations
are used as first stage guidance but there is still a heavy
reliance on experimental testing during R&D. Improved
accuracy in simulations could lead to more rapid R&D
development. IBSim is one approach that can close the
gap between simulation and experiment [16]. By digitising
a real prototype, microscale accurate simulations can be
carried out on the part ‘as manufactured’ rather than ‘as
designed’. This means its geometry is no longer idealised
and simulations account for impact on performance due to
manufacturing processes by inherently including features
such as deviations from tolerance and micro-porosity. A
flowchart is shown in Figure4.
Within R&D, IBSim allows one manufactured prototype
to be tested to destruction multiple times by using a virtual
representation of the manufactured part which is faithful on
the microscale [17]. In addition to replicating laboratory
testing through simulation, it is possible to extrapolate to
scenarios more representative of real-world conditions e.g.,
increased number of cycles, real loading rates and values,
complex loading with multiple mechanisms. This is because
IBSim testing is not constrained by the limitations of the
laboratory. That means much more valuable data can be
obtained from a single prototype, significantly reducing costs.
Through being able to directly compare results from
experimental and simulation results, IBSim benefits from
verified results with increased confidence values compared
with simulations using idealised geometries. This is invalu-
able within the industrial sector when simulating conditions
outside what can be tested in the lab. IBSim is also used in
materials development to perform virtual characterisation
to reduce the number of physical tests and thus the volume
of material required.
The process of manufacturing novel materials can often
be a rapid process which only requires the variation of some
parameters during fabrication. This can create different
microstructures which leads to different material properties.
However, the process for characterising the new properties of
the material can involve a significant effort and thus cost. If
it is desired to measure a range of properties, this can require
fabrication of many samples for a series of experimental tests.
By using IBSim, it is possible to perform virtual testing
with a suite of simulations that emulate laboratory material
characterisation from one manufactured block of material
that is digitally ‘cut’ to the required dimensions. This way,
Fig. 4 Flowchart demonstrating the relationship between ‘as designed’ simulations and physical testing conventionally used with more novel ‘as
manufactured’ virtual testing with IBSim

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
new materials can be rapidly characterised to identify the
strongest candidates. Furthermore, for method and model
validation, actual test samples can be replicated digitally
for a direct comparison of physical and virtual test results
leading to an improved level of confidence.
Because of the digitisation process from a real material,
IBSim can be used for simulations of extremely complex
geometries such as fibre composites or foams with no need
for idealisation. Additionally, it is possible to digitally alter
the structure to see what impact this has on the properties.
For example, the volume of pores or the thickness of foam
ligaments could be increased/decreased to investigate the
benefits of imposing additional control on the material
processing.
To reflect those categories of IBSim usage within HVM,
this review has been structured accordingly. Firstly, IBSim’s
use for material characterisation is presented followed by
how IBSim has been applied to optimise manufacturing
processes. Next, case studies using IBSim to investigate
how deviations from idealised design on both the micro
and macro scales impact product design and finalised part
performance. Also included are two final sections on how
IBSim is used in a broader sense to improve the customi-
sation and personalisation of products and how it is con-
tributing to the field of biomimicry in manufacturing. The
review is concluded by looking at growing trends in IBSim
for HVM that are showing significant promise.
2.1 Material Characterisation
Image-based models and their simulations have been dem-
onstrated to be highly useful in determining morphological
features of materials and, consequently, effective material
properties at various length scales ranging through mac-
roscale (> 10mm), mesoscale (0.1–10mm), microscale
(0.1–100 µm), and nanoscale (< 100nm). The application
of IBSim for material characterisation is used in three main
areas: macroscale topology; homogenisation; and the impact
of microscale features. Macroscale investigations provide
researchers with bulk material properties (e.g., stress–strain
relationships, effective Young’s modulus, Poisson’s ratio,
plastic strength). In cellular materials, such as foams, the
cell morphology (e.g., cell size & shape) and topology (e.g.,
type of cell such as open/closed wall cells or cell connectiv-
ity) can be characterised at the mesoscale [18]. Image-based
numerical models of heterogenous materials (cellular mate-
rials, multiphase rocks, asphalt, fibre reinforced composites
etc.) at the mesoscale can be utilised for homogenisation of
material properties by using unit cells or representative vol-
ume elements (RVE). The resultant material properties are
subsequently used as input for macroscale numerical simu-
lations of larger parts or components. Through analysis of
μCT images of porous media, such as rocks, it is possible to
measure microstructural features (e.g., pore-size distribu-
tions, network connectivity, micro-cracking).
One image-based approach is to gather this type of sta-
tistical data about a material’s microstructure which is then
used in an analytical method to predict its macroscopic
response. For this to be a robust approach, it requires col-
lecting a statistically significant amount of data and thus
provides the response which can be expected on average.
The direct conversion of microscale images into simula-
tions makes predictions with improved accuracy about the
specific part which has been imaged [19]. The limitation in
the direct conversion approach is that, due to results being
part-specific, a new model is required for each part.
2.1.1 Deformation, Damage, andFracture Performance
ofMaterials
2.1.1.1 Heterogenous Composite Materials IBSim can be
used for material characterisation of heterogeneous materi-
als such as asphalt mixtures [20] or concrete [21]. For exam-
ple, Fig.5a illustrates a methodology presenting the stages
in order to compute shear modulus using image-based 2D
and 3D micromechanical FEA models of asphalt mixtures at
high operating temperatures. Where the models were com-
pared with experimental results the 3D models were found
to be more accurate than their 2D counterparts. Fig. 5b
demonstrates a 2D section of the material, where different
material phases (aggregate, mastic, and air void) were dif-
ferentiated by an image-processing method. In Fig.5c the
undeformed and deformed micromechanical model of the
asphalt mixture, which was subjected to a horizontal surface
shear load, is displayed in 3D. Another example of heterog-
enous materials is concrete composed of aggregate, cement
mortar and pores. A micromechanical FEA model based on
CT images of concrete using MATLAB® was presented to
account for micro-damage mechanisms [21] and, the same
model was used to improve on the limitations of approaches
using statistical random aggregate models. Due to the high
computational cost associated with a full IBSim model at
the smaller scales, a multiscale approach was followed. In
another study, the model included a discrete image-based
mesoscale region, where the main concentrations of stress
were found, and a homogenised macroscopic lattice region
for prediction of mesoscopic crack growth in three-point
bending of concrete [22]. In another, a different image-based
approach was used via a two scale homogenisation method
aimed to predict effective elastic properties of high-perfor-
mance fibre reinforced concrete, where the elastic moduli of
each of the constituents was measured by physical micro-
indentation tests [23]. The microscopic level homogenisa-
tion focussed on the mortar of sand, cement paste and small
pores in a range of 10–600µm, whereas the mesoscopic
level homogenisation focussed on a RVE (a 20mm cube)

Ll.M.Evans et al.
1 3
of fibres and pores in a range larger than 600µm. In a fur-
ther IBSim application example in this field [24], mesoscale
damage and fracture behaviour of concrete based on in-situ
CT images was simulated in tension and compression with
continuum damage plasticity, which elucidated crack initia-
tion and propagation in a complex microstructure of aggre-
gate, mortar and initial voids and cracks. Readers interested
in further examples of IBSim applications with cement and
concrete are directed towards research by Wang etal. [25]
and a review on the subject [26].
2.1.1.2 Orthotropic Materials In addition to directly convert-
ing volumetric images into simulation geometries the infor-
mation about the morphology of materials produced by 2D
or 3D image-processing can be used as input data to feed
stochastic models. For orthotropic materials, such as fibre
reinforced composites or wovens, examples of statistical and
mean morphological characteristics are: orientation distribu-
tion of fibres; density of compound materials; pore size and its
density in foams [27]. Three dimensional images of ceramic
matrix textile composites were obtained by synchrotron X-ray
µCT to perform statistical analysis of geometrical and spatial
features of fibre tows in 3D woven architectures [28].
Realistic-virtual textile composite specimens with 3D
tows were generated by using the experimental statistical
data with deviations [29] and a Monte Carlo based algorithm
[30], where textile reinforcements are represented as 1D loci
in 3D space. Fig.6a shows a schematic for generating virtual
C/SiC woven composite specimens using statistical descrip-
tion reported in the same source (stage-1) and virtual speci-
men generator (stage-2). The method on how to compute 1D
tow loci based on Markov Chain Algorithm [30] is briefly
as follows. First, the extracted 1D tow loci were shown par-
tially embedded in a CT volume in Fig.6b, then the com-
posites were swept along the tow loci paths to generate the
3D virtual specimens (Fig.6c). Such realisations based on
statistical data of actual samples enable to create as many
different FEA models, which fulfil the statistical description,
as possible. As noted in Fig.6c, the computational model
demonstrates a homogenised microstructure at fibre scale;
however, the fibre tows are represented in a realistic way.
Deformation and damage responses of materials can
be simulated at macroscale or mesoscale with Continuum
Damage Models (CDM). For instance, Badel etal. ana-
lysed woven textile reinforced composites at the mesoscale,
where bundles of fibres in yarns were homogenised with
an assumption that yarns are transversely isotropic in the
Fig. 5 a A methodology for the
development of micromechani-
cal model of asphalt mixtures
and simulations (redrawn from
[20]); b 2D image of asphalt
mixture before FE-meshing
[20]; c FEA simulation of
image-based heterogenous
asphalt mixture under shear
load [20]

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
direction perpendicular to their fibres [31]. Similarly, fibre
reinforced polymer composites were modelled by generating
a mesoscale model from XCT data which was coupled to a
macroscale model [32].
Constituent fibres can be detected in composite materials
by fibre-segmentation algorithms [33], where fibres can be
tracked in 3D using a Kalman-filter estimator, for further
numerical investigations. Fig.7a shows the reconstructed
volume of a fibre composite and Fig.7b showsits CAD
rendition with orientation distributions generated from the
image data. More recently, deep-learning procedures have
been used to automate segmentation of 3D CT images from
fibre reinforced ceramic composites composed of fibres and
matrix in the same material (SiC) [30]. This same study
managed to segment matrix cracks in in-situ tensile loading
tests with influence of nonuniform spatial fibre distribution.
Ali etal. proposed a methodology to create IBSim FEA
models from µCT images of two 2D woven carbon–carbon
composites for nuclear applications [34]. These composites
consist of multiple phases of the same material. The material
properties of the separate phases are required as input data
for FEA simulations, in this case the mechanical properties
were determined experimentally by physical nano-indenta-
tion material characterisation tests. Kishimoto etal. used
IBSim to study inhomogeneous local deformation of rubber
matrices, where uniformly and non-uniformly distributed
Fig. 6 a A schematic for generating virtual specimens based on statistical data of real samples (redrawn from [29]) and b μCT image of a C/SiC
woven composite and the centres-of-mass of tow sections [30]; c a 3D rendered virtual specimen [29]
Fig. 7 a Volumetric rendering
from CT data of fibre compos-
ites and b CAD rendition of
their fibres [33]

Ll.M.Evans et al.
1 3
silica particles were embedded [35]. By using IBSim mod-
els, more accurate results were obtained showing that the
inhomogeneous local stress fields strengthened the mechani-
cal properties, such as the ultimate strength, of the material.
2.1.1.3 Additive Manufacturing Additive manufactur-
ing (AM) is one area that may particularly benefit from
IBSim due to the significance of variation between the ‘as
designed’ and ‘as manufactured’ states. AM allows highly
complex parts to be manufactured, which is one of its main
benefits compared to traditional manufacturing approaches,
allowing such designs as biomimetic brackets or cellular
structures for light weighting advantages. This complex
design makes prediction of properties challenging for tradi-
tional FEM, especially when manufacturing deviations and
flaws can occur.
One of the most important features is porosity within the
manufactured material, whether intentional or not, and gain-
ing an understanding of the influence of such pores on the
mechanical properties of the material. As an example of a
study using IBSim in this context, the tensile deformation
mechanisms of porous sintered 316 L steel were investi-
gated with three image-processing approaches [36]. The first
two approaches were based on artificially-underestimating
the material properties of the material [37] and altering the
porosity of IBSim models by changing the greyscale thresh-
olds of shapes to meet experimental stress measurements
[38], and the third was a novel approach compensating the
effect of CT inaccuracy in porous materials on numerical
analysis by modifying μCT images and separating shapes
of fissures and small pores. The third approach yielded the
most realistic porous microstructures and consequently val-
ues in stress distributions when compared with experiments.
With this increased level of detail, it was also possible to
estimate the critical stress locations where fracture on mac-
roscopic scale was most likely to occur. This method was
found to be computationally expensive as well as having
issues with convergence. Effective material properties (e.g.,
Young’s modulus, yield strength, shear modulus and Pois-
son’s ratio) at the macroscopic scale and local stress & strain
distributions were found to be strongly influenced by the
image-processing approach applied, with the third approach
providing the most accurate results.
Porosity-induced stress concentration on fatigue scatter
due to remnant porosity within components manufactured
by laser powder bed fusion (L-PBF), an AM method, was
analysed with IBSim [39]. The CT scans of the AM compo-
nents, with a range of pore sizes, were post-processed in a
workflow which included VG Studio Max by Volume Graph-
ics®, Simpleware®, and a + CAD® subroutine to generate
image-based meshes. Then Abaqus® was used for FEA to
compute stress concentration factors using an elasto-plastic
material model. The analyses around the pores showed that
small pores near to the surface were more detrimental to
the material than the pores deeper within the components.
In a further example, deformation and damage behaviour
of tin (Sn) solder alloys were simulated with image-based
FEA models reflecting the exact geometry of pores in solder
joints, and the ductile damage mechanisms (crack nucleation
and propagation) were described with a damage model to
a degree of accuracy not previously possible [40]. The use
of ‘stitching tomography’ enabled Amani etal. to increase
their detector’s field of view and thus image greater volumes
whilst retaining resolution [41]. When coupled with IBSim,
this allowed them to analyse the compressive response of
AM 3D lattice structures on both the macroscale, i.e., global
lattice structure, and microscale, in which micropores and
imperfections in struts were captured. Damage and fracture
behaviours of the ductile struts were homogenised by imple-
menting Gurson-Tvergaard-Needleman (GTN) damage-cou-
pled plasticity, informing the accumulative porosity com-
puted from high-resolution CT. This presents a very accurate
solution to this highly nonlinear multiscale problem and the
predicted fracture locations were in a good agreement with
experimental investigations. In a similar study, the same
two-scale modelling approach and microstructure-informed
GTN plasticity model was also practised for open-cell alu-
minium foams subjected to tensile loading [42].
For full size components the length scale of interest is
usually the macroscale: it has been demonstrated that effec-
tive material properties such as time-dependent or independ-
ent elastoplastic parameters (stress–strain relations), plas-
tic strength (collapse stress) on this scale can be obtained
through IBSim. For instance, damage evolution of L-PBF
-Printed AlSi10Mg alloys was simulated with CAD-based
(as-designed) and image-based (as-manufactured) FEA
models of tensile specimens, directly extracted from µCT
images by using Avizo®, in order to assess the role of imper-
fections on mechanical properties [43]. The as-manufactured
FEA model met the expectations better in comparison to as-
designed model by predicting the higher failure strain due to
the geometrical defects present in the parts.
A strong growth area in AM, especially in L-PBF, is the
use of insitu monitoring. This refers to imaging of the melt
pool and/or the entire build area with optical and infrared
(IR) cameras. The presence of defects is highlighted in this
way directly when they occur in the layer-by-layer process.
It is possible to generate full 3D model data from this in-situ
generated imaging data, that could be used in the same way
as XCT data for further simulation [44, 45].
2.1.1.4 Foams μCT-based FEA models of zirconia foams
were developed to correlate its macroscopic mechanical
response to microscopic features such as thickness of cell
struts (i.e., walls), strut waviness and material properties
of struts [46]. To do this, the local elastoplastic properties

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
were obtained at strut level with physical micro-indentation
tests and plastic deformation was implemented with an iso-
tropic plastic model (Von Mises yield criteria). With a simi-
lar motivation, physical nano-indentation tests were carried
out on stainless steel walls of cellular materials to obtain
material properties as inputs to IBSim models [47]. How
to collect material property data appropriate to the scale
in question is a major outstanding question in the field of
IBSim. This is because of the transition from a homogenised
continuum approach at the macroscale to a more granular
one on the meso to nano scales to a fully discretised one at
the atomistic scale.
Cho etal. conducted a multiscale FEA analyses on tita-
nium (Ti) foams with periodic architecture, the details of
which were obtained with µCT images [48] (see Fig.8a).
Virtual uniaxial compression tests on foam specimens
(macroscale investigation) were simulated with FEA and
the computed local deformation gradients were imposed on
the boundaries of periodic unit cells of Ti foam (microscale
investigation), illustrated in Fig.8b. The computational
expense of this complex mechanical problem decreased
significantly by using this multiscale approach, despite that
the microstructural inhomogeneities were included in the
FEA model which made improved its accuracy over CAD-
based models.
X-ray-based FEA models of closed-cell metallic foams
were compressed under large deformation with nonlinear
elastoplastic material behaviour of foam walls, where a
watershed method and geodesic reconstruction were used
for isolation of cells and identification of missing walls
[49]. The steps of this investigation are presented in Fig.9,
where the microstructural deformation and damage patterns
of IBSim models were compared to CT images collected
in-situ during physical experiments. The collapsing cell-
zones in the IBSim models accurately matched that of the
physical experiments. The numerical model reproduced the
experimental plastic band well in addition to capturing the
buckling, bending and fracture behaviour of the cell walls.
Comparable research reported the simulation of deformation
and plastic collapse mechanisms in closed-cell aluminium
foams with contact interaction [50]. Whereas, damage and
fracture behaviour of Cordierite-mullite-alumina ceramic
foams were simulated with CT-based FEA models and used
to compute Young’s modulus and plastic collapse stress [51].
Once more, it was found that the accuracy of the IBSim FEA
modelling approach is related to the resolution of the X-ray
images used.
Veyhl etal. computed the effective mechanical properties
(elasticity moduli and yield strength) of an open-cell porous
sponge with porosity of 91–93% and closed-cell foam with
porosity of 80–86% by using µCT-based FEA models in the
commercial software MSC Marc® (MSC Software Corpo-
ration USA) [52]. The elastoplastic behaviour of wall mate-
rial of cells was modelled by von Mises yield criterion, the
anisotropic material behaviour in orthogonal directions was
captured by simulations of uniaxial compression tests. Effec-
tive strains and stresses were computed from total forces and
geometric stretches over the loading planes, which is known
as RVE-based homogenisation of material properties.
2.1.1.5 Random Fibre Networks One of the well-known
heterogeneous porous materials, to which IBSim is well
suited, is nonwovens. They are composed of randomly
distributed fibres, where fibres form contacts between
each other. Understanding of their mechanical behaviours
and predicting their effective properties are cumbersome
because of their complex microstructures and randomness.
Therefore, non-destructive characterisation techniques are
used to determine their microstructural features such as ori-
entation and length distributions of constituent fibres from
2D or 3D images. For instance, 2D orientation distribution
of the fibres was computed from Scanning Electron Micros-
copy (SEM) or CT images through Hough-transform based
algorithms, where fibre edges are detected [53] and this data
was later used in another study [54] to simulate tensile and
damage behaviour of planar random fibre networks. Some
researchers directly processed μCT images of these materi-
als for generation of their computational models [55, 56],
Fig. 8 a 3D CT reconstruction of Ti foam with a unit cell and b its homogenised FEA model under compression applied to the unit cell [48]

Ll.M.Evans et al.
1 3
one of which is shown in Fig.10a, and some used such mod-
els in their inverse parameter identification studies to find
bond strength of fibres [57]. What is more, fibre length and
diameter distributions of short fibres were computed from
μCT images of wood fibre networks and 3D discrete FEA
models were generated by implementing these obtained dis-
tributions [58] and similarly this is demonstrated with a 3D
stochastic model [59] (see Fig.10b).
2.1.2 Flow andThermal Performance ofMaterials
In addition to characterisation of mechanical properties,
IBSim is also used to characterise other physical mecha-
nisms. The other main use observed in literature is to study
the impact of imaged features on flow and thermal per-
formance. In addition to direct conversion of images into
meshes, there are simpler examples in this field which use
Fig. 9 A schematic of CT-based
experimental and numerical
investigation of closed-cell
metallic foams [49]
Fig. 10 aCT-based FEA model of nonwoven solid under compression [55] and ba statistical realisation (FEA) model of fibre-glass pack [59]

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
measurements from volumetric images as input to models
which are computationally less expensive. For example,
investigating permeability with the resultant pore network
extracted from an image can allow the consideration of a
larger domain than feasible with full-scale IBSim models if
limited by computational expense.
Blunt etal. analysed three different porous materials
(sand pack, sandstone, and carbonate—Portland limestone)
by using X-ray images to extract their pore-scale models
and solved those with the Stoke equations governing flow
behaviour [60]. An example of (a) pore-space and (b) pore
network models of one of the porous materials from that
study is shown in Fig.11. A pore-scale network model was
used in order to determine macroscopic transport proper-
ties and porosity–permeability evolution during reactive
transport processes in a sample reservoir [61]. Bultreys
etal. discussed well-known methods to extract pore-scale
networks and numerical methods (e.g., traditional CFD, Lat-
tice Boltzmann Methods (LBM), Smoothed Particle Hydro-
dynamics) to solve Navier-Stoke’s equations [19]. A review
of pore network modelling for porous media [62] explored
pore network construction approaches and their applications
(e.g., adsorption, dissolution and precipitation). Single and
two-phase flow behaviour of rock samples were simulated
with their pore-network and unstructured meshed models
for prediction of permeability under different wetting condi-
tions [63].
These void statistics also help to improve the accuracy
of micro-mechanics-based constitutive models predicting
deformation, damage, and fracture behaviour of materials.
For instance, Lu and Chan quantified the three dimensional
micro-voids in warm-forging of biocompatible alloys (stain-
less steel 316 L (SS316L) and a titanium alloy Ti6Al4V)
by analysing reconstructed volumes from µCT images [64].
The spatial distribution and number of micro-voids, porosity
was obtained through an advanced segmentation algorithm
in a commercial software VGStudio MAX 2.2®. AM inter-
penetrating phase composites were characterised by μCT
to detect pores in constituent materials (see Fig.12a) and
their interfacial porosity (Fig.12b) for the prediction of
thermal conductivity [65]. Periodic homogenisation theory
was implemented to compute the effects of porosity and unit
cell structure on the effective thermal conductivity with the
COMSOL® Multiphysics software package.
Geometry and connectivity of pores are a dominant fea-
ture of what controls transport properties of porous medi-
ums. Due to this, pore and throat size distributions of Fon-
tainebleau sandstones were measured using synchrotron
XCT images in an earlier work by Lindquist and Venkataran-
gan [66]. Silin and Patzek introduced an algorithm to ana-
lyse the geometry and connectivity of the pore space mor-
phology of sedimentary rock, where pore space and throats
are distinguished by describing them as inscribed spheres
[67]. This work was extended by Dong and Blunt to extract
pore-network connectivity out of voxel-based models, con-
structed from 3D X-ray images, for predicting permeability
of porous medium that depends on pore geometries and wet-
tability [68]. In similar investigations for significantly dif-
ferent materials and applications, pore characteristics were
computed from the X-ray-based computational models of:
bone substitute materials [69], because bone formation over
a scaffold strongly depends on pore configurations; proton
exchange membrane fuel cells [70] to understand and model
two phase flows in a gas diffusion layer; microporous soils,
sand-bentonite mixtures, and precision glass beads for test-
ing different segmentation methods [71]; various soil sam-
ples for quantification of pore size distribution [72]; soils in
Fig. 11 aPore-space image of Mount Gambier; bits pore-network model extracted/computed from CT images [60]

Ll.M.Evans et al.
1 3
the Antaibao Opencast Coal-mine for distribution character-
istics of the reconstructed soil [73].
Houston etal. reviewed existing methods in literature
estimating pore size distribution and analysed artificial 3D
images and actual CT images of various selections of soils
in order to make a comparison of their performance [72].
In general, according to Xiong etal., the methods aiming
to extract pores and their connecting throats in the recon-
structed geometries were listed as (i) statistical reconstruc-
tions; (ii) grain-based models; (iii) direct mapping models;
(iv) regular network models; (v) two-scale pore network
models [62]. Elsewhere, the IBSim approach made it pos-
sible to assess pore characteristics (e.g., porosity, pore size
distribution, throat size distribution) before/after deforma-
tion and damage, and the effect of microstructural changes in
porous medium on flow characteristics such as pressure-drop
and permeability [74].
As flow permeability, which depends on porous micro-
structure, is of great interest in industrial applications,
morphological statistics of porosity, tortuosity and pore
diameters of fibrous media were obtained from high-reso-
lution XCT for use with LBM simulations [75]. These were
conducted over various sizes of RVE’s to compute macro-
scopic transport properties. It is known that the combination
of LBM and XCT have previously been used for simulat-
ing water flow and chemical transport of porous materials
at the pore-scale [76]. Fig.13a demonstrates the effect of
RVE size (or window size) on permeability. Likewise, Kok
solved mass transfer equations at low Reynolds numbers for
image-based flow models of various fibrous media with a
variety of anisotropic fibre distributions (namely, carbon
felts and two different electrospun carbon networks used
in flow electrodes) by using LBM [77]. Some researchers
directly processed SEM images to obtain their 3D compu-
tational models, where the filtration performance of polyu-
rethane nanofibre filters was investigated (see Fig.13b for
an example SEM image and inferred 3D layered model)
[78]. Saturated fluid flow in packed particle beds [79] was
simulated by implementation of LBM in order to calculate
permeability from μCT images. Porous gas diffusion layers
(GDLs) are key parts of hydrogen fuel cells and, in order to
mimic water flow behaviour of the GDLs a pressure drop
was applied to one surface of a µCT-based LBM model to
simulate the formation of water droplets in the porous micro-
structure to represent water–gas surface tension [80]. In an
alternative example, Navier–Stokes and convection–diffu-
sion equations were solved with the Modified-moving parti-
cle semi-implicit (MMPS) method for unsteady and steady-
state flow in a disordered porous media [81]. More recently,
the effectivity of face masks to filter airborne viruses such
as COVID-19 has been of great interest and has also been
investigated with IBSim [82, 83].
Water distribution in the hydrophobic microporous layer
(MPL) of polymer electrolyte membrane (PEM) fuel cells
was computed from image-based pore geometries and oxy-
gen transport mechanisms was simulated through pore-scale
modelling, where the simulated oxygen concentration and
flux values were averaged to the effective diffusion coef-
ficients of RVEs [84]. Convective drying process, a form of
moisture removal mechanism in porous materials, of porous
asphalt was investigated by CFD simulations of 3D IBSim
models with different airflow speeds and Steady Reynolds-
Averaged Navier–Stokes (RANS) k-ε model accounting for
turbulent flow behaviour [85]. Flow behaviour in highly
porous monolithic alumina columns [13] was simulated by
direct CFD models whose porous structure was obtained
from 3D CT and the governing flow equations were solved
with an open-source CFD tool (OpenFoam) in order to
enhance monolith performance.
Cooper etal. performed heat transfer analysis with IBSim
of LiFePO4 electrodes using the finite-volume method in
Star CCM + ® [86]. After reconstructing the 3D volume
from CT data, they converted the heterogenous microstruc-
ture into surfaces (Standard Tessellation Language (STL)
Fig. 12 a AM A356/316L composite in low resolution and its unit cell in high resolution with microporosities (316L in dark and A356 in bright
contrast) and b the unit cell with interfacial porosity in high resolution [65]

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
format). The file was imported into a CFD pre-processing
module in order to volume-mesh the electrode material (first
re-meshing the surface and then volume meshing). The
workflow is shown pictorially in Fig.14a, b and d, e with
Fig.14c, f showing the temperature distribution in relatively
large and small domain models.
Anisotropic thermal conductivity of a sintered metallic fibre
structure with varying porosity was investigated using µCT-
based FEA models [87]. It was shown that the thermal conduc-
tivity is a function of porosity and fibre orientations. Carbon
fibre networks are effective insulators for applications, where
the materials are exposed to high temperatures [88]. The geo-
metrically accurate flow models of these networks, digitised
from μCT images, were used to calculate temperature-depend-
ent permeabilities. Similarly, room temperature conductivity of
carbon fibre networks was computed with voxel-based IBSim
models, where oxidation behaviour and surface reactions were
incorporated into microscale simulations [89].
Effective anisotropic thermal conductivity of a glasswool
insulation material composed of randomly distributed fibres,
the main source of anisotropy, was characterised by FEA
simulations generated from X-ray images by solving 3D
heat equations and applying different temperature distribu-
tions over surface boundaries in order to create and measure
temperature drops [90]. The thermal conductivity of highly
porous metal foams was analysed with IBSim FEA models
and it was numerically proven that the RVE and FEA ele-
ment size are two parameters which have a non-negligible
impact on the virtually measured thermal conductivity, thus
highlighting the importance of performing sensitivity analy-
ses as part of the methodology [91].
Electrochemical performance of carbon felt electrodes of
redox flow batteries is influenced by the microstructure of
carbon felt. CFD simulations were used to investigate com-
pression of the felts and thus predict the increase in pressure
drop due to microstructural changes [92]. High-resolution
µCT CFD simulations of open-cell aluminium foams with
different pore densities, i.e., number of pores per unit vol-
ume, were conducted to predict permeability and effective
thermal conductivity under incompressible flow and steady
state flow conditions [93].
Evans etal. carried out a thermal analysis of a heat
exchanger component (Fusion Energy Monoblock) by using
a hybrid FEA model containing: a graphite foam interlayer
with microscale accuracy directly derived from CT images;
a CAD-based armour and coolant pipe [94] (see Fig.15).
The graphite foam ring layer was digitally ‘cut’ from a larger
block of imaged material, thus being able to rapidly assess
the design without the need for physical manufacturing. In
another study, the thermal response of a carbon fibre com-
posite-copper monoblock was simulated with IBSim FEA
[16]. The model included a debonding region at the carbon
fibre composite-copper interface. By capturing the debond-
ing at this interface, this ‘as-manufactured’ simulation
predicted the loss in thermal conductivity at the interface,
which would not have been included in an ‘as-designed’
Fig. 13 aEffect of RVE size on flow permeability for three different porosity levels (E) [75]; b SEM image of nanofibre network and its 3D
model [78]

