Conference PaperPDF Available

Augmented BIM based taxonomy for steel design systems: An integrated evolutionary computational approach

Conference Paper

Augmented BIM based taxonomy for steel design systems: An integrated evolutionary computational approach

Abstract and Figures

Within the building industry, approximately 45% of the total material usage arises from the consumption of steel and concrete in construction. Therefore, it is crucial to understand the flow of interactions between materials, processes and activities within buildings’ lifecycle in order to meet national and global environmental emissions’ targets. The present research investigates the potential to reduce the environmental impacts of structural systems though a more efficient use of materials. The main objective of this study is to explore and to develop a novel approach that utilises Building Information Modelling’s capabilities combined with advanced structural analysis and Life Cycle Assessment as well as with a two-staged automated optimisation engine that develops structural efficient and environmentally responsible steel I-beam sections.
Content may be subject to copyright.
Augmented BIM based taxonomy for steel design systems: An integrated
evolutionary computational approach
Stathis Eleftheriadis
University College London
Project objectives and goals
Within the building industry, approximately 45%
of the total material usage arises from the
consumption of steel and concrete in
construction. Therefore, it is crucial to understand
the flow of interactions between materials,
processes and activities within buildings’ lifecycle
in order to meet national and global
environmental emissions’ targets (Ibn-
Mohammed et al, 2013). The present research
investigates the potential to reduce the
environmental impacts of structural systems
though a more efficient use of materials. The
main objective of this study is to explore and to
develop a novel approach that utilises Building
Information Modelling’s (BIM) capabilities
combined with advanced structural analysis and
Life Cycle Assessment (LCA) as well as with a
two-staged automated search engine that
considers structural efficient and environmentally
responsible steel I-beams.
Description of method and results
A prototypical steel framed structural system
under certain loads has been analysed. Four
input parameters of the I-section have been
assumed: web thickness, flange thickness,
section’s width and depth. The obtained optimal
solution is the one that makes more sustainable
use of materials - minimises the weight of the
structure - without compromising its structural
performance - maximises the efficiency of the
tested frame. The system’s architecture utilises
the Application Programming Interface (API) of
Robot and the .NET framework of C♯, and it
inherits several structural functionalities based on
Robot Finite Element Method (FEM) engine. The
main functionalities that have been implemented
within the developed Graphical User Interface
(GUI) are:
Developing a parametric, 3D computer
model of a prototypical steel framed
structural system,
Verifying the structural elements
according to Eurocode BS-EN 1993-
1:2005/NA: 2008/AC: 2009,
Performing sensitivity analysis in order to
identify the correlation matrices between
input and output parameters
Establishing a custom constraint genetic
algorithm that interacts with the
developed computer design model,
Executing