Ll.M.Evans et al.
1 3
model. This led to a rise of over 20% in the peak tempera-
tures which consequently would have increased the ther-
mally induced stresses.
2.1.3 Multiphysics Performance ofMaterials
A review of analytical models to predict electrical con-
ductivity in porous media was published by Cai etal. [95].
These analytical modelling approaches such as pore network
Fig. 14 a initial surface; b surface after Boolean subtraction operation; d re-meshed surface with triangular elements; e polyhedral volume mesh
of porous structure of electrodes; c, f heat transfer analyses over different domain volume size of LiFePO4 electrodes [86]
Fig. 15 Virtual manufactur-
ing workflow from graphite
foam interlayer to CAD pipe
and CAD armour with thermal
boundary conditions [94]

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
and percolation modelling rely strongly on the processing of
detailed microstructural images.
Commercial use of solid oxide fuel cells is limited by
technical issues such as thermal gradients across the cell
developed during operation leading to deteriorating the bat-
tery performance [96]. Electrode polarisation losses of solid
oxide fuel cells, associated with the composition of constitu-
ent materials and their microstructure, reduces performance.
The underlying electrochemical processes (e.g., oxygen dif-
fusion in gas phase and charge transfer at the interface elec-
trolyte–electrode material) were investigated for a porous
mixed ionic-electronic conducting cathode by using 3D FEA
models based on reconstruction of focused-ion beam (FIB)
serial sectioning and SEM imaging to produce tomographic
images (i.e., FIB-SEM) [97]. Furthermore, effective electri-
cal conductivity of composite asphalts with randomly dis-
tributed steels in an epoxy was numerically analysed [98].
Also, multiphysics simulations of solid oxide fuel cells
based on 3D micro/nano reconstructions were performed
by taking conformal boundaries between different phases
into account [99].
Zhao etal. performed a comprehensive review on model-
ling approaches for the coupled chemo-mechanical behav-
iour of Lithium-ion batteries at particle, electrode and cell
levels [100]. In this, the capacity loss in these batteries dur-
ing charging/discharging cycles were associated with some
phenomena such as nonlinear elasticity, plasticity, aniso-
tropic mechanical behaviour and phase separation. Hein’s
electrochemical simulations of Lithium-ion batteries relied
on CT-based parametrised stochastic models and non-par-
ametric realisations extracted from reconstructions of CT
volumes [101]. Numerical methods such as LBM have been
used to compute electrical and species transport properties
of lithium-ion batteries in order to develop new products or
optimise their performance [102]. Since lithium-ion batter-
ies experience electrode failures due to diffusion-induced
stresses occurring in charge and discharge, these processes
were simulated by Lim etal. with micro and nano CT-based
FEA models of active particles for different discharge rates
(C rates) and, the non-uniform/complex shape of the parti-
cles increased in induced von Mises and Tresca stresses lead-
ing to failure [103]. Fig.16a shows a microstructural model
of lithium-ion battery anode with boundary conditions and
Fig.16b, the 3D distribution of electrical current through-
out pore space[104]. The workflow starting from a com-
mercial lithium-ion cells (batteries) down to single particles
extracted from reconstructed CT volumes are presented in
Fig.16c. A similar methodology was implemented to predict
the transient stress-fields over the cathode particles of com-
mercial lithium-ion batteries by coupling electrochemical
processes with mechanical ones [105]. The swelling in
LiCoO2 cathodes was studied with coupled electrochem-
ical-mechanical simulations to unfold the mechanisms of
stress generation and the effect of process parameters along
with microstructure on these stresses [106]. Galvanostatic
discharge processes of LiCoO2 cathodes at various C rates
were simulated with 3D IBSim models by Yan etal. [107]. A
comparison between macro and microscale IBSim models of
lithium-ion porous battery electrodes was made in terms of
their electrical conductivity and diffusion [108]. Elsewhere,
mesoscale multiphysics simulations bringing electrochemis-
try, mechanical deformation and transport processes together
in lithium-ion batteries incorporating conductive binder par-
ticles were presented [109–111]. An X-ray-based realistic
3D microstructure numerical model enabled the authors to
obtain the stress accumulation in nickel-manganese-cobalt
(NMC) half-cell, resulting from the phase transitions and
lithium intercalation [112]. Multiscale investigations on
Lithium-ion batteries revealed porosity from X-ray micros-
copy and effective diffusivity as well as tortuosity from com-
puter simulations in GeoDict® [113].
Fluid and electrical flows through reservoir rock samples
accommodating highly complex pores were simulated with
the COMSOL® multiphysics simulation tool [114]. In this
work they conducted a downsampling study, where the sizes
of volume elements were controlled. This indicated that, as
some pores disappear, and the remaining ones alters geom-
etry, fluid and electrical flow patterns were affected signifi-
cantly. Together with this, IBSim had been used to model
water and oil distribution formations in the microstructure
of a porous rock and to investigate the effect of rock wet-
tability on electrical properties [115]. An IBSim approach
enabled the authors to examine the effect of calcite precipi-
tation on the permeability of a porous media with a Stokes
solver (an inhouse solver implemented in Avizo® under
XLabHydro®), where the precipitated particles and porous
media microstructure were captured by µCT and the pore
network was converted into a flow model [116]. Another
open-source software package for porous materials is PuMA
(Porous Microstructure Analysis), computing effective mate-
rial properties such as thermal and electrical conductivities
by using finite difference Laplace solvers [117]. An effective
thermal conductivity of a composite material with aniso-
tropic constitutive phases was predicted with PuMA [118].
The software package can be used for virtually generating a
computational domain of arbitrary porous structures and the
tortuosity of these artificial models or their 3D IBSim mod-
els can be computed with a random walk algorithm [119].
The package was integrated into an image analysis software
Dragonfly® (Object Research Systems, Canada).

Ll.M.Evans et al.
1 3
2.2 Characterisation ofManufacturing Processes
2.2.1 Defects andManufacturing Process Errors
Different manufacturing processes have different unique
defect types inherent to the process which may occur, and
which require attention to minimise their extent in manufac-
tured products—i.e., optimisation of the processes is often
needed. For example, metal casting processes are prone
to shrinkage porosity and gas porosity (shown in Fig.17)
which are formed due to the entrapped gas during the casting
process (gas porosity) and due to inadequate filling of the
casting mould, with subsequent cooling and shrinkage of the
molten material (shrinkage porosity)[120, 145]. These can
be minimised by varying the casting infill velocity, ingate
geometry and location(s) and cooling of the mould. Forma-
tion of microporosity in the solidification process of Sn-Bi
alloys in a copper mould was investigated with X-ray and
FE modelling and the porosity strongly depends on alloy
composition [121]. Similar porosity formation occurs in
plastic injection moulding processes. These defect types
are conventionally detected by NDT methods such as X-ray
radiography (2D) or CT (3D), and may be used to improve
the manufacturing process or may be used for pass/fail deci-
sions for individual parts [27, 122, 123].
Fig. 16 a A computational model of anode microstructure with
boundary conditions [104]; b 3D current stream line distribution
(red and blue colours are ionic and electronic currents, respectively)
[104]; c individual complex particles extracted from reconstructed
volume of commercial lithium-ion batteries for stress analysis [105]

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
In AM, different types of pores are formed with
very small size in comparison to castings and injection
mouldings, but with a wider distribution in the part (see
Fig.18)[120]. This is due to the track-by-track and layer-
by-layer manufacturing process which creates possibilities
for pore formation in smaller regions but more widespread
in all regions of the part. A recent review explains defects
and anomalies in metal PBF in detail [124]. Another paper
focuses on characteristics and variability of defects occur-
ring in metal laser PBF [125].
The above-mentioned examples are illustrative of the
types of porosities of different sizes, morphologies and
distributions that may occur in various manufacturing pro-
cesses, and which can be improved by optimising the manu-
facturing process. Further examples are found in different
studies showing the presence of pores in different materials
and due to different parameters [126–129]. The intentional
variation of manufacturing process parameters shows clearly
the influence of each parameter on pore formation [130], and
in a recent round robin test, different porosity distributions
were found in samples produced in different laboratories
[131].
Other defect types that occur in manufacturing in gen-
eral, besides porosity, are inclusions, cracks, geometrical
inaccuracy, surface roughness, residual stress, and micro-
structural anisotropy or inhomogeneities. All of these defect
types are known to influence mechanical performance, either
by reduction of yield strength, reduction in ductility, or in
lower fatigue strength [132, 133]. They are discussed below
briefly in the context of IBSim and the possibility for process
optimisation.
Surface roughness, indentations, scratches or surface
damage of any kind can influence the mechanical proper-
ties especially acting as fatigue crack initiation sites [134]. In
this work, FEA simulations of the ideal geometry were used
in combination with local surface depression depth, to create
Fig. 17 Examples of casting porosity including (left) shrinkage porosity and (right) gas porosity, image from [120]. The samples are a commer-
cial sand-cast aluminium alloy automotive part (left) and investment cast titanium alloy machined to a tensile dogbone geometry [145]

Ll.M.Evans et al.
1 3
a modified stress intensity factor which correlated well with
fatigue crack initiation site, despite differences in residual
stress and microstructure between samples (due to different
build orientations). Similar work was reported for lattice
structures manufactured by AM [135, 136]. Further work is
needed to make direct simulations utilising the actual sur-
face morphology, as there may be shielding effects where
adjacent depressions or pores create stress shielding or may
enhance the local stress in some places. This was prelimi-
narily investigated already in 2D [137]. Because achievable
CT resolution is limited to sample size, the incorporation of
relatively small features in images and subsequent models
is challenging. A good example is the surface roughness for
a macroscale component—by scanning the whole part the
surface roughness details are not included. Small coupon
samples may be used in addition to the full-scale part, to
provide some inputs, despite its limits (possible variations
from larger part and no direct correlation).
Residual stress is another strong influencer of mechani-
cal performance and is difficult to characterise and incor-
porate into simulation models, because of the challenges in
its measurement [138, 139]. Since the characterisation of
residual stress is either destructive (by hole drilling or simi-
lar methods), or in laboratory instruments only providing
limited depth information, no work so far has incorporated
the influence of residual stress into simulation models, to
the knowledge of the authors. What is often done, however,
is to compare predictions of stress based on manufacturing
process simulations with stress maps obtained from X-ray
diffraction imaging at synchrotron sources [140]. In AM,
much effort is made in process optimisation by thermal
simulations to predict residual stress and minimise this by
simulation and variation of scan strategies [141].
Microstructure of metals (grain sizes, orientations, granu-
lar structure, etc.) is similarly challenging to characterise
non-destructively and hence difficult to incorporate in IBSim
Fig. 18 Examples of porosity in metal AM including (left) lack of
fusion porosity and (right) keyhole porosity, image from [120]. The
samples are small cubes of titanium alloy, manufactured using dif-
ferent process parameters—such cubes are often used to optimise the
process allowing up to 99.99% dense parts under optimal conditions
[125]

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
models. Technically this is possible using destructive imag-
ing and correlating microstructural mechanical properties in
averaged volumetric regions, or non-destructively using dif-
fraction contrast imaging. However, to the authors’ knowl-
edge, there are no examples where direct incorporation of
these approaches with IBSim has been reported yet.
2.2.2 Effects ofDefects
The influence of porosity on fatigue performance has been
reviewed comprehensively [133, 142, 143] and more spe-
cifically for metal AM by Sanaei etal. [144]. Irregular-
shaped pores (as in Fig.17a or 18a) are more detrimental to
mechanical performance, as are larger pores and those closer
to the surface of the part. Cracks (e.g., from manufacturing
induced stresses) are similarly detrimental and more so when
they are closer to the surface or larger in size (or both).
The effects of pores on mechanical performance may
be investigated by IBSim. When compared directly with a
physical test (i.e., performing a simulation of the test on the
digital representation of the sample), the physical results
may be used to verify the IBSim model which can then be
interrogated in greater detail than the results from the physi-
cal counterpart, allowing localised measurements through
the sample’s full 3D volume with microscale accuracy. For
example, the stress distribution around casting pores were
evaluated as shown in Fig.19, before and after tensile testing
[145–148]. Similar work was reported for brackets fabricated
by AM with pores [149], and for pores in high pressure die
castings in recent work [14], and for prediction of mechani-
cal properties in aluminium castings [150, 151], and for
mechanical characterisation of AM nickel–chromium alloy
samples [152]. A Bayesian-based statistical analysis was
conducted for uncertainty quantification of pore distributions
in AM components [153], which was later used for devel-
oping a probabilistic constitutive damage model. A recent
study made use of AM to artificially create defects in tensile
samples, and made use of XCT and IBSim to investigate
the effects of the defects on tensile behaviour [128]. Similar
investigations were reported using artificially induced pores
and XCT [154, 155], though these do not include simula-
tion. The influence of pores on fatigue performance is also
widely acknowledged, as the pores act as stress concentra-
tion locations for crack initiation. This was studied using
in-situ synchrotron imaging in castings with pores, finding
the exact crack initiation location at pore boundaries and
applying IBSim to complement the study [147].
The effect of defects in metal AM was reviewed recently
in the context of XCT imaging insights [132], where it is
evident that most of the small porosity in these materials
influence the ductility of the parts but not the strength,
unless present in excessive amounts (> 1%). It has also been
found that lack of fusion pores with irregular shapes are
especially detrimental to fatigue properties, as are all large
pores near the surface [144]. As one can expect, these influ-
ences become difficult to predict when the part geometry is
complex, and/or when loading scenarios are not simple (e.g.,
multiaxial loading).
In cases where the geometry or loading scenario is com-
plex, simulation is highly valuable for the performance and
property prediction. Examples of highly complex geometries
are cellular porous “lattice structures” manufactured by AM
[156–158]. IBSim models have been used to compare dif-
ferent ideal geometries of such lattices of different designs,
showing differences in permeability and stiffness, factors
important for medical implant applications [159]. This
allows an improved design choice to be made. Engineering
simulation is widely used already to check performance of
designs prior to manufacturing, this is even more important
with complex geometries becoming possible through AM
[160, 161]. It is also possible to incorporate expected defects
into such ideal models to predict the effect of manufactur-
ing defects and predict a critical size of such defects, as was
done for a pore in the middle of a single strut of a lattice
structure [162].
Despite the capabilities of simulations of idealised
“design” models, all manufactured parts inevitably have
some geometrical inaccuracies, defects, and deviation
from ideal design geometry. Here IBSim of realistic mod-
els from μCT or other 3D image data is particularly use-
ful, as the simulation of the actual geometry including its
defects and inaccuracies provides insight into the influence
of these defects on the performance. One example is shown
in Fig.20 where a load simulation was applied to a gyroid
lattice structure manufactured by L-PBF[163], similar to
that reported by Plessis etal. [164]. This highlights the loca-
tions of highest stress in relation to the local rough surface.
Fig. 19 IBSim model before (left) and after (right) tensile testing to
failure [145]

Ll.M.Evans et al.
1 3
AM single lattice struts with process-induced geometrical
imperfections were analysed with IBSim and multiscale
modelling [165]. The use of IBSim in order to incorporate
the influences of porosity and surface roughness into the
predicted performance was previously suggested [166], and
was used to correlate stress concentrations to failure loca-
tions in compression tests [167]. Amani etal. used a similar
approach for lattice structures incorporating defects, surface
roughness and using a GTN model to include void nuclea-
tion and growth into the simulation model [41]. A simpli-
fied approach was also used for modelling irregularity in
strut diameter to model the realistic manufacturing quality
of struts [168] and nodes [169] on mechanical properties of
lattice structures. Numerical studies of cellular structures
incorporating defects have also been reported [170–172].
Foams and stochastic porous materials have also been the
subject of IBSim studies in the past [173].
2.3 Impact ofDeviations fromIdealised Design
Geometry onProduct Design andPerformance
The simulation of full-scale components in the HVM sector
is not new. As already noted, real parts often deviate from
idealised design, adding uncertainty to the results obtained
from conventional simulations. Consequently, large safety
margins are often imposed upon in-service components.
Here, IBSim comes to the fore, incorporating various object-
specific deviations and unique aspects into the simulation,
for better prediction and characterisation. This increases
confidence in performance prediction (or reducing uncer-
tainty) and thus allows smaller safety margins to be imposed.
When considering issues relating to full-scale components,
this refers to deviations in actual part size, warping, surface
roughness, micro-cracking, or bonding interfaces. Especially
for complex shaped parts, the influence of such defects or
deviations might be unpredictable, and hence the need for
further quantification by IBSim.
At the macroscopic level a material or group of materi-
als might be chosen for a task based upon idealised mac-
roscopic properties. Important macroscopic metrics may
include overall volume and weight, with further consid-
erations of geometric tolerances, dimensions, and surface
areas. One can also look at the microscopic aspects of mate-
rials, particularly their structural arrangement, and subse-
quently attempt to design an idealised microstructure, with
key metrics such as the arrangement of pores and struts, or
the size, shape, and orientation of grains and fibres. For an
engineering product, as already discussed, these microscopic
metrics play an important role in determining macroscopic
behaviour.
Assessing the impact of deviations from idealised design
geometry is a multiscale problem, and the methods by
which geometries are digitally acquired (e.g., surface, or
volumetric scanning) and transferred to the computational
domain must, therefore, accurately capture deviations across
multiple length scales for IBSim to be representative. This
is a complex task as no single method covers all scales of
interest.
One conceptual approach is to use the materials clas-
sification shown in Fig.21 (reproduced from [174]). This
allows HVM to be viewed not only in terms of macroscopic
material type: non-porous solid or porous solid, but also in
terms of microscopic (microstructural) type. This simplifica-
tion enables a unifying link between the macroscopic and the
microscopic, making it easier to perceive common design
elements across the various materials used within different
HVM sectors. Since this is a geometry-based approach for
the assessment of deviations from idealised design, it links
well with IBSim.
Today, NDT of parts is widely used for evaluating poros-
ity, deviations from design, cracks, or other flaw types.
The pass/fail decision is made, however, based upon pre-
determined design rules depending on the material type, the
intended application, and the industry concerned. Although
this approach can provide a qualitative ‘rule of thumb’, it
does not take into consideration the full multi-physics com-
bination of the detected features, e.g., the combination of
a small pore that passes the design rule near an allowable
deviation from tolerance, which may lead to a combined
impact on performance that is greater than would otherwise
be allowable. Furthermore, the use of pass/fail testing poten-
tially leads to many parts being scrapped which could still
be serviceable if used under slightly different loading condi-
tions (e.g., in a different location within the assembly or in
an assembly not expected to undergo the same extremes).
Fig. 20 IBSim model of loading applied to a gyroid lattice structure
of titanium alloy produced by L-PBF. The small section viewed here
is cropped from the larger structure showing the location of high
stresses, and the rough surface exacerbates this [163]. Compressive
loading is applied in the vertical direction in the image

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
The use of part-specific simulation to provide a more quan-
titative evaluation is the next generation for this type of test-
ing. It is important to realise that despite the inclusion of
macroscale deviations (porosity, etc.) into the simulation
model, which is a significant improvement over the simula-
tion of idealised design geometries, some flaw types may
still go undetected. For instance, residual stress is not visible
to XCT, and some microcracks may go unnoticed below the
scan resolution. This section presents published applications
of IBSim in HVM according to the industrial sector to which
they’re most relevant.
2.3.1 Examples intheEnergy Sector
Electrochemical energy devices are complex multi-phase,
multiscale systems, consisting overall of solids (both porous
and non-porous), liquids, and gases. Whilst the gases are
not ‘designed’, the solid structures that contain them are.
Liquid electrolytes are designed, but since IBSim is not used
explicitly for this process, they will not be discussed further.
Interested readers are directed to further reading [175, 176].
Electrodes found in lithium-ion batteries are porous sol-
ids, typically consisting of electrochemically active mate-
rial and conductive additives, held together with a binder
(often polymeric) [177, 178]. The pore network within
these agglomerated structures contains mostly electrolyte,
but depending upon the battery chemistry, gases evolve at
the electrodes during operation that may also inhabit the
pore space. Electrodes can also swell during operation as
ions are intercalated. Thus, the battery is a pressure vessel
that requires careful thermomechanical management. From
a macroscopic perspective, then, deviations from idealised
designs of certain safety features could be catastrophic, and
IBSim has been performed alongside physical testing to vali-
date safety models and to update standards [179]. From a
microscopic perspective, no two electrodes are manufactured
identical, but their microstructural metrics can be made sim-
ilar with existing products conforming to a predefined range
to guarantee performance. This range is the culmination of
extensive previous research, and efforts continue to optimise
existing solutions and search for alternatives. Thus, when we
think of the deviations from the ideal, and how this impacts
products and their performance, we need to investigate how
IBSim is used for characterising microstructural metrics for
electrochemical energy devices.
A limiting factor of note within battery electrodes is mass
transport restriction, where constrictions within the three-
dimensional pore network can cause flow paths to be highly
tortuous [86]. In one study the 3D microstructure of a sin-
gle LiFePO4 electrode was acquired using XCT [110]. The
effects of tortuosity within the pore network were examined
using IBSim to assess the impact on ionic diffusion, a key
performance parameter, showing the specific role of each
microstructural phase (see Fig.22, reproduced from [110]).
In work by Trembacki etal., an XCT dataset from [180]
was used to simulate binder-phase morphology in nickel-
manganese-cobalt cathodes [110]. The amount of binder
can be varied within an electrode to alter electronic con-
duction. The calendering pressure can also affect porosity
and contact between particles within the active material.
These manufacturing parameters may be chosen to enhance
energy or power density. Thus, deviations from these will
impact intended performance. Interestingly, this study com-
pares the finite volume method (FVM) and FEA on the same
Fig. 21 Reproduced from [174] where heterogeneous solids are grouped and further subdivided by microstructure classes

Ll.M.Evans et al.
1 3
mesh, uncovering discrepancies when simulating electronic
conductions at material interfaces where singularities can
arise. This points to the possible use of adaptive meshes
to improve simulation fidelity. Further examples of IBSim
applied to research of batteries have also been published
[181–185].
An area of interest within the energy sector, for mac-
roscale simulation, is that of turbine blades. In a study of
composite turbine blades, containing imperfections, wires
and sensors, IBSim models from μCT data were performed
with and without the inclusion of these features provid-
ing information on the effective properties and influence
of these features on the performance [186]. At the micro-
scale, IBSim has been used to assess cast materials, such
as ductile cast iron (a non-porous crystalline solid), where
it is normally assumed that the crack initiation stage has a
negligible effect upon fatigue life, since early fatigue cracks
are often observed in these materials. However, gas bub-
bles can be trapped inside casts, and shrinkage can lead
to the formation of cavities, both of which serve to act as
fatigue crack initiation sites. These defects are precursors
to pre-existing cracks, acting like localised porous materi-
als within the global non-porous solid. Simulating this pore
space, its tortuosity and evolution, is therefore important
for predicting fatigue life scatter. In one study a compari-
son between experimental fractography data and simulated
fatigue life scatter was made [187]. X-ray μCT was used to
obtain defect distributions within a range of specimens taken
from rejected wind turbine hub castings, allowing a random
defect analysis to be performed to predict fatigue life scatter.
The nuclear industry, another highly regulated sector,
has stringent requirements on the quality and qualification
of manufactured parts. In this context, IBSim models have
been demonstrated to be useful to predict the performance
of AM parts for nuclear applications [188]. IBSim was used
to characterise a component manufactured with a bonding
procedure for dissimilar materials used in water-cooled heat
exchange components, identifying a defective joining pro-
cess within ‘digital twins’ that would otherwise have com-
prised the component and surrounding substructure [189].
IBSim was used with a digital twin approach for detecting
in-situ flaw formation in stainless steel (316L) impeller-
shaped parts manufactured by L-PBF [190]. The digital
twin approach was shown to be effective for detection of the
three types of flaw formation causes studied in this research.
Whilst in work by Evans etal. it was employed for high-heat
flux components used within experimental nuclear fusion
plants, where the debonding regions within carbon fibre
composite-copper interfaces can be detected and quantified
in silico from image data captured via high-resolution XCT
[16]. At the microscale it has been used to assess the effect
of microstructure and crystalline structure upon the thermal
conductivity of graphite foams [94].
To round-off the breadth of applications of IBSim for the
energy sector, we consider its application to semi-crystalline
polymers. IBSim has been used to improve the performance
of insulative porous polymeric coatings for offshore pipe-
lines [191]. A common problem when simulating such mul-
tiscale systems is the computational complexity involved;
molecular weight and distribution, size of crystallites, and
microstructure all affect mechanical behaviour under vari-
ous loading scenarios, and temporal effects such as work
hardening can present themselves. The homogenisation of
multiscale systems is an obvious approach to reducing simu-
lation complexity, but continuum-level materials properties
are not always representative at lower length scales. This
is an outstanding challenge with the IBSim approach and
common across application spaces.
2.3.2 Examples intheAerospace Sector
A primary functional requirement for engineering prod-
ucts used in aerospace is for them to be lightweight. Thus,
composite materials such as carbon fibre reinforced plastics
Fig. 22 Showing the specific role of each microstructural phase (reproduced from [110])

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
(CFRP) have found widespread use. At the microscale, the
fibre network of which they are composed is a key element in
their performance, with factors such as directionality affect-
ing damage evolution [192] and permeability [193]. X-ray
μCT is invaluable for characterising these networks, but one
particular challenge is to have a robust method to quantify
microstructural features during postprocessing of images to
obtain topologically accurate volumetric representations for
IBSim models [194]. For example, loss of fibre edge defi-
nition can occur when imaging fibres made of low-atomic
numbers, where X-ray phase contrast effects can cause them
to appear thinner than their true diameter. Another area of
significant industrial interest at this scale is the quality of
welded regions within parts, with the potential for large
pore spaces in these weld seams to create possible failure
locations. An example of a welded light aircraft engine
bracket shows the presence of a pore with diameter more
than 1mm in size [195]. Incorporating this into a simula-
tion with ‘in-service’ loading conditions would identify the
criticality of this pore for the intended application. In a study
of welded seams using XCT and tensile tests, the authors
compared simulations of the weld with tensile tests of the
same samples, incorporating pores and surface irregularities
into the simulation [14]. The stress–strain curves obtained
from numerical simulations were in good agreement with
experimental ones within the linear elastic regime; however,
they deviated from each other when non-linear behaviour
was observed experimentally since plastic behaviour was
not implemented in the numerical models. The amount of
pore content (high or low) and type of surface appearance
(irregular or regular surface) was demonstrated to change
the quantity of plastically-deformed areas and load-bearing
capacity.
An example of IBSim applied to the macroscale involved
a study to evaluate the performance of parts which contained
intentionally controlled defects [196]. The authors applied
a simplified FEA method with a linear elastic assumption to
μCT scans of AM aerospace brackets with varying sizes and
locations of defects. The incorporation of these defects, in
addition to the surface imperfections, into the simulations
led to accurate predictions of failure locations using a stress-
hotspot evaluation approach. The utility of XCT continues,
as it also allows for the comparison of actual geometry to
designed geometry, giving insight into deviations. The color-
map, shown in Fig.23, clearly identifies the largest devia-
tions, which is especially important in critical locations of
components, such as load-bearing sections with thin walls.
The geometry shown is obtained by simulation-based design
(also called topology optimisation or generative design) and
the component is manufactured by metal AM. The arms are
warped towards one another, affecting the alignment of the
two holes, which is critical for its practical application. The
decision, based on this analysis, was that this component
required additional machining, and the ‘as manufactured’
bent arms might induce higher stresses when subjected to
planned loads, which would not have been anticipated dur-
ing design.
The ability to optimise the design of complex parts for
specific loading regimes or functionality is particularly use-
ful for targets such as reducing mass. This approach usually
requires multiple rounds of simulation during the design
process. Besides the use of simulation in the design process,
the application of simulations to the final design for a quality
control step is also important for highlighting possible limi-
tations of the design and to check minimum safety factors
[162, 197]. Given that the aerospace industry is yet another
tightly regulated sector, it is key to the acceptance of novel
techniques like AM that uncertainty in performance is lim-
ited. The advantages of AM for engineering structures with
increasingly complex geometries are clear. For example, a
range of designs for shape-changing thin-walled cylindri-
cal composite structures were subjected to non-linear static
FEA simulations using quadratic hexahedral elements [198].
FEA simulations were combined with experimental test
results using digital image correlation (DIC), which allowed
strain maps to be correlated with physical structures that
had been fabricated by AM. Being able to predict failure in
AM engineering products is also vital. Immersed-boundary
finite elements can be used to predict the tensile strength
of designs, simulating stress distributions from local stress
concentrations and the location of crack initiation sites. This
method does not require a conforming simulation mesh
and is therefore suitable for complex porous solids where
meshing may introduce singularities. This approach was
taken by Fieres etal. using aeronautic parts consisting of
AM AlSi10Mg aluminium alloy, with physical specimens
destructively tested in tension [149]. They found there was
good agreement between IBSim models and experiments for
Fig. 23 CAD variance analysis of an actual AM bracket (with colour
coding showing deviation) compared to its CAD design (shown in
yellow mesh). This example is from round robin tests [196] whereby
AM parts were analysed by fixed XCT workflows, one of which is
to evaluate differences between actual geometry and design geometry

Ll.M.Evans et al.
1 3
predicted and measured tensile strengths. Furthermore, they
concluded that crack initiation location and onset could be
forecast accurately.
Further examples of the application of IBSim within this
sector include design improvements for composite Hart-II
blades [186], biomimetic insect-sized micro wings [199],
and nanoscale hexagonal plated wings for next-generation
microflyers [200]. IBSim models of these structures require
complex simulations coupling high-speed fluid flow,
fluid–structure interaction, thermal flow, and mechanical
vibration analysis. There exists a range of readily available
software packages to explore different simulation options,
but it will be important for anyone attempting to perform
high-fidelity simulations on these IBSim models to balance
meshing intricacy, component geometry complexity, and run
time analysis.
2.3.3 Examples intheMedical Sector
Biomedical IBSim applications in general are based on
patient medical imaging (e.g., CT/MRI) and are outside the
scope of this review, and will, therefore, only be referred to
for the purposes of context; the primary focus being manu-
factured parts used in the medical sector. IBSim has great
potential for reducing material waste due to part rejection.
A review was made of CT-based measurement techniques
used for assessment of quality of bioengineering compo-
nents [201]. Due to the critical nature of medical applica-
tions, FEA is often used [202, 203], especially in the design
phase for new device development [204]. Originally FEA
was widely used in orthopaedic studies for improved under-
standing of bone stress distributions, including bone-pros-
thesis structures and similar devices [205]. FEA has found
particular use in dentistry and orthodontics [206–208] and
spine research has also benefitted from FEA for improved
understanding of the spine and spinal implants [209], whilst
simulations involving the human skull have supported the
understanding of head injuries even assisting forensic inves-
tigations [210, 211].
Due to the wide scope of applications of FEA in biomedi-
cal and biomechanical applications, some guidelines have
been suggested almost a decade ago [212–217]. The medi-
cal sector is understandably highly regulated, and as such,
use of novel methodology usually requires certification from
a regulatory body (depending on the nation) before use is
permitted with patients. Despite this, and due to the com-
plexity in geometries and variability between cases in bio-
medical applications, IBSim has already found widespread
use, including in the study of scaffolds for bone and tissue
regeneration [218–222], implants [223–227], intracorpor-
eal structures [228–230], biomechanics [231], and medical
device design verification [232].
In a patient-based IBSim study of a medical implant, high
stress locations were identified that led to eventual failure
of the implant in the patient [12]. In this work, a medical
CT scan of a patient was used to design a suitable mandibu-
lar implant geometry (patient-specific and porous) which
unfortunately later fractured in the patient. An IBSim study
was therefore conducted to correlate the actual failure loca-
tion (from subsequent CT scans) with the location of high
stresses by using IBSim models of the designed implant.
This study highlights the potential of macroscale geometri-
cal simulation of complex objects, informing the design,
irrespective of manufacturing defects and imperfections. In
this case, the manufacturing process was not checked, rather
the design was flawed, though manufacturing flaws or imper-
fections can contribute to such failures. Similarly, fracture
behaviour of human and sheep mandibular diastema fixated
with titanium miniplates and screws under physiological
muscular loads and variety of clenching modes (intercuspal,
incisal, and unilateral) were simulated using IBSim mod-
els [233]. IBSim models of mandibular constructs made of
Titanium scaffolds were simulated by using 3D FEA and
multiscale modelling to pre-clinically determine the opti-
mal mandibular geometry for a specific patient, where the
influence of strut diameter and inter-strut distance in porous
architecture on stress and strain distributions was quantified
[234].
A particular challenge with IBSim, in comparison to
conventional engineering simulations, is the computational
expense involved. Reducing complexity of IBSim models
is one solution and can be required if access to high-perfor-
mance computing resources is a limitation. As with mate-
rial characterisation, a popular method when considering
microstructure in biomedical engineering, is to use RVEs
with periodic boundary conditions to reduce computational
expense. For example, one work used RVEs to homogenise
the mechanical properties of Ti6Al4V structures for bio-
medical applications [223]. The IBSim models were subse-
quently solved using a four-node tetrahedral mesh via com-
mercially available software. To validate their approach,
models were manufactured by AM with electron beam
melting (EBM), then subjected to mechanical tests. Whilst
reported results appear limited, it does show that homog-
enisation of key design elements through RVEs can be a
useful method for design improvements with reduced com-
putational complexity. Another example of reducing compu-
tational complexity includes methods such as the extended
finite element method (XFEM) [224], where the initiation
and propagation of fracture paths in specimen-specific bone
can be predicted. The benefit of XFEM is that it does not
require a priori information about the crack path, and thus
model re-meshing is not needed. For those interested, a
review of FEA models and their validation for tibiofemoral
joints is available [231].

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
2.4 Customisation andPersonalisation ofProducts
Nowadays, customers expect products tailored to their per-
sonal preferences, tastes, needs and lifestyle. The paradigm
of mass production has therefore shifted in favour of cus-
tomised production, which caters to the needs of individual
clients or patients [235, 236]. The notion of personalisation
is not unequivocally defined, although the common features
of this approach indicated in the literature are customer
preferences, customer participation in the product design
process, customisation and information flow between cus-
tomer and manufacturer [237, 238]. The level of customer
involvement in the production cycle is thought to play a criti-
cal role in determining the degree of customisation, whereby
the earlier the initial involvement of the client, the greater
the customisation.
The term ‘mass customisation’ refers to products which
are mass produced but where the consumer is offered options
to customise the product. Mintzberg views customisation
as taking one of three forms: pure, tailored, or standard-
ised [236]. Pure customisation includes the consumer in
the entire cycle, from design through fabrication, assembly
and delivery and it provides a highly customised product
(Fig.24). Altering a basic design to meet the specific needs
of a particular client is known as tailored customisation, and
in standardised customisation a final product is assembled
from a predetermined set of standard components [239].
To avoid confusion, and to ensure a clear distinc-
tion between customisation and personalisation, the term
‘pure customisation’ will no longer be used throughout
this section. Instead, the term ‘personalisation’ will refer
to bespoke/individualised products which are fulfilled at
the personal level (i.e., a market of one). The automotive
industry is a well-documented example of low and high
degree of customisation [240]. Here, the customer can opt
to customise the car from a plethora of options (e.g., colour,
model, trim level etc.). The Porsche automotive company
have recently ventured into custom car seating to allow the
customer to choose between three firmness levels (hard,
medium, soft). The 3D printed bucket seat is set to become
a personalised product in the future, based on the customer’s
specific body measurements.
To move from mass production to mass customisation,
a company needs to invest in the right technological capa-
bilities. Traditional manufacturing methods are restricted in
their ability to create customised products as new moulds
are typically required for each product. This subsequently
leads to increased change-over costs associated with tooling
and fixtures and extended timelines. AM has been widely
acknowledged as the most appropriate manufacturing
method of production of customised products, due to the
lack of associated tooling required and the ability to pro-
duce highly complex geometries. AM covers a broad range
of production technologies that fabricate products layer-by-
layer, enabling 3D objects to be ‘printed’ on demand in a
variety of materials. AM technologies such as stereolithog-
raphy (SLA), 3D printing, L-PBF, selective laser sintering
(SLS) and EBM lend themselves to manufacturing complex
anatomic parts without any barriers of design constraints.
Examples of customised and personalised products include
implants [241], bone and tissue scaffolds [242] and prosthet-
ics [243] within the medical sector, which appears to be the
one currently most active in this field. Then closely related
to the medical sector are customised personal protective
equipment [244] and protective sportswear [245].
Fig. 24 Customer involvement
and modularity in the produc-
tion cycle of mass customisa-
tion [239]

Ll.M.Evans et al.
1 3
The most accurate analysis of processes and mate-
rial behaviours comes from in-situ imaging and diffrac-
tion techniques. Combining imaging technologies such as
XCT, FIB-SEM and MRI with the design freedom of AM
has opened up new and exciting opportunities to customise
and personalise products to many applications [246–250].
This has been particularly advantageous for industry such
as healthcare to improve the effectiveness of diagnosis, plan-
ning, surgery, and clinical outcomes perfectly adapted to the
patient’s specific anatomy or needs. However, a critical stage
of this process is simulation modelling, e.g., using FEA.
The ability to analyse and test how a product will react to
certain environments (e.g., heat, force, microclimate etc.)
and to predict structural strength and prevent failures is cru-
cial where safety is paramount, e.g., both in patient-specific
devices or custom wire baskets in aerospace.
Understanding the environment is important to all appli-
cations, but it is particularly complex in the case of medical
applications where devices are intended to be used inside
the human body. As organs and soft tissues already exist
in the patient, it is imperative that the bone and tissue scaf-
folds integrate fully and do not cause infection or become
‘rejected’ by the body. Customised or patient-specific scaf-
fold geometry can be gained by applying CAD software
along with known individual patient anatomy parameters
related to the defect site to create a 3D model. Computer
modelling and FEA before 3D printing of a composite bone
or tissue scaffold allows accurate identification of patient-
specific anatomy and any variation in defect shape and size;
subsequently ensuring the quality of the final medical model
and product is not impaired [251].
The continuous demand for efficient and adaptive cus-
tomised and personalised products relies heavily on IBSim
techniques. In this sub-section we have therefore presented
a variety of applications which have implemented IBSim in
the development of customised and personalised products
below. These include: Healthcare, Personalised Medicine,
Prosthetics and Orthotics, Sport and Lifestyle, and Automo-
tive sectors. Note, that this is not intended to be an exhaus-
tive list, but instead an overview of the key applications
observed as currently being of interest.
2.4.1 Healthcare
Arguably, healthcare is one of the most complex and chal-
lenging industries to produce customised and personalised
products. Due to the differences that exist between humans,
it is essential that all aspects of design and manufacturing
are considered to ensure the device not only functions as
required, but that it does not cause damage and/or harm to
the body. The complex anatomy, sensitivity of the surround-
ing bones and soft tissues, and uniqueness of the defect or
malfunction means IBSim can play a critical role.
Obtaining patient- or object- specific surface or volumet-
ric geometry is important to assess the size, location, and
overall fit of the device/product in relation to the individual.
Static or dynamic simulations can then be run to investigate
short- or long- term outcomes. The ability to run simulations
that mimic the likely response of the device against patient-
specific geometry is crucial for predicting the success of the
proposed product and determining any potential flaws in the
design which could be optimized. Simulations are also use-
ful in understanding the behaviour of a material, particularly
in time dependency models where wear and tear can occur
over a given time period.
Applications of XCT with AM in a medical context are
extensive, particularly in dentistry, where there is great
demand for personalised products [207, 252–255]. Primary
applications of XCT with AM in medicine include: the pro-
duction of anatomical models, surgical guides, endoprosthet-
ics and orthotics, stand-alone implants and scaffold implants
[256]. These applications rely on the principles of reverse
engineering, using patient XCT data to inform the design
process. These advantages make AM invaluable in tissue
engineering applications, where the production of micro-
scale lattice structures is an intrinsic requirement.
It is also very common to see other imaging modalities
such as MRI used in medicine as access to pre-operative data
is readily available. In the same way as XCT, MRI data can
be used to reconstruct patient-specific 3D geometry. Whilst
the main focus of this review is XCT due to HVM applica-
tions, it is worth acknowledging that MRI has several ben-
efits over XCT whereby it does not exposure the patient to
ionizing radiation. It is also extremely useful in examining
soft biological tissues, whereas XCT is particularly useful
for examining materials with a high atomic number [27].
Acquisition methods such as cone beam computed tomogra-
phy (CBCT) have considerably reduced the radiation dosage
to the patient, but there are increased concerns regarding the
collective radiation dose given for medical purposes [257].
XCT and CBCT have proven particularly useful for complex
dental cases involving surgical planning, detection and treat-
ment of tumours and reconstructive surgery of the mandi-
ble; where personalised geometry is acquired for point cloud
data processing and analysis using FEA [247, 258]. Publicly
available tools such as the MATLAB ‘Torsion Tool’ and the
‘Bone deformation Tool’ allow personalised geometry to be
generated quickly in an OpenSim musculoskeletal model
[259]. Such tools can estimate personalised measurements
within seconds but are often limited to a single model.
Yu etal., integrated CBCT, reverse engineering, CAD,
FEA and rapid prototyping to fabricate an accurate custom-
ised surgical template for orthodontic mini screws [260].
IBSim was of particular importance to this study as the
use of CBCT was able to measure the interradicular spaces
with greater accuracy and reproducibility than other imaging

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
modalities (such as multidetector CT), whilst the FE models
allowed biomechanical evaluation of the customised surgical
template with increased clinical stability. Whilst full IBSim
was utilised in this study, further simulations could have
been implemented to assess the biomechanics of the surgi-
cal template post-surgery in order to reflect the effect of
soft tissue inflammation and screw loosening. The authors
did acknowledge, however, that there were large gap sizes
between the surgical template and teeth/mucosa due to fitting
errors with the soft tissue image reconstruction [260]. This
could also potentially be rectified by further FE analysis.
Another example is presented in a recent paper by Dot
etal., whereby CBCT, intra-oral scans and subject-specific
FEA were used to track the 3D orthodontic tooth movement
in a patient undergoing canine retraction over a seven-month
period [261]. An iterative closest point (ICP) algorithm was
used in Geomagic Studio software to align and register the
scans and segmented canines at different stages (i.e., initial,
and intermediate). Open-source software ITK-SNAP, 3D
Slicer and Mimics were then used to create 3D models and
calculate rigid body displacements of the canines. The 3D
models enabled preliminary FE models to be developed and
validated [261].
There is a breadth of literature that focuses on the design
of implants, particularly for craniomaxillofacial surgery
where gunshot wounds or tumours have required mandibu-
lar or cranial reconstruction. Parthasarathy reviews a number
of articles related to 3D modelling and custom/personal-
ised implants in craniofacial surgery, specifically in relation
to application of CAD and computer-aided manufacturing
(CAM) technologies with different materials [241]. The con-
version of CAD models to STL format for manufacturing is
common. The use of CAD/CAM in dentistry has allowed
various morphologies from different devices with high accu-
racy, thereby increasing treatment opportunities in some
clinical situations. This technology combined with a L-PBF
machine can provide porous titanium structures with com-
plex geometries that control the internal architecture [262].
In addition to L-PBF, SLS and EBM have also been used to
facilitate the direct production of titanium, chrome cobalt
and polyetheretherketone (PEEK) implants with engineered
properties that match properties of the tissues at the region
of implantation [250, 263, 264].
Wu etal., used a combined methodologic approach to
assess the biomechanical performance of a conventional and
custom angled dental abutment without the need for wax and
cast [265]. Numerical models were acquired through optical
scanning of the dental cast, with the optical gaging prod-
ucts video measuring system used to obtain detailed shape
parameters. A virtual prosthesis was then preliminarily
positioned and probed for interference using the “Collision
Detection” function based on the “Least Square method” to
fix the prosthesis in the required position. Geometry of the
patient’s bone was taken via medical CT images and an FE
model generated in ANSYS Workbench 11.0. The results of
the von-mises stress distribution simulation can be seen in
Fig.25 where no distinct difference in the stress distribu-
tion was found using the custom or the conventional angled
abutment [265].
XCT-based FEA has evolved into a standard tool for
the biomechanical evaluation and optimisation of porous
bone tissue scaffolds. Systems such as the Skyscan 1272,
Bruker-MicroCT are commonly used to obtain high resolu-
tion 3D scan data of the scaffold. The reconstructed images
and morphometric and structural analysis can then be per-
formed using Bruker proprietary software NRecon® and
CT-Analyser CTAn® [266]. The primary advantage of XCT
for bone tissue scaffolds is that it is a non-destructive imag-
ing technique which is capable of providing a comprehen-
sive set of data. However, the accuracy of the analysis, is
highly dependent upon several parameters such as specimen
preparation, parameter settings during the acquisition, and
reconstruction of the images [267].
The majority of literature surveyed in this paper utilised a
combination of imaging techniques e.g., XCT and scanning
electron microscopy (SEM) in order to examine the char-
acteristics of the scaffold. 3D CAD designs and theoretical
equations/simulations were also common for designing the
porosity of the scaffold but there have been reports that cer-
tain CAD simulations over-predict scaffold performance due
to limitations in simulating micro-topologies [268]. FEM
proved useful in investigating and optimising the mechani-
cal behaviour of the scaffolds [269, 270]. The use of FEM
was also able to measure the sensitivity of scaffold proper-
ties (e.g., to the filament diameter, the variations of porosity
and surface area). The review paper by Podshivalov etal.
describes the state of the art in multiscale computational
methods used in analysing bone tissue for personalised med-
icine is summarised by [271]. Challenges on optimization
of 3D-printed customised bone scaffolds is presented in a
recent review paper by Bahraminasab [251].
2.4.2 Personalised Medicine
Personalised medicine (often referred to as “precision medi-
cine”) is an emerging field which will significantly benefit
from the implementation of IBSim. In the US, the Food and
Drug Administration reported 38% to 75% of patients for
whom medication was ineffective for a number of conditions
from depression to cancer [272]. Therefore, the ability to tai-
lor a drug dosage specific to an individual where ingredients
can be adjusted based on the patient’s age, gender, weight,
genetic factors and previous responses to different dosage
levels, rather than using the conventional dosage forms may
diminish all potential adverse effects [273–275].

Ll.M.Evans et al.
1 3
For example, IBSim techniques are invaluable for radio-
therapy treatment planning as they can be used to; predict
tumour response to radiation, reduce uncertainty in the pre-
scribed dose distribution and spare organs at risk [276–280].
However, accurately calculating the perturbation effects of
the interfaces between materials of vastly differing ana-
tomic number (e.g., lung, bone and/or air) is complex, and
as such, has often been solved using the Monte Carlo method
[281]. A recent study by Roncali etal. looked at personal-
ised dosimetry for liver cancer Y-90 radioembolization for a
single patient [282]. CBCT was used to segment the hepatic
arterial tree to predict microsphere transport using multi-
scale CFD modelling and Monte Carlo simulation. Bespoke
manufacturing is also used within personalised radiotherapy
treatment, mainly shielding blocks to protect regions of the
body not intended to receive a dose [283]; and boluses to
alter dosing received from the beam [284]. There are exam-
ples where the performance of custom boluses have been
modelled with IBSim [285](Fig.26). Firstly, the head geom-
etry of a phantom was captured via CT scanning, the bolus
was designed to fit the topography of the ear and manufac-
tured with AM. Once placed on the phantom, the head was
re-scanned, and the data used in a simulation of a CT scan
to evaluate the dosimetric properties of the custom bolus. It
was found that the custom boluses better fitted the irregular
surfaces, and this enhanced the dose that would have been
received by the patient.
IBSim has not only played a pivotal role in the develop-
ment of personalised medicine but also in the storage sta-
bility of pharmaceutical products. One example of this is a
study by Zhang etal., which characterized a lyophilized drug
product (a freeze-drying process that removes water from
a drug product via sublimation) [220]. The study involved
the use of high-resolution X-Ray microscopy to collect 3D
volume data from lyophilized drug samples and quantita-
tively characterise the microstructures of the lyophilized
drug [286].
Examples of digital twin applications and the ethical
issues that arise when digital twins are applied to model
humans for personalised medicine are presented in a recent
review paper by Kamel Boulos and Zhang [287]. Of note, is
the study by Cho etal., who assessed the facial profiles of
Korean adult females using facial scans and CBCT imag-
ing (Fig.27). Digital twins were reconstructed to evaluate
and compare the sagittal relationship between the maxillary
central incisors and the forehead before and after orthodontic
treatment [288]. This technology is somewhat in its infancy
with regard to personalised medicine, but it is interesting
to see this implemented with full IBSim. The concept of
the digital twin for mass customisation enables new options
for product manufacturers as seen in the aerospace industry
Fig. 25 Von-Mises stress distribution of implant, custom abutment, conventional abutment and screw under loading along the abutment long
axis (row 1) and along the implant long axis (row 2) [265]

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
[289]. Over the next decade, we are likely to see more
research involving digital twins to personalise medicine, but
this will require solving a wide range of technical, medical,
ethical, and theoretical challenges.
2.4.3 Prosthetics andOrthotics
Researchers have been interested in computer-aided pros-
thetic socket design since the early 1960s, however, FEA
was not introduced to prosthetic socket and orthosis design
until the late 1980s [290]. Prefabricated prosthetic and
orthotic products are readily available and less expensive
than custom products; however, customised, or personalised
products that take individual characteristics into considera-
tion are considerably more comfortable and functional for
the wearer.
Examples of mass customisation of orthoses to date have
largely consisted of foot and/or ankle–foot orthoses (AFOs)
[291]. All studies from 1990 to 2015/16 have been summa-
rised in the review papers by Jin etal. [292] and Chen etal.
[293]. Each of which have discussed the progress of AM
of custom prosthetics and orthotics and the benefits over
traditional plaster moulding techniques. More recently, Mali
and Vasistha presented an efficient solution for the manu-
facture of an AFO using reverse engineering software to
obtain a refined model via data repair [294]. A Steinbichler
Comet3D™ structured blue light scanner was used to obtain
the geometry of the diseased foot and the generation and
post processing of the cloud data points were conducted
using the proprietary software Cometplus™ and Autodesk®
Meshmixer. FEA of the AFO was performed in Autodesk®
Fusion360™ for three different materials. Optimization
of the orthosis resulted in increase in safety factor, higher
strength and lesser displacement when compared to a non-
optimized AFO [294]. Agudelo-Ardila etal., proposed a
similar solution for an upper limb orthosis whereby a sub-
ject’s hand and forearm were scanned using a structured
light 3D scanning system [295].
The STL model was processed in Canfit and Meshmixer
software. FEA was utilised to perform structural simula-
tions to determine when the material will deform or collapse
(Fig.28). Biomimetics were implemented from Voronoi
structure (as an alternative for modelling cellular structures)
and met the objective of material reduction, consequently
leading to a lighter orthosis (Fig.28). Both these studies
demonstrated the effectiveness of IBSim not only as a useful
tool for evaluating structural feasibility but also in making
design decisions, reducing problems associated with plaster
moulds and thus achieving a custom-made orthotic that is
optimised for the patient [294, 295].
In addition to AFOs, there has been the development of
custom prosthesis for the management of entero-atmospheric
fistulas whereby an Einscan pro + , 3D shining scanner cou-
pled with CAD software was used to capture the geometry
of the fistula and create a polycaprolactone personalised
ring-shaped device [243]. The device was then placed on
the image of the wound to verify the customisation and
Fig. 26 Isodose lines calculated by means of IBSim CT simulations of a phantom head a with no bolus, b with a commercial bolus and c a
bespoke AM bolus [285]
Fig. 27 A digital twin reconstructed by the fusion of facial scan and
CBCT images. a Coronal view of the face. b Sagittal view of the face
[288]

Ll.M.Evans et al.
1 3
placement (Fig.29C–E). Virtual simulations allowed the
tolerance margin to be calculated which was important in
ensuring the prosthesis did not press on the fistulous tissue.
Ideally, any product that is designed for the treatment of
an individual patient with specific illness, disease or injury
needs to be fully personalised. The emergence of AM tech-
nologies allows the fabrication of custom-made orthoses in
a cost-effective way [296]. Interestingly, in a 2020 review on
the use of AM to produce lower limb orthoses, only three
of the seven AM technologies available have been explored
(vat photo-polymerisation, material extrusion and powder
bed fusion). Material extrusion was found to be the most
affordable AM technology but limited to the use of polymers
[296]. SLA was considered an unsuitable method for manu-
facturing AFOs due to limited bending and fatigue strength.
Instead, FDM was selected for customising AFOs. One of
the interesting points mentioned in the review paper was
the ability of Cyber Design and AM (CDAM) to assist the
design phase of orthoses [296]. Fig.30 shows an overview
of the CDAM system developed by Shih etal., which aims to
improve the fit and comfort of custom orthoses and prosthe-
ses, and enable users to solve complex design and analysis
problems (e.g., FEA, optimization, visualisation) [297].
Sharma etal., recently proposed a methodology for
designing highly customised 3D printed facial protection
orthosis for rehabilitative management in patients with
sports-related maxillofacial injuries [298]. A postoperative
CBCT scan was imported into Materialise Interactive Medi-
cal Image Control System (MIMICS) medical software in
order to segment the region of interest and generate a 3D
volumetric reconstruction (Fig.31). A 3D optical scan was
Fig. 28 FEA analysis showing the stress and temperature analysis of the lower and upper part of the orthosis (top and middle row respectively),
and the application of Voronoi patterns to reduce the material in the resulting 3D printed custom orthosis (bottom row) [295]

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
also taken to corroborate the soft tissue components and
create digitised surface geometry of the face with a trian-
gular mesh. An ICP algorithm, surface registration protocol
(n-point and global registration) was accomplished between
the CBCT scan and optical face model [298]. Conventional
methods for fabricating a mask were eliminated using
IBSim, thus eliminating a very time-consuming process. The
imaging and AM workflow demonstrated in this study could
be applied to numerous applications, but computational
simulations are not presented. The maxillofacial orthotic
was based on an adult male whose face shape is unlikely
to change (as the bones are fully fused). For a child or ado-
lescent who has not reached skeletal maturity, simulations
could help predict the optimum design of the orthotic and
also identify when the orthotic is no longer effective. Digital
workflows similar to the abovementioned study have been
found for cervical collars [299].
2.4.4 Sport andLifestyle
With the growing concern surrounding head injuries in
sport, there has been increased research into personal protec-
tive equipment [245, 300]. 3D surface scans, medical imag-
ing and/or 3D anthropometric data, can be used to acquire
the head geometry and proposed protective headwear. The
detail of this geometry depends on the type of methodology
implored and the material properties assigned to the head
and proposed helmet design. Corrales etal., addressed these
limitations by developing a numerical model of a modern
football helmet by integrating two headforms and assessing
a range of impact conditions [301].
Virtual impact test simulations can assess which design is
most effective at protecting the head (in terms of structural
and kinematic response to impact) and can be evaluated by
experimental impact testing. Fig.32 shows an example of
a custom-fit bicycle helmet model proposed by Ellena etal.
whereby 3D anthropometry, reverse engineering techniques
Fig. 29 A Process of taking pictures with the bioscanner. B Images
obtained with the bioscanner. C Measurement of the exposed intes-
tinal surface dimensions for device design. D Verification of the
suitability of the prosthesis by extrusion of the fistulous surface. E
Placement of the device on the image of the bioscanned wound to
determine the correct adaptation to the patient. F 3D printing of the
bioprosthesis [244]
Fig. 30 a Overview of the Cyber Design and Additive Manufacturing (CDAM) system for custom Ankle Foot Orthoses (AFO). b Illustration of
the interaction between the hardware and software systems with the cloud storage system [297]

Ll.M.Evans et al.
1 3
and computational analysis methods were used to assess
accuracy of fit [302]. This study demonstrated that the fit
accuracy of the custom-helmet models was significantly
increased compared to three commercially available helmets,
and their method complied with the relevant drop impact
test standards. However, the authors acknowledge that the
mechanical properties of the available materials used in their
custom design differ significantly from well-known foam
material, and that a combination of AM with moulding tech-
niques would be a likely outcome in future studies.
Safety standards and certification have most likely con-
tributed to the lack of mass customisation systems of hel-
mets to date. Industry customised helmets, by brands such
as Bell Sports® (Rantoul, Illinois, USA) currently meet the
US Standard, but the information in how the Standard was
achieved is not disclosed. In addition to helmet designs,
there has been a growing trend among shoe manufactur-
ers (e.g., Nike, Adidas) to introduce customised shoes to
improve fit, comfort and performance [303]. Custom-fit
mouthguard designs, have also been studied extensively to
Fig. 31 Top: An overview of the schematic representation of the
digital workflow with Postoperative CBCT 3D volumetric reconstruc-
tions. Bottom: A 3D computer-aided design and planning B FDM
printed carbon-reinforced PLA face mask C A professional soccer
player with a customized face mask during his sport’s practice session
[298]
Fig. 32 Impact locations of the customized helmet: side, front and top (left) and deviation analysis of a participant’s customised helmet (right)
[302]

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
evaluate the tooth stresses and strains, shock absorption, and
displacement during impact [304, 305]
2.4.5 Automotive
The automotive industry is a highly competitive market
where there is a constant and ever-increasing demand for
personalised products. The CES 2020 Survey by CITE
Research Dassault Systèmes found that 83% of 3000 con-
sumers in the US, China and France expect products or ser-
vices to adapt in a matter of moments or hours [306]. Only
21% will wait four or more days for a personalised product
or service to be delivered, but they are willing to pay more
(an average 25.3%) for personalisation [306].
Consumers’ increasing demand for personalisation
capabilities, coupled with their refusal to incur any extra
wait time for delivery, sets up a major design challenge for
engineers. Some luxury carmakers have embraced this chal-
lenge and launched customisation schemes to deliver tai-
lored designs to meet customer requirements. MINI Yours in
2018 enabled customisation of small parts such as steering
wheels, decals, and colour combinations [307].
FEA of vehicle crash analysis and crash test dummy
simulation uses software such as LS Dyna to run explicit
analysis. To predict and assess the response of occupants in
a vehicle crash, specific models can be generated to include
a number of individual characteristics (e.g., gender, age,
height, weight, etc.). Automatic seat belt draping, pre-ten-
sioning and body interaction with the seat will influence the
sophistication of these simulations. The M50 seated finite
element male is intended for use in simulations of vehicle
crash and was developed in LS-DYNA Rev. 4.2.1 [308].
Multi-modality imaging comprising MRI, CT, and a 3D dig-
itiser (FARO Technologies Inc., Lake Mary, FL) was used
to capture subject-specific anatomy (Fig.33). Non-linear
dynamic FEA simulations have also been used to predict the
magnitude of impact forces, G loading, deformation, stresses
as a function of race car velocity and the angle of impact of
a novel airbag technology [309]. Custom clutch designs to
determine the suitability of a specific material to be used in
real production have also been analysed [310].
2.5 Image‑Based Simulations inBiomimicry
When the imaging techniques already discussed are used
to image biomaterials, the detailed multiscale images pro-
duced provide invaluable information, which may be used
to better understand nature and consequently used to solve
engineering problems. Biomaterials are the result of hun-
dreds of millions of years of evolution in nature, a natural
iterative optimisation method. They are known to be multi-
functional, for example, having the functions of impact or
fracture resistance [311, 312], armour and protection [313,
314], strength and durability [315], light weight for flight
[316]. Meyers etal. [317] stated that there are two levels for
the development and implementation of concepts extracted
from nature (i) design and concepts which are inspired by
nature but applied with different materials and conventional
processing techniques such as self-cleaning surfaces inspired
from lotus leaf for hydrophobicity [318], (ii) bioinspired
structures which are mimicked in molecular level by means
of self-assembly and molecular engineering such as high-
performance ceramic–metal composites designed from ice-
templating process [319]. Plessis and Broeckhoven reviewed
the ability of AM in biomimicry and the contribution of
biomimetic structures to new engineered products and appli-
cations [320]. In this section we present papers which have
used IBSim as part of their biomimicry research collated
into three main approaches.
Previous to this review, a few other invaluable works
have concentrated mainly on a specific part of the IBSim
technique. For example, Cooper etal. reviewed IBSim FEA
models of tibiofemoral joint with respect to specific titles
such as generic knee models, mechanical models [331]. Ha
Fig. 33 Overview of data collected for the M50 seated FEA male. Left—conventional MRI thigh cross section and lateral view of neck; mid-
dle—quasi-seated CT scan and external body laser scan in seated posture; right—full human body model (bone and muscle) [308]

Ll.M.Evans et al.
1 3
and Lu looked at recent research on bioinspired structures
and materials in terms of their energy absorption capabili-
ties, this collated examples from bamboo inspired structures
to a bone-mimetic crash box [321]. Vasarhelyi etal. investi-
gated the use of μCT and 3D analysis in characterisation of
advanced materials in application-wise (bioinspired mate-
rials, structural materials, energy and environment) with a
relatively sparse amount of literature [322]. This section, on
the other hand, attempts to collate the broader literature of
biomimicry applications of IBSim under the titles of three
main IBSim approaches used in biomimicry of materials
in order to design new materials or upgrade and/or repair
existing damaged materials.
2.5.1 Imitation ofBiomaterial Architectures inCAD‑Based
Product Design Using 3D Image Data
The first approach relies on the imitation of biomaterial
architectures in product design, where imaging data is
analysed carefully, and image-based numerical models are
established based on this data for elucidating the mechanics
of material architecture under operating conditions. In this
case researchers establish their numerical models inspired
from natural materials using direct measurements of micro-
structural features, for instance, from CT images rather than
converting 3D segmented image data directly into image-
based meshes. This IBSim approach has been applied in
investigations of wood [323], marine animals [313], balsa
trunk and branches [324, 325], hedgehog quill [326] and
spine [327]. New structures can be designed based on the
architecture of these materials. For instance, a study took
into account cancellous bone structure of human tibia as a
bionic object when structurally designing a novel crash box
with improved crashworthiness and energy absorption per-
formance [228]. Similarly, thin-walled structures with supe-
rior crashworthiness features were designed using the bio-
logical structure of bamboo with energy absorption ability
and the deformation behaviour of the new designs was simu-
lated under longitudinal and lateral compression [328]. In
the first approach, the ultimate design does not have to be in
the same length scale as the inspired material. A comprehen-
sive review has been conducted into bioinspired structures
and materials [321]. These structures are often at a different
length scale to that of the original biomaterial, therefore, the
reviewers decided to organise the literature in terms of their
energy absorption performance. To exemplify such applica-
tions, the wood structure of a Manchurian walnut tree was
scanned by μCT to be converted into FEA models of the
microstructure and, then, the anisotropic micro-structure
was imitated on the macroscale by manufacturing with an
AM process [323]. In another example, armours of Chitons,
which are a family of marine animals, inspired artificial
armours with increased flexibility and protective structures
compared with conventional man-made armours which are
highly rigid structures with flexibility and manoeuvrability
trade-off [313]. The methodology of the research consists of
three critical stages (i) imaging (ii) computational modelling
and (iii) manufacturing of designs. Dimensions and geom-
etry of basic components of chiton’s armour were quanti-
fied from SEM and X-ray images (Fig.34a, b). In this type
of modelling approach, computational structures for FEA
simulations are generated from parametric CAD models,
shown in Fig.34c, originated from the imaging data and not
direct discretisation of 3D segmented volume data. Fig.34d
illustrates the comparison between the CT data-based CAD
design and the segmented 3D volume of the actual chiton
armour components. An advantage of having a parametric
computational model is the capability to test, virtually, many
design scenarios in various loads and boundary conditions,
such as compression (i.e., buckling) in Fig.34e. This led to
identifying an optimum design, performing all the necessary
tasks (bending, buckling, stretching) properly without any
functional loss. A prototype for a design of chiton-inspired
armour (Fig.34f) was fabricated with multi-material 3D
printing AM technology and qualitatively tested.
Furthermore, balsa trunk and branches hosting vertical
hexagonal columns (cellular material) can be implemented
for engineering applications because of their ability to hold
high stiffness under tension and shear [325], and strength
under compression [326]. The cylindrical structure of hedge-
hog quills were the inspiration for stiffeners due to their
bending, ovalisation and buckling resistances [326].
Other previous research has worked in the same length
scale where the information used to build the computa-
tional structures was extracted from the CT images or an
alternative imaging method. For example, X-ray images of
hedgehog spine samples were acquired and rendered in 3D
to obtain the detailed microstructural measurements to be
used to generate discrete micro-accurate FEA models [327].
Differing to continuous models, where homogenisation pro-
cesses are applied over the volume of interest, discrete mod-
els aim to capture physics of mechanical problems by imitat-
ing true architecture of materials to as high a resolution as
experimental measurements have allowed. In the case of the
hedgehog, longitudinal stringers and transverse central sup-
port plates were identified from X-ray images and this infor-
mation was introduced into FEA models of spines [327].
Additionally, the imaging datasets of various biomaterials
obtained from CT or μCT are available in online repositories
such as Digimorph, Morphosource, Gigascience for the use
of researchers [329] that can use the potential of biomimicry
to develop their bioinspired products.

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
2.5.2 Virtual Qualification ofBioinspired Design Using
IBSim onManufactured Parts
The second approach of IBSim in biomimicry makes use
of the imagining data with high resolution simulations to
virtually qualify the constructed biomaterial architectures
[330–332]. The alternative architectures, which are ana-
logues to the real biomaterials in terms of statistical data,
can be manufactured as a replacement to the original mate-
rial. Because IBSim can be performed non-destructively,
it provides an opportunity to replicate virtual mechanical
behaviour of biomaterials under numerous loading condi-
tions on the same sample, an advantage when compared
to destructive in-situ physical investigations such as time-
resolved μCT [333, 334]. IBSim is used for developing body
implants and improving their mechanical properties. One
aim in designing such implants is to reduce weight by uti-
lising their porous architecture while preserving function-
ality [222]. Fig.35 shows two different cases of artificial
porous scaffolds designed for a certain area in trabecular
bone (CAD-based geometry) [335] and a mandible (hybrid
CAD/image-based geometry) [336]. Porous micro or macro-
structure is a major parameter influencing the mechanical
properties, such as ultimate strength of implants. A research
paper reported that customised porous titanium plates were
manufactured to cover two human skulls having undergone
trauma or disease [223]. Rigid cell foams are one type of
porous scaffolds, which are frequently used in bioengineer-
ing applications such as metal bone replacement in ortho-
paedic applications [337], polymeric bones in real vertebra
Fig. 34 a SEM image of chi-
ton’s armour and b segmented
X-ray volume; c sketches of
parametric CAD model; d
comparison between the CT
data-based CAD design and the
segmented 3D volume of the
actual chiton armour compo-
nents; e virtually buckled FEA
model of man-made armour; f
a prototype of chiton-inspired
armour (reproduced from [313])

Ll.M.Evans et al.
1 3
[338], glass–ceramics foam in orbital implants [330], zir-
conia scaffolds in bone tissue engineering [46]. They are
designed to provide highly similar microstructural environ-
ments for clinical experiments. IBSim, therefore, can help
to predict the mechanical response of artificial tissues or
structures in order to prevent potential risks that can occur
during and post medical operations.
In some biomedical applications, load-bearing tissues,
such as trabecular bone, are replaced by artificial tissues or
tissue constructs. μCT-based FEA models were generated
and the mechanical behaviour was simulated to optimise
the microstructural design of scaffolds by Jaecques etal.
[339]. Geometrical parameters of open-porous titanium
scaffolds with cubic, diagonal and pyramidal designs were
numerically optimised to reach the elastic properties of
human cortical bone, where minimum pore size was taken
into account [220]. A set of scaffolds with high strength,
stretch-dominated topologies (tetrahedron and octet trusses)
for bone replacements were fabricated with L-PBF and
tested to understand the influence of cell topology, pore-
size, volumetric porosity on mechanical strength and bone
in-growth [337]. The morphological deviations related to
L-PBF technology was analysed via μCT. Triply periodic
surface microstructures of Ti6Al4V were manufactured by
a laser melting process and their mechanical properties were
investigated with CAD-based and μCT-based FEA mod-
els [340]. These structures are lattices with high porosity,
promising sufficient load bearing capacity for bone implants.
Similarly, the mechanical properties of zirconia foams for
bone tissue engineering was investigated with μCT-based
FEA models, where noise in μCT images was removed and
smooth boundaries were applied before construction of 3D
zirconia foams [46]. The smoothing process retained the
fabrication-induced pores, though artefact-based voids were
eliminated. Importantly, it was found that the strength of
image filtering has a non-negligible effect on the geometry,
e.g., porous structures, and hence material properties like
stiffness (Young’s moduli).
Porous scaffolds with different geometrical parameters
can be produced by various manufacturing methods such
as AM. Their simulation geometries, obtained from direct
processing of X-ray images, can be used to search for the
optimum architecture for cell growth and better mechanical
performance. Scaffolds can be designed with various exter-
nal geometries and various tortuous internal architectures
in an idealised CAD environment; however, fabrication pro-
cesses result in discrepancies, such as surface roughness and
micropores, between the manufactured scaffolds and their
CAD design [341].
Virtual topological optimisation of scaffolds might use
IBSim and AM technologies, where voxel-based CAD
geometries are constructed from 3D image data which are
processed into STL surface geometries [218]. At this point,
a topological optimisation algorithm, which considers a
variety of loads and material parameters such as elasticity
and plasticity in the design of scaffolds, can be applied to
discretised CAD-based FEA simulations. The most efficient
design is then selected to be fabricated with one of the AM
Fig. 35 Design steps to generate a 3D porous scaffold: a selecting the
implantable volume zone (or volume of interest), b 3D homogene-
ous model geometry, c inserting porous trabecular bone like scaffold
[335]; d FEA model of a mandible, scaffold and tissue-engineered
bone graft implants [222]

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
technologies. In order to realise CAD geometries of ultimate
designs into real-world parts, the CAD models are math-
ematically sliced into set of thin layers.
In order to provide adequate mechanical properties
to porous scaffolds for biomimetic applications, they can
be designed and simulated mechanically with CAD, then
optimised by using AM for experimental characterisation
to quantify performance parameters such as stiffness and
permeability [219]. In the same paper, IBSim was used
for virtual qualification of the manufactured part and com-
pared with the expected values based on the initial CAD-
based design. The focus was on Poly-e-caprolactone-4%
hydroxyapatite porous scaffolds with various solid volume
fractions which were numerically generated into voxel-based
FEA models of scaffolds. Consequently, the FEA models
were split into dataset of ‘TIFF’ images with higher resolu-
tion that are segmented to create STL scaffolds. STL files
were then prepared for rapid prototyping of the scaffolds
with L-PBF. However, the nature of rapid prototyping results
in differences between the ‘as designed’ and ‘as manufac-
tured’ scaffold geometries [337]. In light of their compari-
sons, architectures of scaffolds were optimised. Steps from
the design to manufacturing of an optimised porous scaffold
are explained in Fig.36, with further details available in the
original research article [219].
2.5.3 Direct Implementation ofIBSim onSource
Biomaterials
Borah etal. investigated: the roles of μCT and image analy-
sis for a quantitative analysis of trabecular bone architec-
ture; FEA for mechanical behaviour of bone at micro and
macro-levels; physical replicas from rapid prototyping for
enhanced visualisation [342]. This allows the researcher to
understand the effect of bone microstructure on osteoporotic
fractures [343]. μCT-based IBSim can be directly used for
quantifying stress and strain analysis in actual bone tissues
and scaffolds [46]. Performance of scaffolds relies on design
and characterisation of their microstructure and, therefore,
μCT is one of the key instruments in the microstructural
characterisation [344].
To demonstrate the use of IBSim with foam materials, a
commercial synthetic foam (open-cell) was considered as
a replacement to human cadaveric bone Synthetic to simu-
late various in-vitro cases for bone infiltration (see Fig.37)
[332]. The procedure to be followed is listed as i) μCT
scanning of the sample materials and saving the images in
DICOM file format, ii) importing the images in Simpleware
Software (Synopsys Inc.) and applying a list of postprocess-
ing operations—in particular, noise reduction, smoothing,
automated segmentation, iii) construction of foam volume,
iv) Boolean operation to obtain complex flow volume, v)
meshing the flow volume with tetrahedral elements, vi)
applying prescribed boundary conditions, inlet flow veloc-
ity, outlet pressure and slip/nonslip boundaries. In another
similar application of the direct use of IBSim, the peel from
the Pomelo fruit, which has a foam-like hierarchical micro-
structure, was investigated by analysing image data from in-
situ compression tests [334]. This research was carried out
to inspire development of novel materials due to their high
energy absorption efficiency. For inspiration of advanced
biomimetic hydraulic systems, the flow mechanisms of a
hydraulic joint in a spider leg was studied with CFD simula-
tions through a commercial software ANSYS Fluent® where
the flow models were extracted from direct processing of
3D μCT images and supported with appropriate boundary
conditions of high-pressure areas such as inlet, joint and
closed leg ends [345].
A review paper from Jones and Wilcox examined various
strategies used to develop IBSim FEA models of human spines
(in particular, vertebra, intervertebral disc and short spinal seg-
ments) [346]. Each model had three major steps: i) verifica-
tion, ii) sensitivity analysis, iii) validation. These models were
based on direct processing of images. In order to validate the
IBSim results against the experimental tests requires accurate
localised material properties rather than the global effective (or
homogenised) properties which are usually obtained.
Fig. 36 Steps from the design to manufacturing: a initial andb optimised FEA meshes, c stack of TIFF images, d STL images of the scaffold,
and e manufactured scaffolds (redrawn from [219])

Ll.M.Evans et al.
1 3
Multiple data sets for the same scaffolds at different
length scales can be achieved with the IBSim approach.
For instance, the exterior geometry of the scaffold can be
generated with macro-details, and the interior geometry
can be represented with microscale architecture [341]. For
instance, polymeric beads were placed into the solid rods
of ceramic scaffolds to characterise the effects of these
defects on their fracture behaviour under uniaxial compres-
sion [347]. The scaffold macrostructure was scanned at
low-resolution, whereas the rods including artificial defects
were the region of interest and analysed with CT at a higher
resolution. With this method, FEA models of segmented CT
images with sufficiently fine meshes can give an account for
deformation and fracture behaviour at two scales (micro and
macroscales).
Permeability of AM scaffolds designed for healing bone
tissues with defects was investigated through flow IBSim
models [348]. Air or liquid permeability in accordance
with flow behaviour relies on major geometrical micro-
structure parameters, for example, total porosity, pore
shape and sizes, their interconnectivity and abundance
[349, 350]. Permeability of porous scaffolds depends on
nano and macrostructures and plays a key role in the bio-
logical performance of the material [350]. Accurate statis-
tical analyses of these microstructure parameters can be
carried out with IBSim. The permeability coefficients of
porous scaffolds can be computed with virtual tests using
CFD simulations of IBSim models in place of their physi-
cal counterparts. Results can be compared with experi-
mental measurements for verification or simple models
can be verified with analytical or empirical solutions from
literature. Such investigations on CAD designs and actual
scaffold geometries manufactured with AM were carried
out to for skeletal tissue engineering [348]. The pressure
and velocity fields of the scaffolds were computed using a
commercial finite-volume based CFD code (Fluent 6.3®)
for both the CAD and image-based models. The computed
permeability of IBSim models were found to have a higher
degree of accuracy than the CAD-based models when com-
pared with the experimental measurements.
Fig. 37 Flow behaviour over porous scaffold geometry directly
obtained from CT images can be achieved with a list of steps: a stack
of μCT images, b binarisation and smoothing of the selected volume
of interest, c 3D reconstruction and segmentation of the structure, d
Boolean operation over the structural geometry to produce the flow
domain (i.e., CFD domain), e meshing process of IBSim CFD model,
f assigning various boundary conditions (inlet velocity, outlet pres-
sure and wall boundary conditions) [332]

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
3 Summary, Discussion andConclusions
This review has set out to report on applications of IBSim
within HVM. In doing so, discussion of the literature has
been grouped into application spaces within HVM (and
subdivided further within those subsections). That is, char-
acterisation of materials and manufacturing techniques,
quantifying the impact of ‘real’ geometries compared with
idealised ones, customisation of products, and biomimicry.
As a ‘first of a kind’ review in this field there was a signifi-
cant volume of literature to consider. As such, the papers
presented in this work is not an exhaustive list but presents
important milestones in the adoption of IBSim within HVM.
For the research using IBSim for material characterisa-
tion, these works investigated materials such as composite
materials (both fibrous and aggregate), AM materials, and
foams. The feature in common between these material types
were that they all exhibited non-negligible variations from
one instance to another, either by design or as a by-product
of manufacturing processes. They were mostly well-suited
for XCT imaging, due to a beneficial attenuation contrast to
facilitate identifying geometric features. Where this does not
hold true is for fibre composites, where the fibres and matrix
use either the same or similar material, or AM materials with
low pore volume fractions. There is ongoing research within
the imaging field to use methods such as phase contrast CT
to improve such data [351]. The literature demonstrated
that the majority of effort is in characterising mechanical
behaviour or the permeability of materials. The other main
areas of interest were thermal and electrical conductance. A
substantial proportion of the work focussed on using RVEs
or unit cells from a region of interest within the imaged
material to reduce computational expense.
Understandably there was a significant overlap with areas
of interest for characterisation of materials. The main dis-
tinction was that efforts using IBSim in this field focussed
on particular features (e.g., a class of defect) to better under-
stand the cause of their formation. The majority of the fea-
tures of interest (e.g., pores) are inherent by-products of
the manufacturing methods, whilst others were caused by
unexpected issues during manufacturing. IBSim is being
used to better understand what level of these features can be
tolerated, and in some cases used to improve performance
(such as surface roughness to increase heat transfer) or to
improve manufacturing efficiency. Examples were observed
where artificially induced defects were included to investi-
gate the impact of defects on behaviour in greater detail.
Of the research observed, many made recommendations
based on IBSim investigations as to how processes may be
improved, however, none were found that closed this loop
by implementing the improvements suggested within their
own studies.
When considering the applications of IBSim to inves-
tigate the impact of deviations from the ideal on perfor-
mance and on a sector-by-sector basis, literature broadly
fell into three sectors: medical, energy, and aerospace. It
was observed that, until recently, IBSim has generally been
used to investigate material coupons or regions of interest
within larger components rather than performing IBSim
analysis of whole components. There is a wealth of litera-
ture using IBSim in the medical sector to investigate aspects
of the human body (e.g., mechanical behaviour of bone or
flow through the cardiovascular network). The scope of
this review was restricted to examples from the medical
sector, which also included HVM, e.g., the design or use
of an implant. From the volume of available literature to
date, it can be observed that this is currently the main sector
where IBSim is used routinely, and a significant portion of
this is for patient specific applications (see below). For the
energy sector, the main area where IBSim has been used
is in the characterisation of materials in batteries and the
development of novel materials. There are some examples
where IBSim is used for NDT/NDE on the component scale,
however, this approach has not yet reached maturity. For
the aerospace sector a similar distribution of research was
observed, i.e., that IBSim is predominantly used for materi-
als or process development, with some examples in compo-
nent scale NDT/NDE.
As noted, where IBSim is used in HVM for biomedi-
cal purposes, a significant proportion of this is patient spe-
cific applications. These applications include healthcare,
personalised medicine, prosthetics, and orthotics. Avail-
able literature shows how IBSim has been used to design
implants, plan procedures, and monitor deployment of per-
sonalised products over time. Compared to applications of
IBSim for HVM in other sectors presented in this review,
these examples are comparatively mature. There is, how-
ever, still significant potential for further uses, and further
work is required to achieve acceptance of highly personal-
ised approaches with regulatory bodies. Examples were also
observed where IBSim was used to personalise or customise
sport, lifestyle, and automotive products outside biomedi-
cal applications. For example, to improve the efficiency of
protective sport equipment for an individual or to improve
the accuracy of digital vehicle crash testing using virtual
humans from image data.
Stemming from the fact that IBSim has its origins in bio-
medical engineering, it was observed that there was a sig-
nificant volume of research using IBSim for HVM in relation
to biomimicry and bio-inspired engineering design. These
works were not only inspired by aspects of the human body,
but also the natural world. Hundreds of millions of years of
evolution have generated structures that are highly efficient
at their given functions. The observed literature demon-
strates that IBSim is enabling researchers to investigate how

Ll.M.Evans et al.
1 3
these organically generated structures perform in a detail not
previously possible. The work surveyed using IBSim with
biomaterials broadly fell into three categories:
– Use of 3D image data from biomaterials to inspire engi-
neering designs, which in turn are tested with conven-
tional simulation analysis methods.
– Virtually testing bioinspired parts ‘as manufactured’ to
gain insight at the microscale.
– Direct implementation of IBSim on source biomaterials
for a better understanding of their behaviour.
Despite the broad range of application spaces where
IBSim has been observed to be used within HVM, there
are a number of commonalities which are noteworthy.
Firstly, the imaging and simulation methods observed in
the research were predominantly XCT and FEA or CFD.
This can be attributed to the fact that this review focuses
IBSim applications in HVM, i.e., a sector where the use
of CAD-based FEA/CFD is already commonplace for com-
ponent design, and there is significant year-on-year growth
of XCT for metrological characterisation. That is, where
IBSim was observed to be used, the imaging and simulation
techniques were already being employed and IBSim was a
method to extract additional value from data that was already
routinely generated. By combining these commonly used
methodologies, it is observed that IBSim has been providing
researchers with improved levels of accuracy with predictive
simulations that were previously unobtainable.
This observation is broadly supported by the recent
growth in publications on these topics, see Fig.38. This data
was collected from the Scopus database of literature, with
the searches being restricted to the ‘Engineering’ and ‘Mate-
rials Science’ subject areas and the ‘Article’ document type.
To account for the general year-on-year growth in published
research articles, in Fig.38 (area) the changes are displayed
as a percentage of the total number of papers published in
the subject areas in question. The ‘IBSim’ curve in Fig.38
(line) is given as the absolute number of articles published.
Data for 2021–2028 are extrapolated using a 3rd order poly-
nomial. The search terms used were:
• Numerical Simulation = “numerical simulation” OR
“computational engineering”.
• Tomography = “tomography”.
• Both = (“numerical simulation” OR “computational engi-
neering”) AND “tomography”.
• Others = The remaining articles published in the ‘Engi-
neering’ and ‘Materials Science’ subject areas.
• IBSim = ‘Both’ search terms OR “image-based simula-
tion” OR “image-based modelling”.
By combining these commonly used methodologies, it
is observed that IBSim has been providing researchers with
Fig. 38 Change in publications by year: (area) as a percentage of the
total publications considered; (line) absolute number. The area data
was found using the search terms (“numerical simulation” OR “com-
putational engineering”); (“tomography”); ‘Both’ denotes, the cross-
over of these two search terms. The line data was found using the
search term ‘(tomography AND “computational engineering”) OR
(tomography AND “numerical simulation”) OR (“image-based simu-
lation” OR “image-based modelling”)’. Data collected from Scopus
whilst restricting the search to the ‘Engineering’ and ‘Materials Sci-
ence’ subject areas and the ‘Article’ document type. Data for 2021-
2028 are extrapolated using a 3rd order polynomial

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
improved levels of accuracy with predictive simulations that
were previously unobtainable.
There were also some caveats, limitations and chal-
lenges that were common to many studies. The availability
of simulation geometries at newly achievable resolutions in
particular brought a set of challenges for consideration. A
major one of these is which material properties should be
used and how to collect these. This is because simulations,
such as FEA, conventionally treat materials as homogeneous
and thus use experimental data that is typically collected on
macroscale samples, whereas those material properties do
not necessarily hold true on the microscale. Additionally,
simulations performed on the same size of component, but
at a significantly higher resolution (i.e., including micro-
features within macro-models), increases the computational
demands both in terms of solving the equations and visu-
alising the results. For a typical FEA analyst, CAD-based
simulations are mostly carried out using desktop PCs which
have higher power hardware that those used for conventional
office work. However, these are only moderately powered
relative to high-end workstations available on the market
and significantly less powerful than supercomputing facili-
ties. On these types of machines, the lack of available RAM
and computing power can make it unfeasibly slow or, at
worst, impossible to perform IBSim research [352]. Another
recurring theme was that simulation results were highly sen-
sitive to image quality and the associated post-processing
of images. A significant factor in this is variability between
imaging systems (even of the same type) and the subjectiv-
ity of operators.
Based on the observed literature, the reported benefits in
applying IBSim within HVM for virtual testing can broadly
be categorised as follows:
– Can use to accurately replicate laboratory experiments
for a direct comparison of results. This is beneficial to
facilitate validation of numerical models or to gain addi-
tional insight to experimental results. That is, experimen-
tal results tend to give the overall global response (or
sparse local data), whereas it is possible to investigate
local response through the full volume with IBSim.
– Can perform virtual tests not possible experimentally.
This can be of benefit to perform a simplified experiment
which more rapidly measures the feature of interest (e.g.,
applying a thermal gradient across a volume to measure
effective thermal conductivity) or to apply conditions
that would be too difficult to apply in a laboratory (e.g.,
extreme loads). Similarly, laboratory tests are often set
up to approximate in-service conditions, whereas these
can be applied directly to the real geometry with IBSim
without the constraints of a lab.
– Can perform a series of virtual destructive tests on the
exact same sample to remove variability of geometry
(and the consequential variability on results). Similarly,
if sufficient computing power is available, many virtual
tests can be run in parallel to accelerate the generation
of data without the requirement to buy duplicate testing
hardware.
The majority of research published in peer review jour-
nals using IBSim applies the technique to coupon scale
samples or RVEs. Alternatively, the research uses a hybrid
CAD+IBSim approach where IBSim is used to investigate
a region of interest, or a component is designed by CAD
to conform to an image-based geometry of the object with
which it is to interact. Comparatively, there is very little
literature on component scale modelling done purely with
IBSim. Although it is difficult to directly demonstrate cau-
sality, it is the authors’ opinion this can be attributed to a
few different factors. For example, a highly detailed IBSim
model can be many orders of magnitude more computa-
tionally expensive than a CAD-based equivalent, leading
to requiring national-level high-performance computing
facilities to which not all researchers readily have available
access. Smaller scale ROI or RVE models can be used to
mitigate this issue. Imaging full scale components, then, has
two associated challenges. Firstly, as the size of the com-
ponent increases, the level of achievable resolution often
reduces. This could mean that the micro-scale features of
interest which have a non-negligible impact on the compo-
nent performance might not be resolvable in a volumetric
image of the whole component. Secondly, for techniques
like XCT, larger components lead to higher signal attenu-
ation. If the component is sufficiently large it could mean
that volumetric imaging is no longer possible with the same
equipment. Despite this, it is known to the authors that there
is an appreciable amount of R&D activity occurring within
the industrial sector which utilises IBSim for component
scale modelling. Unfortunately, this work is considered valu-
able intellectual property and therefore not published in peer
review journals.
Another point noted concerning the current state of IBSim
literature is that none reported implementing the improve-
ments suggested by their own studies. Due to the detrimen-
tal influences of different defect types within HVM, there
is a strong need for defects to be minimised or mitigated
by quality control (e.g., ensuring clean material feedstock
to limit inclusions, ensuring a stable manufacturing process,
etc.) or by process optimisation (adjusting process parameters
e.g., power or scan speed, etc.). The extent to which this pro-
cess optimisation is required depends on the strength of the
influence, i.e., what is the effect of the defect exactly on the
mechanical performance, and what is the expected mechani-
cal performance of the part? This is where IBSim can be
useful in combination with experimental work, to reveal for
a specific material, geometry, and combination of defects,

Ll.M.Evans et al.
1 3
what the influence is on the mechanical properties and evalu-
ate whether this is acceptable or not. This is further possible
not only for improving manufacturing processes, but also for
evaluating post-processing and wear or degradation after use.
Post processing of metal parts often include heat treatment,
machining, and sometimes more advanced processes such as
hot isostatic pressing, which improves the microstructure and
closes porosity, resulting in improved mechanical properties
[353, 354]. IBSim may be used to evaluate the size of pores
or other defects after processing or due to degradation of the
material and evaluate the effect of the defect on the expected
performance of the part. For IBSim to demonstrate its true
value it is important that future work is published reporting
measured benefits gained rather than only potential ones.
Finally, it is worth noting that a significant proportion of
the research in the literature using IBSim does so within the
highly regulated medical and aerospace sectors (see following
sub-section for more details). These sectors are, understand-
ably, known as being risk averse, however, they are also known
for being pioneers of new technology. For IBSim, having a
growing body of work from within these sectors should lead to
increased confidence as the technique matures, which will sup-
port wider adoption in other regulated sectors such as energy.
3.1 Future Trends
In this section, the authors present developments in associ-
ated fields of research and discuss how these could have
a direct impact on the future use of IBSim. These are
divided into developments which might bring about evo-
lutionary and revolutionary changes presented in their own
sub-sub-sections.
3.1.1 Evolutionary Changes
The easily predictable evolutionary advancements are those
associated with improvements in hardware (both imag-
ing and computational) and software, which are continu-
ally being gradually improved and are included in publicly
available manufacturer development roadmaps. For exam-
ple, manufacturers of imaging apparatus regularly release
new versions that produce data with a higher resolution or
improved image quality (e.g., reduced noise). This is obvi-
ously beneficial for producing more detailed volumetric
images, or images at the same resolution, but producing
results with a greater level of confidence. A by-product of
these improvements is the ability to perform faster imag-
ing for a higher throughput, something of great value to
researchers using time-resolved imaging. For XCT in par-
ticular, a development of value for HVM is the availability
of X-ray sources capable of higher energies. This is allow-
ing researchers to image larger components or those made
from materials that are high attenuators of X-rays. Similarly,
computing hardware is constantly improving, allowing
processing at faster rates and for larger datasets, both for
imaging and simulation algorithms. Furthermore, there is
ongoing research into improving algorithms for more effi-
ciently utilising computational hardware (e.g., better paral-
lelism including use of GPUs). Furthermore, algorithms are
being developed to bring additional benefits, e.g., reduced
artefacts during image reconstruction or higher simulation
accuracies with adaptive meshing. There is also research
into a method to convert 2D image projections directly into
FEA meshes [355], which would remove the need for many
of the interim workflow stages. The limitations caused by
computational expense was mentioned in a significant pro-
portion of papers as a factor leading to the choice to investi-
gate ROIs. As already noted, this review has identified that
IBSim is currently only being used in a limited way on the
component-scale. The authors of this review believe there
will be a significant shift towards more activity on the com-
ponent scale through wider availability of higher power XCT
devices and further improvements in computing power.
Other potential evolutional developments of IBSim are
related to the way in which the techniques that are part of
the workflow are being implemented within the HVM sec-
tor. For example, FEA is now a common tool in HVM, and,
therefore, a set of ‘best practices’ have developed within the
community that uses them, which are often formalised as
international standards, e.g., in aerospace [356]. Since the
use of advanced imaging techniques such as XCT in HVM
is a relatively recent addition, the standards surrounding the
methodologies are still in their infancy. As these mature this
will facilitate repeatability in results, both in the individual
method and for IBSim that combines them. It is important for
the acceptance of IBSim within HVM that there is industry-
level confidence in each of the workflow components.
It is also important to consider the apparent conflict in
requirements between regulation and personalisation. For
example, in personalised medicine, when the Federal Food,
Drug, and Cosmetic Act (FDA) was passed in 1938, the
term “Personalised Medicine” had not yet been coined. The
standards for FDA approval for regulating medical devices
(including laboratory tests) were established for traditional
products in 1976, but the complexities associated with
obtaining approvals for personalised medical products
have since proved challenging. This echoes the challenges
also seen in other highly regulated sectors, e.g., aerospace,
whereby ‘designs codes’ and standards exist with the aim
of ensuring a predictable performance through prescribed
approaches (i.e., constraints) in design and manufacturing.
This is in stark contrast to the approach of IBSim, which
is to yield the same level of performance prediction whilst
allowing design freedom.
At the core of precision medicine lies diagnostic tests and
devices, but the regulatory classification of such products

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
varies globally [357, 358]. Each regulatory agency has a
clearly defined definition of what constitutes a custom-made
device (CMD) or personalised medical device (PMD). For
example, in the UK, any mass-produced device adapted
to specific patient requirements post-production does not
fall under the UK Medical Device Regulation definition
for CMDs (e.g., optical glasses, patient-fitted wheelchairs,
hearing aids and orthotic braces). Furthermore, it excludes
mass-produced devices manufactured via industrial pro-
cesses, even if produced according to written prescriptions
[359]. The European Medical Device Coordination Group
(MDCG) have also ensured that the manufacturers are solely
accountable for patient-matched devices in terms of design,
safety, performance and regulatory compliance [360].
The primary issues of these regulatory agencies involved
concerns as to whether patients should be able to purchase
and use unapproved or unregulated tests; and whether man-
ufacturers can be trusted to market tests that conform to
standards for safety and efficacy if the process of regulating
these tests is too slow and expensive. Navigating these con-
voluted regulatory pathways can be challenging, and whilst
the guidance for CMDs has improved these past few years,
the regulation regarding a completely personalised part is
currently impeded by incoherence between policymakers
and regulators [361]. New regulatory reforms, however, spe-
cific to CMD’s (e.g., Software as a Medical Device (SaMD),
PMDs and Medical Device Production Systems (MDPS))
are beginning to form, which highlights that these changes
are occurring, albeit slowly [357]. A similar response is
being observed in other sectors.
The final way in which IBSim is evolving is the applica-
tion space in which it is being used, that is, the products
being manufactured that are driving a demand for advanced
characterisation methods with increased accuracy. A con-
tinual desire to produce more efficient products is bring-
ing about the use of increasingly advanced materials with
complex behaviours using novel manufacturing methods. In
addition to this there is an increasing demand for personali-
sation of products. The combination of these factors presents
an increasing stream of new opportunities where there exists
a significant variation from part to part, or for those which
are unique.
This review has shown that IBSim is ideally suited to
this, and it has been found to play a critical part in obtaining
patient- or subject- specific 3D geometry, where the ability
to run multiple simulations that closely mimic the in-vivo
response of the human body, or of a real-life scenario, are
invaluable. The continued advancement and availability
of imaging techniques, coupled with open access software
and AM, will further enhance research and, subsequently,
products in these areas. The personalisation fulfilment is,
therefore, slowly leading to machine learning-driven intel-
ligent configuration, and Industry 4.0-driven on-demand
production, sometimes using manufacturing-as-a-service
or 3D printing.
3.1.2 Revolutionary Changes
The prediction of revolutionary advancements in any tech-
nology is inherently more challenging; however, machine
learning (ML) is one such disruptive technique that has been
noticeably transforming most scientific fields over the past
decade. There is much recent research with relevance to
IBSim. In imaging, for example, optimising imaging setup
[362] and the reconstruction of volumetric images [363]
to improve scanning speed and image quality by orders of
magnitude. Machine learning is proving to be well-suited to
automate tasks. For example, automation of image segmen-
tation not only speeds up this section of the workflow by
orders of magnitude, but, for the first time, generates repro-
ducible results [364, 365]. For simulations, IBSim offers the
potential of actual ‘digital twins’, whereby each manufac-
tured part has a digital equivalent capable of providing real
time feedback for an in-service product. This is challenging
for conventional CAD-based simulations and even more so
for IBSim using FEA models. However, ML is being used
to produce FEA surrogate models capable of making predic-
tions in a fraction of the time of a full simulation [366]. A
recent paper by Ezhov etal. proposed an AI system based on
deep learning methods for dental diagnosis with CBCT. The
AI system was found to significantly improve the diagnostic
capabilities of dentists and has the potential to augment the
dentists’ routine clinical practice [367]. The full realisation
of these applications of ML will transform the potential of
IBSim from being a technique requiring significant resource
to use, preserved for the most well-equipped laboratories,
to one that could be commonplace in a smart factory of the
future.
A technological advancement that offers a significant
opportunity for IBSim is augmented reality (AR). In the
modern, highly competitive manufacturing environment, the
application of AR consists of an innovative and effective
solution to simulate, assist, and improve the manufacturing
and maintenance processes. Today, a growing number of
applications based on AR solutions are being developed for
industrial purpose. A systematic review by Baroroh etal.
reported on recent AR applications in smart manufactur-
ing from a human–machine interaction perspective [368].
In another review, Bottani etal. reviewed the literature from
2006 to 2017 to identify the main areas and sectors where
AR is currently deployed, describe the technological solu-
tions adopted, as well as the main benefits achievable with
this kind of technology [369]. In particular examples, 3D
scanning is being used in conjunction with AR, for example,
to simulate virtual ‘try-on’ technology, where fit and size
issues of mass customised men’s jackets have been explored

Ll.M.Evans et al.
1 3
using 3D body scanning and 3D virtual simulation tech-
nology [370]. AR-based design customization of footwear
for children is also widely documented [371]. Presently, the
combining of this with detailed IBSim is likely to be too
computationally challenging, however, using this technol-
ogy alongside ML surrogate IBSim models could allow
real-time feedback to the user (which could be human, or
an AI-driven robot) that predicts the outcome of a rapidly
developing situation, e.g., during layup of fibre laminates in
composite manufacturing.
Another field that may bring transformative change is the
introduction of novel imaging techniques able to generate
rich image data which includes additional information about
the material within the sample, analogous to imaging with
backscattered electrons, energy dispersive spectroscopy, and
secondary electrons in SEM. Not only do these methods
promise to yield information sufficient to generate volu-
metric maps of material types, but provide additional infor-
mation such as the material’s phase, their stress state, and
porosity of sizes lower than the image resolution [372–376],
all of which can be used to greatly enhance the predictive
capability of IBSim.
The use of IBSim within HVM emerged in the early
2000s but exhibited a low rate of growth during that first
decade. During the 2010s, there was wider use within aca-
demia across a broad range of research fields, coinciding
with wider availability of volumetric imaging hardware. The
authors of this review confidently believe that IBSim will
enjoy widespread growth within the industrial sector during
the 2020s and become an invaluable NDT/NDE method as
part of the prevalence of Industry 4.0.
Funding Funding was provided by Engineering and Physical Sciences
Research Council (Grant Number EP/R012091/1).
Open Access This article is licensed under a Creative Commons Attri-
bution 4.0 International License, which permits use, sharing, adapta-
tion, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons licence, and indicate if changes
were made. The images or other third party material in this article are
included in the article's Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in
the article's Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a
copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.
References
1. Withers PJ, Bouman C, Carmignato S, Cnudde V, Grimaldi D,
Hagen CK, Maire E, Manley M, Du Plessis A, Stock SR (2021)
X-ray computed tomography. Nat Rev Methods Primers 1:1–21.
https:// doi. org/ 10. 1038/ s43586- 021- 00015-4
2. Schlüter S, Sheppard A, Brown K, Wildenschild D (2014) Image
processing of multiphase images obtained via X-ray microtomog-
raphy: a review. Water Resour Res 50:3615–3639. https:// doi. org/
10. 1002/ 2014W R0152 56
3. Rizwan I, Haque I, Neubert J (2020) Deep learning approaches
to biomedical image segmentation. Inform Med Unlocke.
18:100297. https:// doi. org/ 10. 1016/j. imu. 2020. 100297
4. Livesey F (2006) Confederation of British Industry, University
of Cambridge, Institute for Manufacturing, Defining high value
manufacturing, CBI, London. http:// www. cbi. org. uk/ ndbs/ posit
iondoc. nsf/ 1f08e c6171 1f297 68025 672a0 055f7 a8/ D70D2 FE009
BC9B7 08025 71340 042F6 A3/ $file/ hvmfu llrep ort. pdf. Accessed
7 Apr 2022.
5. Pieper SD, Laub DRJ, Rosen JM (1995) A finite-element
facial model for simulating plastic surgery. Plast Reconstr Surg
96:1100–1105
6. Wasserman R, Acharya R, Sibata C, Shin KH (1996) A patient-
specific invivo tumor model. Math Biosci 136:111–140. https://
doi. org/ 10. 1016/ 0025- 5564(96) 00045-4
7. Cebral JR, Castro MA, Appanaboyina S, Putman CM, Millan
D, Frangi AF (2005) Efficient pipeline for image-based patient-
specific analysis of cerebral aneurysm hemodynamics: technique
and sensitivity. IEEE Trans Med Imaging 24:457–467. https://
doi. org/ 10. 1109/ TMI. 2005. 844159
8. Hampel U (2015) 6 - X-ray computed tomography. In: Wang M
(ed) Industrial tomography. Woodhead Publishing, Sawston, pp
175–196
9. Sun W, Wong T (2018) Prediction of permeability and formation
factor of sandstone with hybrid lattice Boltzmann/finite element
simulation on microtomographic images. Int J Rock Mech Min
Sci 106:269–277. https:// doi. org/ 10. 1016/j. ijrmms. 2018. 04. 020
10. Brachetta-Aporta N, Toro-Ibacache V (2021) Differences in
masticatory loads impact facial bone surface remodeling in an
archaeological sample of South American individuals. J Archaeol
Sci Rep 38:103034. https:// doi. org/ 10. 1016/j. jasrep. 2021. 103034
11. Manning PL, Margetts L, Johnson MR, Withers PJ, Sellers WI,
Falkingham PL, Mummery PM, Barrett PM, Raymont DR (2009)
Biomechanics of dromaeosaurid dinosaur claws: application of
X-ray microtomography, nanoindentation, and finite element
analysis. Anat Rec 292:1397–1405. https:// doi. org/ 10. 1002/ ar.
20986
12. Huo J, Dérand P, Rännar L-E, Hirsch J-M, Gamstedt EK (2015)
Failure location prediction by finite element analysis for an addi-
tive manufactured mandible implant. Med Eng Phys 37:862–869.
https:// doi. org/ 10. 1016/j. meden gphy. 2015. 06. 001
13. Pawlowski S, Nayak N, Meireles M, Portugal CAM, Velizarov S,
Crespo JG (2018) CFD modelling of flow patterns, tortuosity and
residence time distribution in monolithic porous columns recon-
structed from X-ray tomography data. Chem Eng J 350:757–766.
https:// doi. org/ 10. 1016/j. cej. 2018. 06. 017
14. Teichmann F, Ziemer A, Leitner M, Hensel J, Dilger K (2020)
Linear elastic FE-analysis of porous laser welded, heat treatable,
aluminium high pressure die castings based on x-ray computed
tomography data. Materials 13:1420. https:// doi. org/ 10. 3390/
ma130 61420
15. Finite Element Analysis Market Share, Size and Industry Growth
Analysis 2021 - 2026 (n.d.). https:// www. indus tryarc. com/ Resea
rch/ Finite- Eleme nt- Analy sis- Mark e t- R esea r c h- 502582. Accessed
6 Jan 2022.
16. Evans LlM, Margetts L, Casalegno V, Lever LM, Bushell J, Lowe
T, Wallwork A, Young P, Lindemann A, Schmidt M, Mummery
PM (2015) Transient thermal finite element analysis of CFC–
Cu ITER monoblock using X-ray tomography data. Fusion Eng
Design 100:100–111. https:// doi. org/ 10. 1016/j. fusen gdes. 2015.
04. 048

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
17. Yu J, Zhou C, Zhang H (2017) A micro-image based recon-
structed finite element model of needle-punched C/C composite.
Compos Sci Technol 153:48–61. https:// doi. org/ 10. 1016/j. comps
citech. 2017. 09. 029
18. Sun Y, Li QM (2018) Dynamic compressive behaviour of cellular
materials: a review of phenomenon, mechanism and modelling.
Int J Impact Eng 112:74–115. https:// doi. org/ 10. 1016/j. ijimp eng.
2017. 10. 006
19. Bultreys T, De Boever W, Cnudde V (2016) Imaging and image-
based fluid transport modeling at the pore scale in geological
materials: a practical introduction to the current state-of-the-art.
Earth Sci Rev 155:93–128. https:// doi. org/ 10. 1016/j. earsc irev.
2016. 02. 001
20. Coleri E, Harvey JT, Yang K, Boone JM (2012) Development of
a micromechanical finite element model from computed tomog-
raphy images for shear modulus simulation of asphalt mixtures.
Constr Build Mater 30:783–793. https:// doi. org/ 10. 1016/j. conbu
ildmat. 2011. 12. 071
21. Yang W (2019) Finite element model of concrete material based
on CT image processing technology. J Vis Commun Image Rep-
resent 64:102631. https:// doi. org/ 10. 1016/j. jvcir. 2019. 102631
22. Sun B, Huang X, Zheng Y, Guo L (2020) Multi-scale lattice
method for mesoscopic crack growth simulation of concrete
structures. Theoret Appl Fract Mech 106:102475. https:// doi.
org/ 10. 1016/j. tafmec. 2020. 102475
23. Qsymah A, Sharma R, Yang Z, Margetts L, Mummery P (2017)
Micro X-ray computed tomography image-based two-scale
homogenisation of ultra high performance fibre reinforced
concrete. Constr Build Mater 130:230–240. https:// doi. org/ 10.
1016/j. conbu ildmat. 2016. 09. 020
24. Huang Y, Yang Z, Ren W, Liu G, Zhang C (2015) 3D meso-scale
fracture modelling and validation of concrete based on in-situ
X-ray computed tomography images using damage plasticity
model. Int J Solids Struct 67–68:340–352. https:// doi. org/ 10.
1016/j. ijsol str. 2015. 05. 002
25. Wang Y-S, Dai J-G (2017) X-ray computed tomography for pore-
related characterization and simulation of cement mortar matrix.
NDT and E Int 86:28–35. https:// doi. org/ 10. 1016/j. ndtei nt. 2016.
11. 005
26. du Plessis A, Boshoff WP (2019) A review of X-ray computed
tomography of concrete and asphalt construction materials. Con-
str Build Mater 199:637–651. https:// doi. org/ 10. 1016/j. conbu
ildmat. 2018. 12. 049
27. De Chiffre L, Carmignato S, Kruth J-P, Schmitt R, Weckenmann
A (2014) Industrial applications of computed tomography. CIRP
Ann 63:655–677. https:// doi. org/ 10. 1016/j. cirp. 2014. 05. 011
28. Bale H, Blacklock M, Begley MR, Marshall DB, Cox BN, Ritchie
RO (2012) Characterizing three-dimensional textile ceramic
composites using synchrotron X-ray micro-computed-tomog-
raphy. J Am Ceram Soc 95:392–402. https:// doi. org/ 10. 1111/j.
1551- 2916. 2011. 04802.x
29. Rinaldi RG, Blacklock M, Bale H, Begley MR, Cox BN (2012)
Generating virtual textile composite specimens using statistical
data from micro-computed tomography: 3D tow representations.
J Mech Phys Solids 60:1561–1581. https:// doi. org/ 10. 1016/j.
jmps. 2012. 02. 008
30. Blacklock M, Bale H, Begley M, Cox B (2012) Generating
virtual textile composite specimens using statistical data from
micro-computed tomography: 1D tow representations for the
binary model. J Mech Phys Solids 60:451–470. https:// doi. org/
10. 1016/j. jmps. 2011. 11. 010
31. Badel P, Vidal-Sallé E, Maire E, Boisse P (2008) Simulation and
tomography analysis of textile composite reinforcement defor-
mation at the mesoscopic scale. Compos Sci Technol 68:2433–
2440. https:// doi. org/ 10. 1016/j. comps citech. 2008. 04. 038
32. Sencu RM, Yang Z, Wang YC, Withers PJ, Soutis C (2020) Mul-
tiscale image-based modelling of damage and fracture in car-
bon fibre reinforced polymer composites. Compos Sci Technol
198:108243. https:// doi. org/ 10. 1016/j. comps citech. 2020. 108243
33. Creveling PJ, Whitacre WW, Czabaj MW (2019) A fiber-segmen-
tation algorithm for composites imaged using X-ray microtomog-
raphy: development and validation. Compos A Appl Sci Manuf
126:105606. https:// doi. org/ 10. 1016/j. compo sitesa. 2019. 105606
34. Ali J, Berre C, Mummery PM (2006) Image based modelling
of stress–strain behaviour in carbon/carbon composites. Energy
Mater 1:179–186. https:// doi. org/ 10. 1179/ 17489 2406X 160633
35. Kishimoto H, Shinohara Y, Naito M, Takeuchi A, Uesugi K,
Suzuki Y, Amemiya Y (2013) Visualization of nanoscale defor-
mation in polymer composites with zernike-type phase-contrast
X-ray microscopy and the finite element method. Polym J 45:64–
69. https:// doi. org/ 10. 1038/ pj. 2012. 201
36. Doroszko M, Seweryn A (2017) A new numerical modelling
method for deformation behaviour of metallic porous materi-
als using X-ray computed microtomography. Mater Sci Eng A
689:142–156. https:// doi. org/ 10. 1016/j. msea. 2017. 02. 055
37. Veyhl C, Fiedler T, Jehring U, Andersen O, Bernthaler T, Belova
IV, Murch GE (2013) On the mechanical properties of sintered
metallic fibre structures. Mater Sci Eng A 562:83–88. https:// doi.
org/ 10. 1016/j. msea. 2012. 11. 034
38. Doroszko M, Seweryn A (2015) Numerical modeling of the ten-
sile deformation process of sintered 316L based on microtomog-
raphy of porous mesostructures. Mater Des 88:493–504. https://
doi. org/ 10. 1016/j. matdes. 2015. 09. 006
39. Siddique S, Imran M, Rauer M, Kaloudis M, Wycisk E, Emmel-
mann C, Walther F (2015) Computed tomography for charac-
terization of fatigue performance of selective laser melted parts.
Mater Des 83:661–669. https:// doi. org/ 10. 1016/j. matdes. 2015.
06. 063
40. Padilla E, Jakkali V, Jiang L, Chawla N (2012) Quantifying the
effect of porosity on the evolution of deformation and damage
in Sn-based solder joints by X-ray microtomography and micro-
structure-based finite element modeling. Acta Mater 60:4017–
4026. https:// doi. org/ 10. 1016/j. actam at. 2012. 03. 048
41. Amani Y, Dancette S, Delroisse P, Simar A, Maire E (2018)
Compression behavior of lattice structures produced by selective
laser melting: X-ray tomography based experimental and finite
element approaches. Acta Mater 159:395–407. https:// doi. org/
10. 1016/j. actam at. 2018. 08. 030
42. Amani Y, Dancette S, Maire E, Adrien J, Lachambre J (2018)
Two-scale tomography based finite element modeling of plastic-
ity and damage in aluminum foams. Materials 11:1984. https://
doi. org/ 10. 3390/ ma111 01984
43. Wang P, Lei H, Zhu X, Chen H, Fang D (2019) Influence of
manufacturing geometric defects on the mechanical properties
of AlSi10Mg alloy fabricated by selective laser melting. J Alloys
Compd 789:852–859. https:// doi. org/ 10. 1016/j. jallc om. 2019. 03.
135
44. Zenzinger G, Bamberg J, Ladewig A, Hess T, Henkel B, Satzger
W (2015) Process monitoring of additive manufacturing by using
optical tomography. Boise. https:// doi. org/ 10. 1063/1. 49146 06
45. Grasso M, Colosimo BM, Slattery K, MacDonald E (2021)
Process monitoring of laser powder bed fusion. Fundamentals
of laser powder bed fusion of metals. Elsevier, New York, pp
301–326
46. Askari E, Cengiz IF, Alves JL, Henriques B, Flores P, Fredel MC,
Reis RL, Oliveira JM, Silva FS, Mesquita-Guimarães J (2020)
Micro-CT based finite element modelling and experimental char-
acterization of the compressive mechanical properties of 3-D
zirconia scaffolds for bone tissue engineering. J Mech Behav
Biomed Mater 102:103516. https:// doi. org/ 10. 1016/j. jmbbm.
2019. 103516

Ll.M.Evans et al.
1 3
47. Caty O, Maire E, Youssef S, Bouchet R (2008) Modeling the
properties of closed-cell cellular materials from tomography
images using finite shell elements. Acta Mater 56:5524–5534.
https:// doi. org/ 10. 1016/j. actam at. 2008. 07. 023
48. Cho H-H, Cho Y, Han HN (2015) Finite element analysis for
mechanical response of Ti foams with regular structure obtained
by selective laser melting. Acta Mater 97:199–206. https:// doi.
org/ 10. 1016/j. actam at. 2015. 07. 003
49. Ghazi A, Berke P, Tiago C, Massart TJ (2020) Computed tomog-
raphy based modelling of the behaviour of closed cell metal-
lic foams using a shell approximation. Mater Des 194:108866.
https:// doi. org/ 10. 1016/j. matdes. 2020. 108866
50. Jeon I, Asahina T, Kang K-J, Im S, Lu TJ (2010) Finite element
simulation of the plastic collapse of closed-cell aluminum foams
with X-ray computed tomography. Mech Mater 42:227–236.
https:// doi. org/ 10. 1016/j. mechm at. 2010. 01. 003
51. Zhang L, Ferreira JMF, Olhero S, Courtois L, Zhang T, Maire
E, Rauhe JChr (2012) Modeling the mechanical properties of
optimally processed cordierite–mullite–alumina ceramic foams
by X-ray computed tomography and finite element analysis. Acta
Mater 60:4235–4246. https:// doi. org/ 10. 1016/j. actam at. 2012. 04.
025
52. Veyhl C, Belova IV, Murch GE, Fiedler T (2011) Finite element
analysis of the mechanical properties of cellular aluminium based
on micro-computed tomography. Mater Sci Eng A 528:4550–
4555. https:// doi. org/ 10. 1016/j. msea. 2011. 02. 031
53. Berkache K, Deogekar S, Goda I, Picu RC, Ganghoffer J-F
(2017) Construction of second gradient continuum models for
random fibrous networks and analysis of size effects. Compos
Struct 181:347–357. https:// doi. org/ 10. 1016/j. comps truct. 2017.
08. 078
54. Sozumert E, Farukh F, Sabuncuoglu B, Demirci E, Acar M, Pour-
deyhimi B, Silberschmidt VV (2018) Deformation and damage
of random fibrous networks. Int J Solids Struct. https:// doi. org/
10. 1016/j. ijsol str. 2018. 12. 012
55. Chen Y, Siegmund T (2019) Mechanics of compaction of a
porous non-woven fiber solid. Mech Mater 137:103101. https://
doi. org/ 10. 1016/j. mechm at. 2019. 103101
56. Alimadadi M, Lindström SB, Kulachenko A (2018) Role of
microstructures in the compression response of three-dimen-
sional foam-formed wood fiber networks. Soft Matter 14:8945–
8955. https:// doi. org/ 10. 1039/ C7SM0 2561K
57. Chen N, Silberstein MN (2018) Determination of bond strengths
in non-woven fabrics: a combined experimental and computa-
tional approach. Exp Mech 58:343–355. https:// doi. org/ 10. 1007/
s11340- 017- 0346-3
58. Miettinen A, Hendriks CLL, Chinga-Carrasco G, Gamstedt EK,
Kataja M (2012) A non-destructive X-ray microtomography
approach for measuring fibre length in short-fibre composites.
Compos Sci Technol 72:1901–1908. https:// doi. org/ 10. 1016/j.
comps citech. 2012. 08. 008
59. Yang M, Ji M, Taghipour E, Soghrati S (2018) Cross-linked
fiberglass packs: microstructure reconstruction and finite ele-
ment analysis of the micromechanical behavior. Comput Struct
209:182–196. https:// doi. org/ 10. 1016/j. comps truc. 2018. 08. 014
60. Blunt MJ, Bijeljic B, Dong H, Gharbi O, Iglauer S, Mostaghimi
P, Paluszny A, Pentland C (2013) Pore-scale imaging and model-
ling. Adv Water Resour 51:197–216. https:// doi. org/ 10. 1016/j.
advwa tres. 2012. 03. 003
61. Varloteaux C, Békri S, Adler PM (2013) Pore network modelling
to determine the transport properties in presence of a reactive
fluid: from pore to reservoir scale. Adv Water Resour 53:87–100.
https:// doi. org/ 10. 1016/j. advwa tres. 2012. 10. 004
62. Xiong Q, Baychev TG, Jivkov AP (2016) Review of pore net-
work modelling of porous media: experimental characterisations,
network constructions and applications to reactive transport. J
Contam Hydrol 192:101–117. https:// doi. org/ 10. 1016/j. jconh yd.
2016. 07. 002
63. Song R, Liu J, Cui M (2017) A new method to reconstruct struc-
tured mesh model from micro-computed tomography images
of porous media and its application. Int J Heat Mass Transf
109:705–715. https:// doi. org/ 10. 1016/j. ijhea tmass trans fer. 2017.
02. 053
64. Lu XZ, Chan LC (2018) Micro-voids quantification for damage
prediction in warm forging of biocompatible alloys using 3D
X-ray CT and RVE approach. J Mater Process Technol 258:116–
127. https:// doi. org/ 10. 1016/j. jmatp rotec. 2018. 03. 020
65. Moustafa AR, Dinwiddie RB, Pawlowski AE, Splitter DA, Shyam
A, Cordero ZC (2018) Mesostructure and porosity effects on the
thermal conductivity of additively manufactured interpenetrating
phase composites. Addit Manuf 22:223–229. https:// doi. org/ 10.
1016/j. addma. 2018. 05. 018
66. Lindquist WB, Venkatarangan A, Dunsmuir J, Wong T (2000)
Pore and throat size distributions measured from synchrotron
X-ray tomographic images of Fontainebleau sandstones. J Geo-
phys Res 105:21509–21527. https:// doi. org/ 10. 1029/ 2000J B9002
08
67. Silin D, Patzek T (2006) Pore space morphology analysis using
maximal inscribed spheres. Physica A 371:336–360. https:// doi.
org/ 10. 1016/j. physa. 2006. 04. 048
68. Dong H, Blunt MJ (2009) Pore-network extraction from micro-
computerized-tomography images. Phys Rev E 80:036307.
https:// doi. org/ 10. 1103/ PhysR evE. 80. 036307
69. Klein M, Goetz H, Pazen S, Al-Nawas B, Wagner W, Duschner H
(2009) Pore characteristics of bone substitute materials assessed
by microcomputed tomography. Clin Oral Implant Res 20:67–74.
https:// doi. org/ 10. 1111/j. 1600- 0501. 2008. 01605.x
70. Agaesse T, Lamibrac A, Büchi FN, Pauchet J, Prat M (2016) Val-
idation of pore network simulations of ex-situ water distributions
in a gas diffusion layer of proton exchange membrane fuel cells
with X-ray tomographic images. J Power Sources 331:462–474.
https:// doi. org/ 10. 1016/j. jpows our. 2016. 09. 076
71. Iassonov P, Gebrenegus T, Tuller M (2009) Segmentation of
X-ray computed tomography images of porous materials: a cru-
cial step for characterization and quantitative analysis of pore
structures: X-ray CT image segmentation. Water Resour Res.
https:// doi. org/ 10. 1029/ 2009W R0080 87
72. Houston AN, Otten W, Falconer R, Monga O, Baveye PC, Hapca
SM (2017) Quantification of the pore size distribution of soils:
assessment of existing software using tomographic and synthetic
3D images. Geoderma 299:73–82. https:// doi. org/ 10. 1016/j.
geode rma. 2017. 03. 025
73. Wang J, Guo L, Bai Z, Yang L (2016) Using computed tomog-
raphy (CT) images and multi-fractal theory to quantify the pore
distribution of reconstructed soils during ecological restoration
in opencast coal-mine. Ecol Eng 92:148–157. https:// doi. org/ 10.
1016/j. ecole ng. 2016. 03. 029
74. Shamsi F, Norouzi-Apourvari S, Jafari S (2020) Image-based
simulation of formation damage during suspension injection in
homogeneous and heterogeneous porous media: the impact of
pore-scale characteristics on macroscopic properties. J Petrol Sci
Eng 186:106786. https:// doi. org/ 10. 1016/j. petrol. 2019. 106786
75. Huang X, Wang Q, Zhou W, Deng D, Zhao Y, Wen D, Li J (2015)
Morphology and transport properties of fibrous porous media.
Powder Technol 283:618–626. https:// doi. org/ 10. 1016/j. powtec.
2015. 06. 015
76. Zhang X, Crawford JW, Flavel RJ, Young IM (2016) A multi-
scale Lattice Boltzmann model for simulating solute transport in
3D X-ray micro-tomography images of aggregated porous mate-
rials. J Hydrol 541:1020–1029. https:// doi. org/ 10. 1016/j. jhydr ol.
2016. 08. 013

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
77. Kok MDR, Jervis R, Tranter TG, Sadeghi MA, Brett DJL, Shear-
ing PR, Gostick JT (2019) Mass transfer in fibrous media with
varying anisotropy for flow battery electrodes: direct numerical
simulations with 3D X-ray computed tomography. Chem Eng Sci
196:104–115. https:// doi. org/ 10. 1016/j. ces. 2018. 10. 049
78. Sambaer W, Zatloukal M, Kimmer D (2012) 3D air filtration
modeling for nanofiber based filters in the ultrafine particle size
range. Chem Eng Sci 82:299–311. https:// doi. org/ 10. 1016/j. ces.
2012. 07. 031
79. Videla AR, Lin CL, Miller JD (2008) Simulation of saturated
fluid flow in packed particle beds—the lattice-Boltzmann method
for the calculation of permeability from XMT images. J Chin Inst
Chem Eng, 39:117–128. https:// doi. org/ 10. 1016/j. jcice. 2007. 12.
002
80. Gao Y, Zhang X, Rama P, Chen R, Ostadi H, Jiang K (2013) Lat-
tice Boltzmann simulation of water and gas flow in porous gas
diffusion layers in fuel cells reconstructed from micro-tomogra-
phy. Comput Math Appl 65:891–900. https:// doi. org/ 10. 1016/j.
camwa. 2012. 08. 006
81. Ovaysi S, Piri M (2011) Pore-scale modeling of dispersion in
disordered porous media. J Contam Hydrol 124:68–81. https://
doi. org/ 10. 1016/j. jconh yd. 2011. 02. 004
82. Du W, Iacoviello F, Fernandez T, Loureiro R, Brett DJL, Shear-
ing PR (2021) Microstructure analysis and image-based mod-
elling of face masks for COVID-19 virus protection. Commun
Mater 2:1–10. https:// doi. org/ 10. 1038/ s43246- 021- 00160-z
83. Karakoç A, Miettinen A, Sözümert E, Evans LlM, Yiğitler H,
Bostanci B, Taciroğlu E, Jäntti R (2022) Microstructural evalu-
ation and recommendations for face masks in community use to
reduce the transmission of respiratory infectious diseases. Com-
puter Methods Progr Biomed. https:// doi. org/ 10. 1016/j. cmpb.
2022. 107154
84. Zhang X (n.d.) Modelling water intrusion and oxygen diffusion
in a reconstructed microporous layer of PEM fuel cells. Int J
Hydrogen Energy 9.
85. Lal S, Lucci F, Defraeye T, Poulikakos LD, Partl MN, Derome D,
Carmeliet J (2018) CFD modeling of convective scalar transport
in a macroporous material for drying applications. Int J Therm
Sci 123:86–98. https:// doi. org/ 10. 1016/j. ijthe rmals ci. 2017. 09.
010
86. Cooper SJ, Eastwood DS, Gelb J, Damblanc G, Brett DJL, Brad-
ley RS, Withers PJ, Lee PD, Marquis AJ, Brandon NP, Shearing
PR (2014) Image based modelling of microstructural heterogene-
ity in LiFePO 4 electrodes for Li-ion batteries. J Power Sources
247:1033–1039. https:// doi. org/ 10. 1016/j. jpow s our. 2013. 04. 156
87. Veyhl C, Fiedler T, Andersen O, Meinert J, Bernthaler T, Belova
IV, Murch GE (2012) On the thermal conductivity of sintered
metallic fibre structures. Int J Heat Mass Transf 55:2440–2448.
https:// doi. org/ 10. 1016/j. ijhea tmass trans fer. 2012. 01. 016
88. Borner A, Panerai F, Mansour NN (2017) High temperature
permeability of fibrous materials using direct simulation Monte
Carlo. Int J Heat Mass Transf 106:1318–1326. https:// doi. org/ 10.
1016/j. ijhea tmass trans fer. 2016. 10. 113
89. Panerai F, Ferguson JC, Lachaud J, Martin A, Gasch MJ, Man-
sour NN (2017) Micro-tomography based analysis of thermal
conductivity, diffusivity and oxidation behavior of rigid and
flexible fibrous insulators. Int J Heat Mass Transf 108:801–811.
https:// doi. org/ 10. 1016/j. ijhea tmass trans fer. 2016. 12. 048
90. Akolkar A, Rahmatian N, Unterberger SH, Petrasch J (2017)
Tomography based analysis of conduction anisotropy in fibrous
insulation. Int J Heat Mass Transf 108:1740–1749. https:// doi.
org/ 10. 1016/j. ijhea tmass trans fer. 2016. 12. 083
91. Amani Y, Takahashi A, Chantrenne P, Maruyama S, Dancette
S, Maire E (2018) Thermal conductivity of highly porous metal
foams: experimental and image based finite element analysis.
Int J Heat Mass Transf 122:1–10. https:// doi. org/ 10. 1016/j. ijhea
tmass trans fer. 2018. 01. 050
92. Brown LD, Neville TP, Jervis R, Mason TJ, Shearing PR, Brett
DJL (2016) The effect of felt compression on the performance
and pressure drop of all-vanadium redox flow batteries. Journal
of Energy Storage 8:91–98. https:// doi. org/ 10. 1016/j. est. 2016.
10. 003
93. Ranut P (2015) High resolution X-ray microtomography-based
CFD simulation for the characterization of flow permeability and
effective thermal conductivity of aluminum metal foams. Exp
Thermal Fluid Sci 7
94. Evans LlM, Margetts L, Lee PD, Butler CAM, Surrey E (2019)
Image based in silico characterisation of the effective thermal
properties of a graphite foam. Carbon 143:542–558. https:// doi.
org/ 10. 1016/j. carbon. 2018. 10. 031
95. Cai J, Wei W, Hu X, Wood DA (2017) Electrical conductiv-
ity models in saturated porous media: a review. Earth Sci Rev
171:419–433. https:// doi. org/ 10. 1016/j. earsc irev. 2017. 06. 013
96. Guk E, Venkatesan V, Babar S, Jackson L, Kim JS (2019) Param-
eters and their impacts on the temperature distribution and ther-
mal gradient of solid oxide fuel cell. Appl Energy 11
97. Carraro T, Joos J, Rüger B, Weber A, Ivers-Tiffée E (2012) 3D
finite element model for reconstructed mixed-conducting cath-
odes: I. Performance quantification. Electrochim Acta 77:315–
323. https:// doi. org/ 10. 1016/j. elect acta. 2012. 04. 109
98. Apostolidis P, Liu X, van de Ven M, Erkens S, Scarpas T (2020)
Control the crosslinking of epoxy-asphalt via induction heating.
Int J Pavement Eng 21:956–965. https:// doi. org/ 10. 1080/ 10298
436. 2019. 16527 41
99. Lynch ME, Ding D, Harris WM, Lombardo JJ, Nelson GJ, Chiu
WKS, Liu M (2013) Flexible multiphysics simulation of porous
electrodes: conformal to 3D reconstructed microstructures. Nano
Energy 2:105–115. https:// doi. org/ 10. 1016/j. nanoen. 2012. 08. 002
100. Zhao Y, Stein P, Bai Y, Al-Siraj M, Yang Y, Xu B-X (2019)
A review on modeling of electro-chemo-mechanics in lithium-
ion batteries. J Power Sources 413:259–283. https:// doi. org/ 10.
1016/j. jpows our. 2018. 12. 011
101. Hein S (2016) Stochastic microstructure modeling and electro-
chemical simulation of lithium-ion cell anodes in 3D. J Power
Sources 11.
102. He S, Habte BT, Jiang F (2016) LBM prediction of effective
electric and species transport properties of lithium-ion battery
graphite anode. Solid State Ionics 296:146–153. https:// doi. org/
10. 1016/j. ssi. 2016. 09. 021
103. Lim C, Yan B, Yin L, Zhu L (2012) Simulation of diffusion-
induced stress using reconstructed electrodes particle structures
generated by micro/nano-CT. Electrochim Acta 75:279–287.
https:// doi. org/ 10. 1016/j. elect acta. 2012. 04. 120
104. Kanno D, Shikazono N, Takagi N, Matsuzaki K, Kasagi N (2011)
Evaluation of SOFC anode polarization simulation using three-
dimensional microstructures reconstructed by FIB tomography.
Electrochim Acta 56:4015–4021. https:// doi. org/ 10. 1016/j. elect
acta. 2011. 02. 010
105. Malavé V, Berger JR, Zhu H, Kee RJ (2014) A computational
model of the mechanical behavior within reconstructed LixCoO2
Li-ion battery cathode particles. Electrochim Acta 130:707–717.
https:// doi. org/ 10. 1016/j. elect acta. 2014. 03. 113
106. Mendoza H, Roberts SA, Brunini VE, Grillet AM (2016)
Mechanical and electrochemical response of a LiCoO2 cathode
using reconstructed microstructures. Electrochim Acta 190:1–15.
https:// doi. org/ 10. 1016/j. elect acta. 2015. 12. 224
107. Yan B, Lim C, Yin L, Zhu L (2012) Three dimensional simu-
lation of galvanostatic discharge of LiCoO2 cathode based on
X-ray nano-CT images. J Electrochem Soc 12.
108. Higa K, Wu S-L, Parkinson DY, Fu Y, Ferreira S, Battaglia V,
Srinivasan V (2017) Comparing macroscale and microscale

Ll.M.Evans et al.
1 3
simulations of porous battery electrodes. J Electrochem Soc
164:E3473–E3488. https:// doi. org/ 10. 1149/2. 05017 11jes
109. Trembacki BL, Noble DR, Brunini VE, Ferraro ME, Roberts SA
(2017) Mesoscale effective property simulations incorporating
conductive binder. J Electrochem Soc 15.
110. Trembacki BL, Mistry AN, Noble DR, Ferraro ME, Mukherjee
PP, Roberts SA (2018) Mesoscale analysis of conductive binder
domain morphology in lithium-ion battery electrodes. J Electro-
chem Soc 165:E725–E736. https:// doi. org/ 10. 1149/2. 09818 13jes
111. Ferraro ME, Trembacki BL, Brunini VE, Noble DR, Roberts SA
(2020) Electrode mesoscale as a collection of particles: coupled
electrochemical and mechanical analysis of NMC cathodes. J
Electrochem Soc 167:013543. https:// doi. org/ 10. 1149/ 1945-
7111/ ab632b
112. Wu L, Xiao X, Wen Y, Zhang J (2016) Three-dimensional finite
element study on stress generation in synchrotron X-ray tomog-
raphy reconstructed nickel-manganese-cobalt based half cell.
J Power Sources 336:8–18. https:// doi. org/ 10. 1016/j. jpows our.
2016. 10. 052
113. Gelb J (2017) Multi-scale 3D investigations of a commercial
18650 Li-ion battery with correlative electron- and X-ray micros-
copy. J Power Sources 10.
114. Bird MB, Butler SL, Hawkes CD, Kotzer T (2014) Numerical
modeling of fluid and electrical currents through geometries
based on synchrotron X-ray tomographic images of reservoir
rocks using Avizo and COMSOL. Comput Geosci 73:6–16.
https:// doi. org/ 10. 1016/j. cageo. 2014. 08. 009
115. Liu X, Sun J, Wang H (2009) Numerical simulation of rock elec-
trical properties based on digital cores. Appl Geophys 6:1–7.
https:// doi. org/ 10. 1007/ s11770- 009- 0001-6
116. Noiriel C, Steefel CI, Yang L, Bernard D (2016) Effects of pore-
scale precipitation on permeability and flow. Adv Water Resour
95:125–137. https:// doi. org/ 10. 1016/j. advwa tres. 2015. 11. 013
117. Ferguson JC, Panerai F, Borner A, Mansour NN (2018) PuMA:
the porous microstructure analysis software. SoftwareX 7:81–87.
https:// doi. org/ 10. 1016/j. softx. 2018. 03. 001
118. Semeraro F, Ferguson JC, Acin M, Panerai F, Mansour NN
(2021) Anisotropic analysis of fibrous and woven materials part
2: computation of effective conductivity. Comput Mater Sci
186:109956. https:// doi. org/ 10. 1016/j. comma tsci. 2020. 109956
119. Ferguson JC, Borner A, Panerai F, Close S, Mansour NN (2022)
Continuum to rarefied diffusive tortuosity factors in porous media
from X-ray microtomography. Comput Mater Sci 203:111030.
https:// doi. org/ 10. 1016/j. comma tsci. 2021. 111030
120. A. du Plessis (2021) Porosity in laser powder bed fusion. In:
Fundamenals of laser powder bed fusion of metals. Elsevier, New
York
121. Siroky G, Kraker E, Kieslinger D, Kozeschnik E, Ecker W (2021)
Simulation and experimental characterization of microporosity
during solidification in Sn-Bi alloys. Mater Des 212:110258.
https:// doi. org/ 10. 1016/j. matdes. 2021. 110258
122. Watanabe I, Watkins JH, Nakajima H, Atsuta M, Okabe T (1997)
Effect of pressure difference on the quality of titanium casting. J
Dent Res 76:773–779. https:// doi. org/ 10. 1177/ 00220 34597 07600
31001
123. Sietmann M, Janßen H, Müller M, Lehnert W (2021) An analysis
of the imperfections and defects inside composite bipolar plates
using X-ray computer tomography and resistivity simulations. Int
J Hydrogen Energy 46:25677–25688. https:// doi. org/ 10. 1016/j.
ijhyd ene. 2021. 05. 078
124. Mostafaei A, Zhao C, He Y, Ghiaasiaan SR, Shi B, Shao S,
Shamsaei N, Wu Z, Kouraytem N, Sun T, Pauza J, Gordon JV,
Webler B, Parab ND, Asherloo M, Guo Q, Chen L, Rollett AD
(2022) Defects and anomalies in powder bed fusion metal addi-
tive manufacturing. Curr Opin Solid State Mater Sci 26:100974.
https:// doi. org/ 10. 1016/j. cossms. 2021. 100974
125. Sanaei N, Fatemi A, Phan N (2019) Defect characteristics and
analysis of their variability in metal L-PBF additive manufac-
turing. Mater Des 182:108091. https:// doi. org/ 10. 1016/j. matdes.
2019. 108091
126. Slotwinski JA, Garboczi EJ, Hebenstreit KM (2014) Porosity
measurements and analysis for metal additive manufacturing
process control. J Res Natl Inst Stan 119:494. https:// doi. org/ 10.
6028/ jres. 119. 019
127. Tammas-Williams S, Zhao H, Léonard F, Derguti F, Todd I,
Prangnell PB (2015) XCT analysis of the influence of melt strat-
egies on defect population in Ti–6Al–4V components manufac-
tured by selective electron beam melting. Mater Charact 102:47–
61. https:// doi. org/ 10. 1016/j. match ar. 2015. 02. 008
128. Kim FH, Moylan SP, Phan TQ, Garboczi EJ (2020) Investigation
of the effect of artificial internal defects on the tensile behavior of
laser powder bed fusion 17–4 stainless steel samples: simultane-
ous tensile testing and X-ray computed tomography. Exp Mech
60:987–1004. https:// doi. org/ 10. 1007/ s11340- 020- 00604-6
129. Lu QY, Wong CH (2017) Applications of non-destructive testing
techniques for post-process control of additively manufactured
parts. Virtual Phys Prototyp 12:301–321. https:// doi. org/ 10. 1080/
17452 759. 2017. 13573 19
130. du Plessis A (2019) Effects of process parameters on porosity in
laser powder bed fusion revealed by X-ray tomography. Addit
Manuf 30:100871. https:// doi. org/ 10. 1016/j. addma. 2019. 100871
131. du Plessis A, le Roux SG (2018) Standardized X-ray tomography
testing of additively manufactured parts: a round robin test. Addit
Manuf 24:125–136. https:// doi. org/ 10. 1016/j. addma. 2018. 09. 014
132. du Plessis A, Yadroitsava I, Yadroitsev I (2020) Effects of
defects on mechanical properties in metal additive manufactur-
ing: a review focusing on X-ray tomography insights. Mater Des.
https:// doi. org/ 10. 1016/j. matdes. 2019. 108385
133. Zerbst U, Madia M, Klinger C, Bettge D, Murakami Y (2019)
Defects as a root cause of fatigue failure of metallic components
I: basic aspects. Eng Fail Anal 97:777–792. https:// doi. org/ 10.
1016/j. engfa ilanal. 2019. 01. 055
134. du Plessis A, Beretta S (2020) Killer notches: the effect of as-
built surface roughness on fatigue failure in AlSi10Mg produced
by laser powder bed fusion. Addit Manuf 35:101424. https:// doi.
org/ 10. 1016/j. addma. 2020. 101424
135. Persenot T, Burr A, Dendievel R, Buffière J-Y, Maire E, Lacham-
bre J, Martin G (2020) Fatigue performances of chemically
etched thin struts built by selective electron beam melting:
experiments and predictions. Materialia 9:100589. https:// doi.
org/ 10. 1016/j. mtla. 2020. 100589
136. Burr A, Persenot T, Doutre P-T, Buffiere J-Y, Lhuissier P, Martin
G, Dendievel R (2020) A numerical framework to predict the
fatigue life of lattice structures built by additive manufacturing.
Int J Fatigue 139:105769. https:// doi. org/ 10. 1016/j. ijfat igue.
2020. 105769
137. Nasab MH, Romano S, Gastaldi D, Beretta S, Vedani M (2020)
Combined effect of surface anomalies and volumetric defects on
fatigue assessment of AlSi7Mg fabricated via laser powder bed
fusion. Addit Manuf 34:100918. https:// doi. org/ 10. 1016/j. addma.
2019. 100918
138. Withers PJ, Bhadeshia HKDH (2001) Residual stress. Part 1—
measurement techniques. Mater Sci Technol 17:355–365. https://
doi. org/ 10. 1179/ 02670 83011 01509 980
139. Withers PJ (2001) Residual stress. Part 2 Nat Origins. 17:11.
140. Withers PJ, Webster PJ (2001) Neutron and synchrotron X-ray
strain scanning 37:15
141. Parry LA, Ashcroft IA, Wildman RD (2019) Geometrical effects
on residual stress in selective laser melting. Addit Manuf 25:166–
175. https:// doi. org/ 10. 1016/j. addma. 2018. 09. 026
142. Zerbst U, Madia M, Klinger C, Bettge D, Murakami Y (2019)
Defects as a root cause of fatigue failure of metallic components.

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
II: non-metallic inclusions. Eng Fail Anal 98(2019):228–239.
https:// doi. org/ 10. 1016/j. engfa ilanal. 2019. 01. 054
143. Zerbst U, Madia M, Klinger C, Bettge D, Murakami Y (2019)
Defects as a root cause of fatigue failure of metallic components.
III: cavities, dents, corrosion pits, scratches. Eng Fail Anal
97:759–776. https:// doi. org/ 10. 1016/j. engfa ilanal. 2019. 01. 034
144. Sanaei N, Fatemi A (2020) Defects in additive manufactured
metals and their effect on fatigue performance: a state-of-the-art
review. Progr Mater Sci. https:// doi. org/ 10. 1016/j. pmats ci. 2020.
100724
145. du Plessis A, Yadroitsava I, le Roux SG, Yadroitsev I, Fieres J,
Reinhart C, Rossouw P (2017) Prediction of mechanical per-
formance of Ti6Al4V cast alloy based on microCT-based load
simulation. J Alloys Compd 724:267–274. https:// doi. org/ 10.
1016/j. jallc om. 2017. 06. 320
146. Wicke M, Luetje M, Bacaicoa I, Brueckner-Foit A (2016)
Characterization of casting pores in Fe-rich Al-Si-Cu alloys by
microtomography and finite element analysis. Procedia Struct
Integr 2:2643–2649. https:// doi. org/ 10. 1016/j. prostr. 2016. 06. 330
147. Serrano-Munoz I, Buffiere J-Y, Verdu C (2018) Casting defects
in structural components: are they all dangerous? A 3D study. Int
J Fatigue 117:471–484. https:// doi. org/ 10. 1016/j. ijfat igue. 2018.
08. 019
148. Liu P, Wang C, Yang S, Jiao Y, Li X, Li H, Zhang Y, Liu Y, Wu
S, Zhao J (2022) Coupled influence of pore defects on the failure
site for high-speed railway gearbox material. Eng Fract Mech
261:108216. https:// doi. org/ 10. 1016/j. engfr acmech. 2021. 108216
149. Fieres J, Schumann P, Reinhart C (2018) Predicting failure in
additively manufactured parts using X-ray computed tomogra-
phy and simulation. Procedia Eng 213:69–78. https:// doi. org/ 10.
1016/j. proeng. 2018. 02. 008
150. Ridgeway CD, Gu C, Ripplinger K, Detwiler D, Ji M, Soghrati S,
Luo AA (2020) Prediction of location specific mechanical prop-
erties of aluminum casting using a new CA-FEA (cellular autom-
aton-finite element analysis) approach. Mater Des 194:108929.
https:// doi. org/ 10. 1016/j. matdes. 2020. 108929
151. Kuchariková L, Tillová E, Samardžiová M, Uhríčik M, Belan J,
Švecová I (2019) Quality assessment of Al castings produced in
sand molds using image and CT analyses. J Materi Eng Perform
28:3966–3973. https:// doi. org/ 10. 1007/ s11665- 019- 04040-z
152. Ziółkowski G, Gruber K, Tokarczyk E, Roszak R, Ziegenhorn M
(2021) X-ray computed tomography for the ex-situ mechanical
testing and simulation of additively manufactured IN718 sam-
ples. Addit Manuf 45:102070. https:// doi. org/ 10. 1016/j. addma.
2021. 102070
153. Hu D, Pan J, Mao J, Hu S, Liu X, Fu Y, Wang R (2021) Mechani-
cal behavior prediction of additively manufactured components
based on defect evolution observation by synchrotron radiation
X-ray tomography. Mater Des 198:109353. https:// doi. org/ 10.
1016/j. matdes. 2020. 109353
154. Salarian M, Asgari H, Vlasea M (2020) Pore space character-
istics and corresponding effect on tensile properties of Inconel
625 fabricated via laser powder bed fusion. Mater Sci Eng A
769:138525. https:// doi. org/ 10. 1016/j. msea. 2019. 138525
155. Carlton HD, Haboub A, Gallegos GF, Parkinson DY, MacDowell
AA (2016) Damage evolution and failure mechanisms in addi-
tively manufactured stainless steel. Mater Sci Eng A 651:406–
414. https:// doi. org/ 10. 1016/j. msea. 2015. 10. 073
156. Maconachie T, Leary M, Lozanovski B, Zhang X, Qian M,
Faruque O, Brandt M (2019) SLM lattice structures: proper-
ties, performance, applications and challenges. Mater Des
183:108137. https:// doi. org/ 10. 1016/j. matdes. 2019. 108137
157. Zadpoor AA (2019) Mechanical performance of additively manu-
factured meta-biomaterials. Acta Biomater 85:41–59. https:// doi.
org/ 10. 1016/j. actbio. 2018. 12. 038
158. Murr LE (2019) Strategies for creating living, additively manu-
factured, open-cellular metal and alloy implants by promoting
osseointegration, osteoinduction and vascularization: an over-
view. J Mater Sci Technol 35:231–241. https:// doi. org/ 10. 1016/j.
jmst. 2018. 09. 003
159. du Plessis A, Yadroitsava I, Yadroitsev I, le Roux S, Blaine D
(2018) Numerical comparison of lattice unit cell designs for
medical implants by additive manufacturing. Virtual Phys Pro-
totyp 13:266–281. https:// doi. org/ 10. 1080/ 17452 759. 2018. 14917
13
160. DebRoy T, Wei HL, Zuback JS, Mukherjee T, Elmer JW,
Milewski JO, Beese AM, Wilson-Heid A, De A, Zhang W (2018)
Additive manufacturing of metallic components—process, struc-
ture and properties. Prog Mater Sci 92:112–224. https:// doi. org/
10. 1016/j. pmats ci. 2017. 10. 001
161. du Plessis A, Broeckhoven C, Yadroitsava I, Yadroitsev I, Hands
CH, Kunju R, Bhate D (2019) Beautiful and functional: a review
of biomimetic design in additive manufacturing. Addit Manuf
27:408–427. https:// doi. org/ 10. 1016/j. addma. 2019. 03. 033
162. du Plessis A, Yadroitsava I, Kouprianoff DP, Yadroitsev I
(2018) Numerical and experimental study of the effect of arti-
ficial porosity in a lattice structure manufactured by laser based
powder.
163. Benedetti M, du Plessis A, Ritchie RO, Dallago M, Razavi SMJ,
Berto F (2021) Architected cellular materials: a review on their
mechanical properties towards fatigue-tolerant design and fab-
rication. Mater Sci Eng R Rep 144:100606. https:// doi. org/ 10.
1016/j. mser. 2021. 100606
164. du Plessis A, Razavi SMJ, Berto F (2020) The effects of micr-
oporosity in struts of gyroid lattice structures produced by laser
powder bed fusion. Mater Des 194:108899. https:// doi. org/ 10.
1016/j. matdes. 2020. 108899
165. Sombatmai A, Uthaisangsuk V, Wongwises S, Promoppatum P
(2021) Multiscale investigation of the influence of geometrical
imperfections, porosity, and size-dependent features on mechani-
cal behavior of additively manufactured Ti-6Al-4V lattice struts.
Mater Des 209:109985. https:// doi. org/ 10. 1016/j. matdes. 2021.
109985
166. Xiao Z, Yang Y, Xiao R, Bai Y, Song C, Wang D (2018) Evalu-
ation of topology-optimized lattice structures manufactured via
selective laser melting. Mater Des 143:27–37. https:// doi. org/ 10.
1016/j. matdes. 2018. 01. 023
167. du Plessis A, Yadroitsava I, Yadroitsev I (2018) Ti6Al4V light-
weight lattice structures manufactured by laser powder bed fusion
for load-bearing applications. Opt Laser Technol 108:521–528.
https:// doi. org/ 10. 1016/j. optla stec. 2018. 07. 050
168. Lozanovski B, Leary M, Tran P, Shidid D, Qian M, Choong P,
Brandt M (2019) Computational modelling of strut defects in
SLM manufactured lattice structures. Mater Des 171:107671.
https:// doi. org/ 10. 1016/j. matdes. 2019. 107671
169. Lozanovski B, Downing D, Tino R, du Plessis A, Tran P, Jake-
man J, Shidid D, Emmelmann C, Qian M, Choong P, Brandt
M, Leary M (2020) Non-destructive simulation of node defects
in additively manufactured lattice structures. Addit Manuf
36:101593. https:// doi. org/ 10. 1016/j. addma. 2020. 101593
170. Dong G, Tang Y, Zhao YF (2017) A survey of modeling of lat-
tice structures fabricated by additive manufacturing. J Mech Des
139:100906. https:// doi. org/ 10. 1115/1. 40373 05
171. Dallago M, Winiarski B, Zanini F, Carmignato S, Benedetti M
(2019) On the effect of geometrical imperfections and defects on
the fatigue strength of cellular lattice structures additively manu-
factured via selective laser melting. Int J Fatigue 124:348–360.
https:// doi. org/ 10. 1016/j. ijfat igue. 2019. 03. 019
172. Novak N, Biasetto L, Rebesan P, Zanini F, Carmignato S,
Krstulović-Opara L, Vesenjak M, Ren Z (2021) Experimental
and computational evaluation of tensile properties of additively

Ll.M.Evans et al.
1 3
manufactured hexa- and tetrachiral auxetic cellular structures.
Addit Manuf 45:102022. https:// doi. org/ 10. 1016/j. addma. 2021.
102022
173. Youssef S, Maire E, Gaertner R (2005) Finite element modelling
of the actual structure of cellular materials determined by X-ray
tomography. Acta Mater 53:719–730. https:// doi. org/ 10. 1016/j.
actam at. 2004. 10. 024
174. Bargmann S, Klusemann B, Markmann J, Schnabel JE, Schneider
K, Soyarslan C, Wilmers J (2018) Generation of 3D representa-
tive volume elements for heterogeneous materials: a review. Prog
Mater Sci 96:322–384. https:// doi. org/ 10. 1016/j. pmats ci. 2018.
02. 003
175. Che H, Chen S, Xie Y, Wang H, Amine K, Liao X-Z, Ma Z-F
(2017) Electrolyte design strategies and research progress for
room-temperature sodium-ion batteries. Energy Environ Sci
10:1075–1101. https:// doi. org/ 10. 1039/ C7EE0 0524E
176. Logan ER, Dahn JR (2020) Electrolyte design for fast-charging
Li-ion batteries. Trends Chem 2:354–366. https:// doi. org/ 10.
1016/j. trechm. 2020. 01. 011
177. Le Houx J, Kramer D (2021) X-ray tomography for lithium ion
battery electrode characterization—a review. Energy Rep 7:9–14.
https:// doi. org/ 10. 1016/j. egyr. 2021. 02. 063
178. Chen Y, Kang Y, Zhao Y, Wang L, Liu J, Li Y, Liang Z, He X, Li
X, Tavajohi N, Li B (2021) A review of lithium-ion battery safety
concerns: the issues, strategies, and testing standards. J Energy
Chem 59:83–99. https:// doi. org/ 10. 1016/j. jechem. 2020. 10. 017
179. Ryan G, McGarry P, Pandit A, Apatsidis D (2009) Analysis of
the mechanical behavior of a titanium scaffold with a repeating
unit-cell substructure. J Biomed Mater Res 90B:894–906. https://
doi. org/ 10. 1002/ jbm.b. 31361
180. Ebner M, Geldmacher F, Marone F, Stampanoni M, Wood V
(2013) X-ray tomography of porous, transition metal oxide based
lithium ion battery electrodes. Adv Energy Mater 3:845–850.
https:// doi. org/ 10. 1002/ aenm. 20120 0932
181. Zenyuk IV (2019) Bridging X-ray computed tomography and
computational modeling for electrochemical energy-conversion
and –storage. Curr Opin Electrochem 13:78–85. https:// doi. org/
10. 1016/j. coelec. 2018. 10. 016
182. Ryan EM, Mukherjee PP (2019) Mesoscale modeling in electro-
chemical devices—a critical perspective. Prog Energy Combust
Sci 71:118–142. https:// doi. org/ 10. 1016/j. pecs. 2018. 11. 002
183. Lu X, Bertei A, Finegan DP, Tan C, Daemi SR, Weaving JS,
O’Regan KB, Heenan TMM, Hinds G, Kendrick E, Brett DJL,
Shearing PR (2020) 3D microstructure design of lithium-ion bat-
tery electrodes assisted by X-ray nano-computed tomography
and modelling. Nat Commun 11:2079. https:// doi. org/ 10. 1038/
s41467- 020- 15811-x
184. Fathiannasab H, Zhu L, Chen Z (2021) Chemo-mechanical mod-
eling of stress evolution in all-solid-state lithium-ion batteries
using synchrotron transmission X-ray microscopy tomography.
J Power Sources 483:229028. https:// doi. org/ 10. 1016/j. jpows our.
2020. 229028
185. Kashkooli AG, Farhad S, Lee DU, Feng K, Litster S, Babu SK,
Zhu L, Chen Z (2016) Multiscale modeling of lithium-ion battery
electrodes based on nano-scale X-ray computed tomography. J
Power Sources 307:496–509. https:// doi. org/ 10. 1016/j. jpows our.
2015. 12. 134
186. Jung SN, Dhadwal MK, Kim YW, Kim JH, Riemenschneider J
(2015) Cross-sectional constants of composite blades using com-
puted tomography technique and finite element analysis. Compos
Struct 129:132–142. https:// doi. org/ 10. 1016/j. comps truct. 2015.
03. 074
187. Shirani M, Härkegård G (2012) Damage tolerant design of cast
components based on defects detected by 3D X-ray computed
tomography. Int J Fatigue 41:188–198. https:// doi. org/ 10. 1016/j.
ijfat igue. 2011. 09. 011
188. Hensley C, Sisco K, Beauchamp S, Godfrey A, Rezayat H,
McFalls T, Galicki D, List F, Carver K, Stover C, Gandy DW,
Babu SS (2021) Qualification pathways for additively manu-
factured components for nuclear applications. J Nucl Mater
548:152846. https:// doi. org/ 10. 1016/j. jnucm at. 2021. 152846
189. Barrett TR, Evans LlM, Fursdon M, Domptail F, McIntosh SC,
Iglesias D, Surrey E (2019) Virtual engineering of a fusion
reactor: application to divertor design, manufacture, and test-
ing. IEEE Trans Plasma Sci 47:889–896. https:// doi. org/ 10. 1109/
TPS. 2018. 28568 88
190. Yavari R, Riensche A, Tekerek E, Jacquemetton L, Halliday
H, Vandever M, Tenequer A, Perumal V, Kontsos A, Smoqi Z,
Cole K, Rao P (2021) Digitally twinned additive manufactur-
ing: detecting flaws in laser powder bed fusion by combining
thermal simulations with in-situ meltpool sensor data. Mater Des
211:110167. https:// doi. org/ 10. 1016/j. matdes. 2021. 110167
191. Vestrum O, Langseth M, Børvik T (2019) Finite element mod-
eling of porous polymer pipeline coating using X-ray micro com-
puted tomography. Compos B Eng 172:406–415. https:// doi. org/
10. 1016/j. compo sitesb. 2019. 04. 028
192. Chai Y, Wang Y, Yousaf Z, Vo NT, Lowe T, Potluri P, Withers
PJ (2020) Damage evolution in braided composite tubes under
torsion studied by in-situ X-ray computed tomography. Compos
Sci Technol 188:107976. https:// doi. org/ 10. 1016/j. comps citech.
2019. 107976
193. Ali MA, Umer R, Khan KA (2020) A virtual permeability meas-
urement framework for fiber reinforcements using micro CT gen-
erated digital twins. Int J Lightweight Mater Manuf 3:204–216.
https:// doi. org/ 10. 1016/j. ijlmm. 2019. 12. 002
194. Sencu RM, Yang Z, Wang YC, Withers PJ, Rau C, Parson A,
Soutis C (2016) Generation of micro-scale finite element models
from synchrotron X-ray CT images for multidirectional carbon
fibre reinforced composites. Compos A Appl Sci Manuf 91:85–
95. https:// doi. org/ 10. 1016/j. compo sitesa. 2016. 09. 010
195. Senck S, Happl M, Scheerer M, Glinz J, Reiter T, Kastner J (n.d.)
Quantification of surface-near porosity in additively manufac-
tured aluminum brackets using X-ray microcomputed tomog-
raphy. In: AIAA Scitech 2021 Forum. American Institute of
Aeronautics and Astronautics
196. du Plessis A, le Roux SG, Waller J, Sperling P, Achilles N, Beer-
link A, Métayer J-F, Sinico M, Probst G, Dewulf W, Bittner F,
Endres H-J, Willner M, Drégelyi-Kiss Á, Zikmund T, Laznovsky
J, Kaiser J, Pinter P, Dietrich S, Lopez E, Fitzek O, Konrad P
(2019) Laboratory X-ray tomography for metal additive manu-
facturing: round robin test. Addit Manuf 30:100837. https:// doi.
org/ 10. 1016/j. addma. 2019. 100837
197. Blakey-Milner B, Gradl P, Snedden G, Brooks M, Pitot J, Lopez
E, Leary M, Berto F, du Plessis A (2021) Metal additive manu-
facturing in aerospace: a review. Mater Des 209:110008. https://
doi. org/ 10. 1016/j. matdes. 2021. 110008
198. Chapkin WA, Walgren P, Frank GJ, Seifert DR, Hartl DJ, Baur
JW (2020) Design and optimization of high-strain, cylindrical
composite skins for morphing fuselages. Mater Des 187:108395.
https:// doi. org/ 10. 1016/j. matdes. 2019. 108395
199. Rubio JE, Chakravarty UK (2016) An investigation of the aerody-
namic performance of a biomimetic insect-sized wing for micro
air vehicles. In: Advances in aerospace technology, vol 1. Ameri-
can Society of Mechanical Engineers, Phoenix, p. V001T03A015
200. Jiao P, Nicaise SM, Azadi M, Cortes J, Lilley DE, Cha W, Purohit
PK, Bargatin I (2020) Tunable tensile response of honeycomb
plates with nanoscale thickness: testing and modeling. Extreme
Mech Lett 34:100599. https:// doi. org/ 10. 1016/j. eml. 2019. 100599
201. Rathore JS, Saxena P (2021) Non-destructive quality assessment
of bio-engineering parts using industrial micro X-ray computed
tomography: a review. Mater Lett 287:129252. https:// doi. org/
10. 1016/j. matlet. 2020. 129252

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
202. Moratal D (2012) Finite element analysis—from biomedical
applications to industrial developments. Books on Demand.
https:// doi. org/ 10. 5772/ 2552
203. Barkaoui A, Oumghar IA, Kahla RB (2021) Review on the use
of medical imaging in orthopedic biomechanics: finite element
studies. Comput Methods Biomech Biomed Eng 9:535–554.
https:// doi. org/ 10. 1080/ 21681 163. 2021. 18883 17
204. Driscoll M (2019) The impact of the finite element method on
medical device design. J Med Biol Eng 39:171–172. https:// doi.
org/ 10. 1007/ s40846- 018- 0428-4
205. Kelly ES, Worsley PR, Bowen CJ, Cherry LS, Keenan BE,
Edwards CJ, O’Brien N, King L, Dickinson AS (2021) Predict-
ing forefoot-orthosis interactions in rheumatoid arthritis using
computational modelling. Front Bioeng Biotechnol 9:803725.
https:// doi. org/ 10. 3389/ fbioe. 2021. 803725
206. Geng J-P, Tan KBC, Liu G-R (2001) Application of finite ele-
ment analysis in implant dentistry: a review of the literature.
J Prosthet Dent 85:585–598. https:// doi. org/ 10. 1067/ mpr. 2001.
115251
207. Özcan C, Muraille C, Lestriez P, Josset Y (2019) A new approach
in 3D finite element analysis in restorative dentistry. In: Bioden-
tal engineering V. CRC Press, Boca Raton
208. Rodrigues FP, Li J, Silikas N, Ballester RY, Watts DC (2009)
Sequential software processing of micro-XCT dental-images
for 3D-FE analysis. Dent Mater 25:e47–e55. https:// doi. org/ 10.
1016/j. dental. 2009. 02. 007
209. Faizan A, Goel VK, Garfin SR, Bono CM, Serhan H, Biyani A,
Elgafy H, Krishna M, Friesem T (2012) Do design variations in
the artificial disc influence cervical spine biomechanics? A finite
element investigation. Eur Spine J 21:653–662. https:// doi. org/
10. 1007/ s00586- 009- 1211-6
210. Motherway J, Doorly MC, Curtis M, Gilchrist MD (2009) Head
impact biomechanics simulations: a forensic tool for reconstruct-
ing head injury? Leg Med 11:S220–S222. https:// doi. org/ 10.
1016/j. legal med. 2009. 01. 072
211. Yang J, Lei Z (2019) Chapter18—human head modeling and
applications. In: Scataglini S, Paul G (eds) DHM and posturog-
raphy. Academic Press, Cambridge, pp 217–242
212. Mathur VP, Atif M, Duggal I, Tewari N, Duggal R, Chawla A
(2022) Reporting guidelines for in-silico studies using finite
element analysis in medicine (RIFEM). Comput Methods Pro-
grams Biomed 216:106675. https:// doi. org/ 10. 1016/j. cmpb. 2022.
106675
213. Erdemir A, Guess TM, Halloran J, Tadepalli SC, Morrison TM
(2012) Considerations for reporting finite element analysis stud-
ies in biomechanics. J Biomech 45:625–633. https:// doi. org/ 10.
1016/j. jbiom ech. 2011. 11. 038
214. Hu B-W, Lv X, Chen S-F, Shao Z-W (2019) Application of finite
element analysis for investigation of intervertebral disc degenera-
tion: from laboratory to clinic. Curr Med Sci 39:7–15. https:// doi.
org/ 10. 1007/ s11596- 019- 1993-7
215. Zhang Z, Li Y, Liao Z, Liu W (2016) Research progress and
prospect of applications of finite element method in lumbar
spine biomechanics. Sheng Wu Yi Xue Gong Cheng Xue Za Zhi
33:1196–1202
216. Jin B, Hu Y, Han L (2019) Progress in finite element analysis
of meniscus. China J Orthopaed Traumatol. https:// doi. org/ 10.
3969/j. issn. 1003- 0034. 2019. 05. 019
217. Oefner C, Herrmann S, Kebbach M, Lange H-E, Kluess D, Woic-
zinski M (2021) Reporting checklist for verification and vali-
dation of finite element analysis in orthopedic and trauma bio-
mechanics. Med Eng Phys 92:25–32. https:// doi. org/ 10. 1016/j.
meden gphy. 2021. 03. 011
218. de Almeida A, da Silva Bártolo PJ (2010) Virtual topological
optimisation of scaffolds for rapid prototyping. Med Eng Phys
32:775–782. https:// doi. org/ 10. 1016/j. meden gphy. 2010. 05. 001.
219. Dias MR, Guedes JM, Flanagan CL, Hollister SJ, Fernandes
PR (2014) Optimization of scaffold design for bone tissue engi-
neering: a computational and experimental study. Med Eng Phys
36:448–457. https:// doi. org/ 10. 1016/j. meden gphy. 2014. 02. 010
220. Wieding J, Wolf A, Bader R (2014) Numerical optimization of
open-porous bone scaffold structures to match the elastic prop-
erties of human cortical bone. J Mech Behav Biomed Mater
37:56–68. https:// doi. org/ 10. 1016/j. jmbbm. 2014. 05. 002
221. Roseti L, Parisi V, Petretta M, Cavallo C, Desando G, Bartolotti
I, Grigolo B (2017) Scaffolds for bone tissue engineering: state
of the art and new perspectives. Mater Sci Eng, C 78:1246–1262.
https:// doi. org/ 10. 1016/j. msec. 2017. 05. 017
222. Luo D, Rong Q, Chen Q (2017) Finite-element design and opti-
mization of a three-dimensional tetrahedral porous titanium scaf-
fold for the reconstruction of mandibular defects. Med Eng Phys
47:176–183. https:// doi. org/ 10. 1016/j. meden gphy. 2017. 06. 015
223. Parthasarathy J, Starly B, Raman S (2011) A design for the addi-
tive manufacture of functionally graded porous structures with
tailored mechanical properties for biomedical applications. J
Manuf Process 13:160–170. https:// doi. org/ 10. 1016/j. jmapro.
2011. 01. 004
224. Ali AA, Cristofolini L, Schileo E, Hu H, Taddei F, Kim RH,
Rullkoetter PJ, Laz PJ (2014) Specimen-specific modeling of hip
fracture pattern and repair. J Biomech 47:536–543. https:// doi.
org/ 10. 1016/j. jbiom ech. 2013. 10. 033
225. Bujtár P, Simonovics J, Váradi K, Sándor GKB, Avery CME
(2014) The biomechanical aspects of reconstruction for segmen-
tal defects of the mandible: a finite element study to assess the
optimisation of plate and screw factors. J Cranio-Maxillofac Surg
42:855–862. https:// doi. org/ 10. 1016/j. jcms. 2013. 12. 005
226. Lowther M, Louth S, Davey A, Hussain A, Ginestra P, Carter L,
Eisenstein N, Grover L, Cox S (2019) Clinical, industrial, and
research perspectives on powder bed fusion additively manufac-
tured metal implants. Addit Manuf 28:565–584. https:// doi. org/
10. 1016/j. addma. 2019. 05. 033
227. Li J, Cui X, Hooper GJ, Lim KS, Woodfield TBF (2020) Rational
design, bio-functionalization and biological performance of
hybrid additive manufactured titanium implants for orthopaedic
applications: a review. J Mech Behav Biomed Mater 105:103671.
https:// doi. org/ 10. 1016/j. jmbbm. 2020. 103671
228. Wang C, Li Y, Zhao W, Zou S, Zhou G, Wang Y (2018) Structure
design and multi-objective optimization of a novel crash box
based on biomimetic structure. Int J Mech Sci 138–139:489–501.
https:// doi. org/ 10. 1016/j. ijmec sci. 2018. 01. 032
229. Vilardell AM, Takezawa A, du Plessis A, Takata N, Krakhma-
lev P, Kobashi M, Yadroitsava I, Yadroitsev I (2019) Topology
optimization and characterization of Ti6Al4V ELI cellular lattice
structures by laser powder bed fusion for biomedical applica-
tions. Mater Sci Eng, A 766:138330. https:// doi. org/ 10. 1016/j.
msea. 2019. 138330
230. Yang J, Gu D, Lin K, Yang Y, Ma C (2019) Optimization of
bio-inspired bi-directionally corrugated panel impact-resistance
structures: numerical simulation and selective laser melting pro-
cess. J Mech Behav Biomed Mater 91:59–67. https:// doi. org/ 10.
1016/j. jmbbm. 2018. 11. 026
231. Cooper RJ, Wilcox RK, Jones AC (2019) Finite element models
of the tibiofemoral joint: a review of validation approaches and
modelling challenges. Med Eng Phys 74:1–12. https:// doi. org/
10. 1016/j. meden gphy. 2019. 08. 002
232. Müller P, Cantatore A, Andreasen JL, Hiller J, De Chiffre L
(2013) Computed tomography as a tool for tolerance verifica-
tion of industrial parts. Procedia CIRP 10:125–132. https:// doi.
org/ 10. 1016/j. procir. 2013. 08. 022
233. Orassi V, Duda GN, Heiland M, Fischer H, Rendenbach C, Checa
S (2021) Biomechanical assessment of the validity of sheep as a
preclinical model for testing mandibular fracture fixation devices.

Ll.M.Evans et al.
1 3
Front Bioengand Biotechnol. https:// doi. org/ 10. 3389/ fbioe. 2021.
672176
234. Dutta A, Mukherjee K, Dhara S, Gupta S (2019) Design of
porous titanium scaffold for complete mandibular reconstruc-
tion: the influence of pore architecture parameters. Comput Biol
Med 108:31–41. https:// doi. org/ 10. 1016/j. compb iomed. 2019. 03.
004
235. Koren Y, Shpitalni M, Gu P, Hu SJ (2015) Product design for
mass-individualization. Procedia CIRP 36:64–71. https:// doi. org/
10. 1016/j. procir. 2015. 03. 050
236. Lampel J, Mintzberg H (1996) Customizing customization. Sloan
Manag Rev 38:21–30
237. Saniuk S, Grabowska S, Gajdzik B (2020) Personalization of
products in the industry 40 concept and its impact on achieving
a higher level of sustainable consumption. Energies. https:// doi.
org/ 10. 3390/ en132 25895
238. Zhou F, Ji Y, Jiao RJ (2013) Affective and cognitive design
for mass personalization: status and prospect. J Intell Manuf
24:1047–1069. https:// doi. org/ 10. 1007/ s10845- 012- 0673-2
239. Duray R, Ward PT, Milligan GW, Berry WL (2000) Approaches
to mass customization: configurations and empirical validation.
J Oper Manag 18:605–625. https:// doi. org/ 10. 1016/ S0272-
6963(00) 00043-7
240. Alford D, Sackett P, Nelder G (2000) Mass customization—an
automotive perspective. Int J Prod Econ 65:99–110. https:// doi.
org/ 10. 1016/ S0925- 5273(99) 00093-6
241. Parthasarathy J (2014) 3D modeling, custom implants and its
future perspectives in craniofacial surgery. Ann Maxillofac Sur-
gery 4:9–18. https:// doi. org/ 10. 4103/ 2231- 0746. 133065
242. Yu N, Nguyen T, Cho YD, Kavanagh NM, Ghassib I, Giannobile
WV (2019) Personalized scaffolding technologies for alveolar
bone regenerative medicine. Orthodont Craniofac Res 22(Suppl
1):69–75. https:// doi. org/ 10. 1111/ ocr. 12275
243. Muñoz-Cruzado VD, Castro FJC, Eguía AP, Aguilar LT,
González JT, Puyana JC, Ciuró FP, Padillo-Ruiz J (2020) Using
a bio-scanner and 3D printing to create an innovative custom
made approach for the management of complex entero-atmos-
pheric fistulas. Sci Rep 10:19862. https:// doi. org/ 10. 1038/
s41598- 020- 74213-7
244. Li S, Waheed U, Bahshwan M, Wang LZ, Kalossaka LM, Choi J,
Kundrak F, Lattas A, Ploumpis S, Zafeiriou S, Myant CW (2021)
A scalable mass customisation design process for 3D-printed res-
pirator mask to combat COVID-19. Rapid Prototyp J 27:1302–
1317. https:// doi. org/ 10. 1108/ RPJ- 10- 2020- 0231
245. Harman D, Young PG, Cottona R, Gerber B, Quan C etal (2016)
Reducing the risk of concussion through personalized helmet
design reducing the risk of concussion through personalized hel-
met design
246. du Plessis A, Yadroitsev I, Yadroitsava I, Roux SGL (2018) X-ray
microcomputed tomography in additive manufacturing: a review
of the current technology and applications. 3D Print Addit Manuf
5:227–247. https:// doi. org/ 10. 1089/ 3dp. 2018. 0060.
247. Ali SH (2017) 3D-design of artificial mandible using CBCT,
CAD and FEA. Curr Trend Biomed Eng Biosci 8:28–35
248. Parry EJ, Best JM, Banks CE (2020) Three-dimensional (3D)
scanning and additive manufacturing (AM) allows the fabrication
of customised crutch grips. Mater Today Commun 25:101225.
https:// doi. org/ 10. 1016/j. mtcomm. 2020. 101225
249. Colpani A, Fiorentino A, Ceretti E (2020) Design and fabrica-
tion of customized tracheal stents by additive manufacturing.
Procedia Manuf 47:1029–1035. https:// doi. org/ 10. 1016/j. promfg.
2020. 04. 318
250. Sharma N, Ostas D, Rotar H, Brantner P, Thieringer FM (2021)
Design and additive manufacturing of a biomimetic custom-
ized cranial implant based on Voronoi diagram. Front Physiol
12:647923. https:// doi. org/ 10. 3389/ fphys. 2021. 647923
251. Bahraminasab M (2020) Challenges on optimization of
3D-printed bone scaffolds. Biomed Eng Online 19:69. https://
doi. org/ 10. 1186/ s12938- 020- 00810-2
252. Salmi M (2021) Additive manufacturing processes in medical
applications. Materials (Basel) 14:191. https:// doi. org/ 10. 3390/
ma140 10191
253. Obaton A-F, Fain J, Djemaï M, Meinel D, Léonard F, Mahé E,
Lécuelle B, Fouchet J-J, Bruno G (2017) Invivo XCT bone char-
acterization of lattice structured implants fabricated by additive
manufacturing. Heliyon 3:e00374. https:// doi. org/ 10. 1016/j. heliy
on. 2017. e00374
254. Allen C, Meyer CA, Yoo E, Vargas JA, Liu Y, Jalali P (2018)
Stress distribution in a tooth treated through minimally invasive
access compared to one treated through traditional access: a finite
element analysis study. J Conserv Dent 21:505–509. https:// doi.
org/ 10. 4103/ JCD. JCD_ 260_ 18
255. Cresswell-Boyes AJ, Barber AH, Mills D, Tatla A, Davis GR
(2018) Approaches to 3D printing teeth from X-ray microtomog-
raphy. J Microscopy 272:207–212. https:// doi. org/ 10. 1111/ jmi.
12725
256. Thompson A, McNally D, Maskery I, Leach RK (2017) X-ray
computed tomography and additive manufacturing in medicine:
a review. Int J Metrol Qual Eng 8.
257. Loubele M, Bogaerts R, Van Dijck E, Pauwels R, Vanheusden S,
Suetens P, Marchal G, Sanderink G, Jacobs R (2009) Comparison
between effective radiation dose of CBCT and MSCT scanners
for dentomaxillofacial applications. Eur J Radiol 71:461–468.
https:// doi. org/ 10. 1016/j. ejrad. 2008. 06. 002
258. Ay M, Kubat T, Delilbbasi C, Ekici B, Yuzbasioglu HE, Har-
tomad1olu S (2013) 3D Bio-Cad modeling of human mandible
and fabrication by rapid-prototyping technology. Usak University
Journal of Material Sciences
259. Modenese L, Barzan M, Carty CP (2021) Dependency of lower
limb joint reaction forces on femoral version. Gait Posture
88:318–321. https:// doi. org/ 10. 1016/j. gaitp ost. 2021. 06. 014
260. Yu J-H, Wang Y-T, Lin C-L (2018) Customized surgical tem-
plate fabrication under biomechanical consideration by integrat-
ing CBCT image, CAD system and finite element analysis. Dent
Mater J 37:6–14. https:// doi. org/ 10. 4012/ dmj. 2016- 312
261. Dot G, Licha R, Goussard F, Sansalone V (2021) A new proto-
col to accurately track long–term orthodontic tooth movement
and support patient-specific numerical modeling. J Biomech
129:110760. https:// doi. org/ 10. 1016/j. jbiom ech. 2021. 110760
262. Cucchi A, Giavatto MA, Giannatiempo J, Lizio G, Corinaldesi
G (2019) Custom-made titanium mesh for maxillary bone aug-
mentation with immediate implants and delayed loading. J Oral
Implantol 45:59–64. https:// doi. org/ 10. 1563/ aaid- joi-D- 18- 00141
263. Honigmann P, Sharma N, Okolo B, Popp U, Msallem B, Thier-
inger FM (2018) Patient-specific surgical implants made of 3D
printed PEEK: material, technology, and scope of surgical appli-
cation. Biomed Res Int 2018:4520636. https:// doi. org/ 10. 1155/
2018/ 45206 36
264. Palines PA, Ferrer JR, Yoo A, St Hilaire H, Stalder MW (2021)
Simplifying bony midface reconstruction with patient-specific
titanium plates plastic and reconstructive surgery. Global Open
9:e3555. https:// doi. org/ 10. 1097/ GOX. 00000 00000 003555
265. Wu T, Liao W, Dai N, Tang C (2010) Design of a custom angled
abutment for dental implants using computer-aided design and
nonlinear finite element analysis. J Biomech 43:1941–1946.
https:// doi. org/ 10. 1016/j. jbiom ech. 2010. 03. 017
266. Ciocca L, Lesci IG, Mezini O, Parrilli A, Ragazzini S, Rinnovati
R, Romagnoli N, Roveri N, Scotti R (2017) Customized hybrid
biomimetic hydroxyapatite scaffold for bone tissue regeneration.
J Biomed Mater Res B 105:723–734. https:// doi. org/ 10. 1002/
jbm.b. 33597

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
267. Bedini R, Pecci R, Meleo D, Campioni I (2020) Bone substitutes
scaffold in human bone: comparative evaluation by 3D micro-CT
technique. Appl Sci. https:// doi. org/ 10. 3390/ app10 103451
268. Uth N, Mueller J, Smucker B, Yousefi AM (2017) Validation of
scaffold design optimization in bone tissue engineering: finite
element modeling versus designed experiments. Biofabrication.
https:// doi. org/ 10. 1088/ 1758- 5090/9/ 1/ 015023
269. Soufivand AA, Abolfathi N, Hashemi A, Lee SJ (2020) The effect
of 3D printing on the morphological and mechanical properties
of polycaprolactone filament and scaffold. Polym Adv Technol
31:1038–1046. https:// doi. org/ 10. 1002/ pat. 4838
270. Pandithevan P, Kumar GS (2010) Finite element analysis of a
personalized femoral scaffold with designed microarchitecture.
Proc Inst Mech Eng [H] 224:877–889. https:// doi. org/ 10. 1243/
09544 119JE IM633
271. Podshivalov L, Fischer A, Bar-Yoseph PZ (2014) On the road to
personalized medicine: multiscale computational modeling of
bone tissue. Archiv Comput Methods Eng 21:399–479. https://
doi. org/ 10. 1007/ s11831- 014- 9120-1
272. FDA (2013) Paving the way for personalized medicine paving
the way for personalized medicine—FDA’s role in a new era of
medical product development
273. Alomari M, Mohamed FH, Basit AW, Gaisford S (2015) Person-
alised dosing: printing a dose of one’s own medicine. Int J Pharm
494:568–577. https:// doi. org/ 10. 1016/j. ijpha rm. 2014. 12. 006
274. Florence AT, Lee VHL (2011) Personalised medicines: more
tailored drugs, more tailored delivery. Int J Pharm 415:29–33.
https:// doi. org/ 10. 1016/j. ijpha rm. 2011. 04. 047
275. Warsi MH, Yusuf M, Al Robaian M, Khan M, Khan S (2018) 3D
printing methods for pharmaceutical manufacturing: opportunity
and challenges. Curr Pharm Design 24:4949–4956. https:// doi.
org/ 10. 2174/ 13816 12825 66618 12061 21701
276. Albuquerque ASA, Thalhofer JL, Silva AX, Lopes AM, Reis
Junior JP (2019) Dose distribution in healthy tissues by com-
puter modeling of the Alderson Rando phantom in Monte Carlo
method. In: Sociedad Mexicana de Irradiacion y Dosimetria,
Mexico
277. Lê M, Delingette H, Kalpathy-Cramer J, Gerstner ER, Batch-
elor T, Unkelbach J, Ayache N (2017) Personalized radiotherapy
planning based on a computational tumor growth model. IEEE
Trans Med Imaging 36:815–825. https:// doi. org/ 10. 1109/ TMI.
2016. 26264 43
278. Fuerst B, Mansi T, Khurd P, Zhang J, Declerck J, Boettger T,
Navab N, Bayouth J, Kamen A (2012) Patient-specific finite-
element simulation of respiratory mechanics for radiotherapy
guidance, a first evaluation study. In: 2012 9th IEEE interna-
tional symposium on biomedical imaging (ISBI), pp. 1212–1215.
https:// doi. org/ 10. 1109/ ISBI. 2012. 62357 79.
279. Boman E, Tervo J, Vauhkonen M (2004) Modelling the transport
of ionizing radiation using the finite element method. Phys Med
Biol 50:265–280. https:// doi. org/ 10. 1088/ 0031- 9155/ 50/2/ 006
280. Mohan R (1989) Dose computations for three-dimensional radia-
tion treatment planning. Australas Phys Eng Sci Med 12:241–251
281. Gifford KA, Horton JL, Wareing TA, Failla G, Mourtada F
(2006) Comparison of a finite-element multigroup discrete-ordi-
nates code with Monte Carlo for radiotherapy calculations. Phys
Med Biol 51:2253–2265. https:// doi. org/ 10. 1088/ 0031- 9155/
51/9/ 010
282. Roncali E, Taebi A, Foster C, Vu CT (2020) Personalized dosim-
etry for liver cancer y-90 radioembolization using computational
fluid dynamics and Monte Carlo simulation. Ann Biomed Eng
48:1499–1510. https:// doi. org/ 10. 1007/ s10439- 020- 02469-1
283. Mekdash H, Shahine B, Jalbout W, Chehab C, Khalek HA,
Youssef B (2017) A simple technique for an accurate shielding
of the lungs during total body irradiation. TPMNRO 3:13–18.
https:// doi. org/ 10. 1016/j. tipsro. 2017. 07. 001
284. Lu Y, Song J, Yao X, An M, Shi Q, Huang X (2021) 3D printing
polymer-based bolus used for radiotherapy. Int J Bioprint 7:414
285. Gomez G, Baeza M, Mateos JC, Rivas JA, Simon FJL, Ortega
DM, de los Ángeles Flores Carrión M, del Campo ER, Gómez-
Cía T, Guerra JLL (2021) A three-dimensional printed custom-
ized bolus: adapting to the shape of the outer ear. Rep Pract
Oncol Radiother 26:211–217. https:// doi. org/ 10. 5603/ RPOR.
a2021. 0030
286. Zhang L, Goldman S, Chen J, Rowe X, Lin J, Zhou S (2020)
Image-based characterization—non-invasive, quantitative char-
acterization of lyophilized drug product using three-dimensional
X-ray microscopy analyticsimage-based characterization—
non-invasive, quantitative characterization of lyophilized drug
product, drug development and delivery. https:// drug- dev. com/
image- based- chara cteri zation- non- invas ive- quant itati ve- chara
cteri zation- of- lyoph ilized- drug- produ ct- using- three- dimen
sional- x- ray- micro scopy- analy tics/.
287. Boulos MNK, Zhang P (2021) Digital twins: from personalised
medicine to precision public health. J Personal Med. https:// doi.
org/ 10. 3390/ jpm11 080745
288. Cho S-W, Byun S-H, Yi S, Jang W-S, Kim J-C, Park I-Y, Yang
B-E (2021) Sagittal relationship between the maxillary cen-
tral incisors and the forehead in digital twins of Korean adult
females. J Personal Med. https:// doi. org/ 10. 3390/ jpm11 030203
289. Wang X, Wang Y, Tao F, Liu A (2021) New paradigm of data-
driven smart customisation through digital twin. J Manuf Syst
58:270–280. https:// doi. org/ 10. 1016/j. jmsy. 2020. 07. 023
290. Zhang M, Mak AF, Roberts VC (1998) Finite element model-
ling of a residual lower-limb in a prosthetic socket: a survey of
the development in the first decade. Med Eng Phys 20:360–373.
https:// doi. org/ 10. 1016/ s1350- 4533(98) 00027-7
291. Pallari JHP, Dalgarno KW, Woodburn J (2010) Mass customiza-
tion of foot orthoses for rheumatoid arthritis using selective laser
sintering. IEEE Trans Biomed Eng 57:1750–1756. https:// doi.
org/ 10. 1109/ TBME. 2010. 20441 78
292. Jin Y, Plott J, Chen R, Wensman J, Shih A (2015) Additive manu-
facturing of custom orthoses and prostheses—a review. Procedia
CIRP 36:199–204. https:// doi. org/ 10. 1016/j. procir. 2015. 02. 125
293. Chen RK, Jin Y, Wensman J, Shih A (2016) Additive manu-
facturing of custom orthoses and prostheses—a review. Addit
Manuf 12:77–89. https:// doi. org/ 10. 1016/j. addma. 2016. 04. 002
294. Mali HS, Vasistha S (2020) Fabrication of customized ankle foot
orthosis (AFO) by reverse engineering using fused deposition
modelling. In: Shunmugam MS, Kanthababu M (eds) Advances
in additive manufacturing and joining. Springer, Singapore, pp
3–15
295. Agudelo-Ardila CP, Prada-Botía GC, Rodrigues PHG (2019)
Orthotic prototype for upper limb printed in 3D: a efficient solu-
tion. J Phys. https:// doi. org/ 10. 1088/ 1742- 6596/ 1388/1/ 012016
296. Alqahtani MS, Al-Tamimi A, Almeida H, Cooper G, Bartolo P
(2020) A review on the use of additive manufacturing to produce
lower limb orthoses. Progr Addit Manuf 5:85–94. https:// doi. org/
10. 1007/ s40964- 019- 00104-7
297. Shih A, Park DW, Yang Y-Y, Chisena R, Wu D (2017) Cloud-
based design and additive manufacturing of custom orthoses.
Procedia CIRP 63:156–160. https:// doi. org/ 10. 1016/j. procir.
2017. 03. 355
298. Sharma N, Welker D, Cao S, von Netzer B, Honigmann P, Thier-
inger F (2021) An interactive, fully digital design workflow for
a custom 3d printed facial protection orthosis (Face Mask)
BT—industrializing additive manufacturing. Springer, Cham,
pp 26–36
299. Hale L, Linley E, Kalaskar DM (2020) A digital workflow for
design and fabrication of bespoke orthoses using 3D scanning
and 3D printing, a patient-based case study. Sci Rep 10:7028.
https:// doi. org/ 10. 1038/ s41598- 020- 63937-1

Ll.M.Evans et al.
1 3
300. Decker W, Baker A, Ye X, Brown P, Stitzel J, Gayzik FS (2020)
Development and multi-scale validation of a finite element foot-
ball helmet model. Ann Biomed Eng 48:258–270. https:// doi. org/
10. 1007/ s10439- 019- 02345-7
301. Corrales MA, Gierczycka D, Barker J, Bruneau D, Bustamante
MC, Cronin DS (2020) Validation of a football helmet finite
element model and quantification of impact energy distribu-
tion. Ann Biomed Eng 48:121–132. https:// doi. org/ 10. 1007/
s10439- 019- 02359-1
302. Ellena T, Mustafa H, Subic A, Pang TY (2018) A design frame-
work for the mass customisation of custom-fit bicycle helmet
models. Int J Ind Ergon 64:122–133. https:// doi. org/ 10. 1016/j.
ergon. 2018. 01. 005
303. Xiong S, Zhao J, Jiang Z, Dong M (2010) A computer-aided
design system for foot-feature-based shoe last customization.
Int J Adv Manuf Technol 46:11–19. https:// doi. org/ 10. 1007/
s00170- 009- 2087-7
304. Duarte-Pereira DMV, Del Rey-Santamaria M, Javierre-Garcés
C, Barbany-Cairó J, Paredes-Garcia J, Valmaseda-Castellón E,
Berini-Aytés L, Gay-Escoda C (2008) Wearability and physi-
ological effects of custom-fitted vs self-adapted mouthguards.
Dent Traumatol 24:439–442. https:// doi. org/ 10. 1111/j. 1600-
9657. 2008. 00595.x
305. Verissimo C, Costa PVM, Santos-Filho PCF, Tantbirojn D, Ver-
sluis A, Soares CJ (2016) Custom-Fitted EVA Mouthguards:
what is the ideal thickness? A dynamic finite element impact
study. Dent Traumatol 32:95–102. https:// doi. org/ 10. 1111/ edt.
12210
306. D. Systèmes, CES 2020 Survey by CITE Research/Dassault Sys-
tèmes : consumers want personalized products but won’t wait for
them and expect a cost benefit for their data, 3–5 (2020). https://
www. 3ds. com/ newsr oom/ press- relea ses/ ces- 2020- survey- cite-
resea rch- dassa ult- syste mes- consu mers- want- pers o naliz ed- p r odu
cts- wont- wait- them- and- expect- cost- benefi t- their- data.
307. Pickering C (2018) Mini yours customised: the british car gets
personal. www. theen gineer. co. uk/ mini- custo mised- briti sh- car/.
308. Gayzik FS, Moreno DP, Danelson KA, McNally C, Klinich
KD, Stitzel JD (2012) External landmark, body surface, and
volume data of a mid-sized male in seated and standing pos-
tures. Ann Biomed Eng 40:2019–2032. https:// doi. org/ 10. 1007/
s10439- 012- 0546-z
309. Gupta R, Kelkar AD (2005) Nonlinear crash dynamics simulation
of novel airbag based next generation energy absorbing barrier.
310. Glodová I, Lipták T, Bocko J (2014) Usage of finite element
method for motion and thermal analysis of a specific object in
solidworks environment. Procedia Eng 96:131–135. https:// doi.
org/ 10. 1016/j. proeng. 2014. 12. 131
311. Grunenfelder LK, Suksangpanya N, Salinas C, Milliron G,
Yaraghi N, Herrera S, Evans-Lutterodt K, Nutt SR, Zavattieri
P, Kisailus D (2014) Bio-inspired impact-resistant composites.
Acta Biomater 10:3997–4008. https:// doi. org/ 10. 1016/j. actbio.
2014. 03. 022
312. Tran P, Ngo TD, Mendis P (2014) Bio-inspired composite struc-
tures subjected to underwater impulsive loading. Comput Mater
Sci 82:134–139. https:// doi. org/ 10. 1016/j. comma tsci. 2013. 09.
033
313. Connors M, Yang T, Hosny A, Deng Z, Yazdandoost F, Mas-
saadi H, Eernisse D, Mirzaeifar R, Dean MN, Weaver JC, Ortiz
C, Li L (2019) Bioinspired design of flexible armor based on
chiton scales. Nat Commun 10:5413. https:// doi. org/ 10. 1038/
s41467- 019- 13215-0
314. Signetti S, Pugno NM (2018) Modeling and simulation of bio-
inspired nanoarmors. In: Schmauder S, Chen C-S, Chawla KK,
Chawla N, Chen W, Kagawa Y (eds) Handbook of mechanics of
materials. Springer, Singapore, pp 1–29
315. Achrai B, Wagner HD (2013) Micro-structure and mechanical
properties of the turtle carapace as a biological composite shield.
Acta Biomater 9:5890–5902. https:// doi. org/ 10. 1016/j. actbio.
2012. 12. 023
316. Chen P-Y, McKittrick J, Meyers MA (2012) Biological materials:
functional adaptations and bioinspired designs. Prog Mater Sci
57:1492–1704. https:// doi. org/ 10. 1016/j. pmats ci. 2012. 03. 001
317. Meyers MA, Chen P-Y, Lopez MI, Seki Y, Lin AYM (2011)
Biological materials: a materials science approach. J Mech Behav
Biomed Mater 4:626–657. https:// doi. org/ 10. 1016/j. jmbbm. 2010.
08. 005
318. Koch K, Bhushan B, Barthlott W (2009) Multifunctional surface
structures of plants: an inspiration for biomimetics. Prog Mater
Sci 54:137–178. https:// doi. org/ 10. 1016/j. pmats ci. 2008. 07. 003
319. Launey ME, Munch E, Alsem DH, Saiz E, Tomsia AP, Ritchie
RO (2010) A novel biomimetic approach to the design of high-
performance ceramic–metal composites. J R Soc Interface
7:741–753. https:// doi. org/ 10. 1098/ rsif. 2009. 0331
320. du Plessis A, Broeckhoven C (2022) Chapter10—functional
synergy of biomimicry and additive manufacturing: toward a
bio-enhanced engineering approach. In: Eggermont M, Shyam
V, Hepp AF (eds) Biomimicry for materials, design and habitats.
Elsevier, pp 269–289
321. Ha NS, Lu G (2020) A review of recent research on bio-inspired
structures and materials for energy absorption applications. Com-
pos B Eng 181:107496. https:// doi. org/ 10. 1016/j. compo sitesb.
2019. 107496
322. Vásárhelyi L, Kónya Z, Kukovecz Á, Vajtai R (2020) Microcom-
puted tomography–based characterization of advanced materials:
a review. Mater Today Adv 8:100084. https:// doi. org/ 10. 1016/j.
mtadv. 2020. 100084
323. Tao Y, Li Z, Li P (2020) A design and fabrication method for
wood-inspired composites by micro X-ray computed tomogra-
phy and 3D printing. Appl Sci 10:1400. https:// doi. org/ 10. 3390/
app10 041400
324. Da Silva A, Kyriakides S (2007) Compressive response and fail-
ure of balsa wood. Int J Solids Struct 44:8685–8717. https:// doi.
org/ 10. 1016/j. ijsol str. 2007. 07. 003
325. Shishkina O, Lomov SV, Verpoest I, Gorbatikh L (2014) Struc-
ture–property relations for balsa wood as a function of density:
modelling approach. Arch Appl Mech 84:789–805. https:// doi.
org/ 10. 1007/ s00419- 014- 0833-2
326. Karam GN, Gibson LJ (1994) Biomimicking of animal quills and
plant stems: natural cylindrical shells with foam cores. Mater
Sci Eng C 2:113–132. https:// doi. org/ 10. 1016/ 0928- 4931(94)
90039-6
327. Drol CJ, Kennedy EB, Hsiung B-K, Swift NB, Tan K-T (2019)
Bioinspirational understanding of flexural performance in hedge-
hog spines. Acta Biomater 94:553–564. https:// doi. org/ 10. 1016/j.
actbio. 2019. 04. 036
328. Zou M, Xu S, Wei C, Wang H, Liu Z (2016) A bionic method
for the crashworthiness design of thin-walled structures inspired
by bamboo. Thin-Walled Struct 101:222–230. https:// doi. org/ 10.
1016/j. tws. 2015. 12. 023
329. du Plessis A, Broeckhoven C (2019) Looking deep into nature:
a review of micro-computed tomography in biomimicry. Acta
Biomater 85:27–40. https:// doi. org/ 10. 1016/j. actbio. 2018. 12. 014
330. Baino F, Labate GFD, di Confiengo GG, Faga MG, Vitale-Bro-
varone C, Catapano G (2018) Microstructural characterization
and robust comparison of ceramic porous orbital implants. J Eur
Ceram Soc 38:2988–2993. https:// doi. org/ 10. 1016/j. jeurc erams
oc. 2017. 12. 047
331. Labate GFD, Catapano G, Vitale-Brovarone C, Baino F (2017)
Quantifying the micro-architectural similarity of bioceramic
scaffolds to bone. Ceram Int 43:9443–9450. https:// doi. org/ 10.
1016/j. ceram int. 2017. 04. 121

A Review ofImage-Based Simulation Applications inHigh-Value Manufacturing
1 3
332. González SG, Jiménez JFV, Bastida GC, Vlad MD, López JL,
Aguado EF (2020) Synthetic open cell foams versus a healthy
human vertebra: anisotropy, fluid flow and μ-CT structural stud-
ies. Mater Sci Eng C 108:110404. https:// doi. org/ 10. 1016/j. msec.
2019. 110404
333. Leißner T, Diener A, Löwer E, Ditscherlein R, Krüger K, Kwade
A, Peuker UA (2020) 3D ex-situ and in-situ X-ray CT process
studies in particle technology—a perspective. Adv Powder Tech-
nol 31:78–86. https:// doi. org/ 10. 1016/j. apt. 2019. 09. 038
334. Wang B, Pan B, Lubineau G (2018) Morphological evolution
and internal strain mapping of pomelo peel using X-ray com-
puted tomography and digital volume correlation. Mater Des
137:305–315. https:// doi. org/ 10. 1016/j. matdes. 2017. 10. 038
335. Gómez S, Vlad MD, López J, Fernández E (2016) Design and
properties of 3D scaffolds for bone tissue engineering. Acta Bio-
mater 42:341–350. https:// doi. org/ 10. 1016/j. actbio. 2016. 06. 032
336. Luo G, Shi L, Zhong Y, Yang L, Ma X, Li M, He X (2018) Com-
pressive damage of three-dimensional random fibrous ceramic
materials: a meso-mechanics modeling and experimental study.
Ceram Int 44:15690–15699. https:// doi. org/ 10. 1016/j. ceram int.
2018. 05. 241
337. Arabnejad S, Johnston RB, Pura JA, Singh B, Tanzer M, Pasini
D (2016) High-strength porous biomaterials for bone replace-
ment: a strategy to assess the interplay between cell morphology,
mechanical properties, bone ingrowth and manufacturing con-
straints. Acta Biomater 30:345–356. https:// doi. org/ 10. 1016/j.
actbio. 2015. 10. 048
338. Gómez S, Vlad MD, López J, Navarro M, Fernández E (2013)
Characterization and three-dimensional reconstruction of syn-
thetic bone model foams. Mater Sci Eng C 33:3329–3335.
https:// doi. org/ 10. 1016/j. msec. 2013. 04. 013
339. Jaecques SVN, Van Oosterwyck H, Muraru L, Van Cleynen-
breugel T, De Smet E, Wevers M, Naert I, Vander Sloten J (2004)
Individualised, micro CT-based finite element modelling as a tool
for biomechanical analysis related to tissue engineering of bone.
Biomaterials 25:1683–1696. https:// doi. org/ 10. 1016/ S0142-
9612(03) 00516-7
340. Yan C, Hao L, Hussein A, Young P (2015) Ti–6Al–4V triply
periodic minimal surface structures for bone implants fabricated
via selective laser melting. J Mech Behav Biomed Mater 51:61–
73. https:// doi. org/ 10. 1016/j. jmbbm. 2015. 06. 024
341. Giannitelli SM, Accoto D, Trombetta M, Rainer A (2014) Cur-
rent trends in the design of scaffolds for computer-aided tissue
engineering. Acta Biomater 10:580–594. https:// doi. org/ 10.
1016/j. actbio. 2013. 10. 024
342. Borah B, Gross GJ, Dufresne TE, Smith TS, Cockman MD,
Chmielewski PA, Lundy MW, Hartke JR, Sod EW (2001) Three-
dimensional microimaging (MR?I and ?CT), finite element mod-
eling, and rapid prototyping provide unique insights into bone
architecture in osteoporosis. Anat Rec 265:101–110. https:// doi.
org/ 10. 1002/ ar. 1060
343. Rainer A, Giannitelli SM, Accoto D, De Porcellinis S, Gug-
lielmelli E, Trombetta M (2012) Load-adaptive scaffold architec-
turing: a bioinspired approach to the design of porous additively
manufactured scaffolds with optimized mechanical proper-
ties. Ann Biomed Eng 40:966–975. https:// doi. org/ 10. 1007/
s10439- 011- 0465-4
344. Cengiz IF, Oliveira JM, Reis RL (2018) Micro-CT—a digital 3D
microstructural voyage into scaffolds: a systematic review of the
reported methods and results. Biomater Res 22:26. https:// doi.
org/ 10. 1186/ s40824- 018- 0136-8
345. Liu C, Chen S, Sheng C, Ding P, Qian Z, Ren L (2019) The
art of a hydraulic joint in a spider’s leg: modelling, compu-
tational fluid dynamics (CFD) simulation, and bio-inspired
design. J Comp Physiol A 205:491–504. https:// doi. org/ 10. 1007/
s00359- 019- 01336-2
346. Jones AC, Wilcox RK (2008) Finite element analysis of the
spine: towards a framework of verification, validation and sensi-
tivity analysis. Med Eng Phys 30:1287–1304. https:// doi. org/ 10.
1016/j. meden gphy. 2008. 09. 006
347. Petit C, Meille S, Maire E, Gremillard L, Adrien J, Lau GY, Tom-
sia AP (2017) Fracture behavior of robocast HA/β-TCP scaffolds
studied by X-ray tomography and finite element modeling. J Eur
Ceram Soc 37:1735–1745. https:// doi. org/ 10. 1016/j. jeurc erams
oc. 2016. 11. 035
348. Truscello S, Kerckhofs G, Van Bael S, Pyka G, Schrooten J, Van
Oosterwyck H (2012) Prediction of permeability of regular scaf-
folds for skeletal tissue engineering: a combined computational
and experimental study. Acta Biomater 8:1648–1658. https:// doi.
org/ 10. 1016/j. actbio. 2011. 12. 021
349. Liang Z, Ioannidis MA, Chatzis I (2000) Permeability and elec-
trical conductivity of porous media from 3D stochastic replicas
of the microstructure. Chem Eng Sci 55:5247–5262. https:// doi.
org/ 10. 1016/ S0009- 2509(00) 00142-1
350. Pennella F, Cerino G, Massai D, Gallo D, Labate GFD, Schiavi
A, Deriu MA, Audenino A, Morbiducci U (2013) A survey of
methods for the evaluation of tissue engineering scaffold perme-
ability. Ann Biomed Eng 41:2027–2041. https:// doi. org/ 10. 1007/
s10439- 013- 0815-5
351. Sun J, Zhang Y, Lyckegaard A, Bachmann F, Lauridsen EM,
Jensen DJ (2019) Grain boundary wetting correlated to the
grain boundary properties: a laboratory-based multimodal X-ray
tomography investigation. Scr Mater 163:77–81. https:// doi. org/
10. 1016/j. scrip tamat. 2019. 01. 007
352. Evans LlM (2013) Thermal finite element analysis of ceramic/
metal joining for fusion using X-ray tomography data, PhD The-
sis, University of Manchester. https:// www. escho lar. manch ester.
ac. uk/ api/ datas tream? publi catio nPid= uk- ac- man- scw: 21415 9&
datas tream Id= FULL- TEXT. PDF
353. Masuo H, Tanaka Y, Morokoshi S, Yagura H, Uchida T, Yama-
moto Y, Murakami Y (2018) Influence of defects, surface rough-
ness and HIP on the fatigue strength of Ti-6Al-4V manufactured
by additive manufacturing. Int J Fatigue 117:163–179. https://
doi. org/ 10. 1016/j. ijfat igue. 2018. 07. 020
354. du Plessis A, Macdonald E (2020) Hot isostatic pressing in metal
additive manufacturing: X-ray tomography reveals details of
pore closure. Addit Manuf 34:101191. https:// doi. org/ 10. 1016/j.
addma. 2020. 101191
355. Biguri A, Towsyfyan H, Boardman R, Blumensath T (2020)
Numerically robust tetrahedron-based tomographic forward and
backward projectors on parallel architectures. Ultramicroscopy
214:113016. https:// doi. org/ 10. 1016/j. ultra mic. 2020. 113016
356. BS EN 16603-32-03:2014 Space engineering. Structural finite
element models, BSI, 2014. https:// bsol. bsigr oup. com/ Bibli ograp
hic/ Bibli ograp hicIn foData/ 00000 00000 30288 226. Accessed 18
Jan 2022.
357. Therapeutic Goods Administration (2021) Personalised medical
devices (including 3D-printed devices)—regulatory changes for
custom-made medical devices Version 4.0, August
358. FDA (2014) Custom device exemption
359. MDCG (2021) MDCG 2021–3 questions and answers on custom-
made devices, guidance on classification of medical devices. 1–8.
https:// ec. europa. eu/ health/ md_ sector/ new_ regul ations/ guida
nce_ en.
360. Eisenhart S (2021) European regulators publish new Q&A on
custom-made medical devices
361. Pettitt D, Smith J, Meadows N, Arshad Z, Schuh A, DiGiusto D,
Bountra C, Holländer G, Barker R, Brindley D (2016) Regula-
tory barriers to the advancement of precision medicine. Expert
Rev Precis Med Drug Dev 1:319–329. https:// doi. org/ 10. 1080/
23808 993. 2016. 11765 26

Ll.M.Evans et al.
1 3
362. Cherukara MJ, Zhou T, Nashed Y, Enfedaque P, Hexemer A,
Harder RJ, Holt MV (2020) AI-enabled high-resolution scanning
coherent diffraction imaging. Appl Phys Lett 117:044103. https://
doi. org/ 10. 1063/5. 00130 65
363. Micieli D, Minniti T, Evans LlM, Gorini G (2019) Accelerat-
ing neutron tomography experiments through artificial neural
network based reconstruction. Sci Rep 9:2450–2450. https:// doi.
org/ 10. 1038/ s41598- 019- 38903-1
364. Badran A, Marshall D, Legault Z, Makovetsky R, Provencher B,
Piché N, Marsh M (2020) Automated segmentation of computed
tomography images of fiber-reinforced composites by deep learn-
ing. J Mater Sci. https:// doi. org/ 10. 1007/ s10853- 020- 05148-7
365. Luengo I, Darrow MC, Spink MC, Sun Y, Dai W, He CY, Chiu
W, Pridmore T, Ashton AW, Duke EMH, Basham M, French AP
(2017) SuRVoS: super-region volume segmentation workbench.
J Struct Biol 198:43–53. https:// doi. org/ 10. 1016/j. jsb. 2017. 02.
007
366. Kim SH, Boukouvala F (2020) Machine learning-based surrogate
modeling for data-driven optimization: a comparison of subset
selection for regression techniques. Optim Lett 14:989–1010.
https:// doi. org/ 10. 1007/ s11590- 019- 01428-7
367. Ezhov M, Gusarev M, Golitsyna M, Yates JM, Kushnerev E,
Tamimi D, Aksoy S, Shumilov E, Sanders A, Orhan K (2021)
Clinically applicable artificial intelligence system for dental diag-
nosis with CBCT. Sci Rep 11:15006. https:// doi. org/ 10. 1038/
s41598- 021- 94093-9
368. Baroroh DK, Chu C-H, Wang L (2021) Systematic literature
review on augmented reality in smart manufacturing: collabora-
tion between human and computational intelligence. J Manuf
Syst 61:696–711. https:// doi. org/ 10. 1016/j. jmsy. 2020. 10. 017
369. Bottani E, Vignali G (2019) Augmented reality technology in the
manufacturing industry: a review of the last decade. IISE Trans
51:284–310. https:// doi. org/ 10. 1080/ 24725 854. 2018. 14932 44
370. Sohn J-M, Lee S, Kim D-E (2020) An exploratory study of fit and
size issues with mass customized men’s jackets using 3D body
scan and virtual try-on technology. Text Res J 90:1906–1930.
https:// doi. org/ 10. 1177/ 00405 17520 904927
371. Luh Y-P, Wang J-B, Chang J-W, Chang S-Y, Chu C-H (2013)
Augmented reality-based design customization of footwear for
children. J Intell Manuf 24:905–917. https:// doi. org/ 10. 1007/
s10845- 012- 0642-9
372. Warr R, Ametova E, Cernik RJ, Fardell G, Handschuh S, Jør-
gensen JS, Papoutsellis E, Pasca E, Withers PJ (2021) Enhanced
hyperspectral tomography for bioimaging by spatiospectral
reconstruction. Sci Rep 11:20818. https:// doi. org/ 10. 1038/
s41598- 021- 00146-4
373. Viganò N, Ludwig W (2020) X-ray orientation microscopy using
topo-tomography and multi-mode diffraction contrast tomogra-
phy. Curr Opin Solid State Mater Sci 24:100832. https:// doi. org/
10. 1016/j. cossms. 2020. 100832
374. Kockelmann W, Minniti T, Pooley DE, Burca G, Ramadhan
R, Akeroyd FA, Howells GD, Moreton-Smith C, Keymer DP,
Kelleher J, Kabra S, Lee TL, Ziesche R, Reid A, Vitucci G,
Gorini G, Micieli D, Agostino RG, Formoso V, Aliotta F, Pon-
terio R, Trusso S, Salvato G, Vasi C, Grazzi F, Watanabe K, Lee
JWL, Tremsin AS, McPhate JB, Nixon D, Draper N, Halcrow W,
Nightingale J (2018) Time-of-flight neutron imaging on IMAT@
ISIS: a new user facility for materials science. J Imaging 4:47.
https:// doi. org/ 10. 3390/ jimag ing40 30047
375. Tran KV, Woracek R, Kardjilov N, Markötter H, Hilger A, Kock-
elmann W, Kelleher J, Puplampu SB, Penumadu D, Tremsin AS,
Banhart J, Manke I (2021) Spectral neutron tomography. Mater
Today Adv 9:100132. https:// doi. org/ 10. 1016/j. mtadv. 2021.
100132
376. Shoukroun D, Massimi L, Endrizzi M, Bate D, Fromme P, Olivo
A (2021) Composite porosity characterization using x-ray edge
illumination phase contrast and ultrasonic techniques. In: Health
monitoring of structural and biological systems XV, SPIE, pp
505–512. https:// doi. org/ 10. 1117/ 12. 25825 61.
Publisher's Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.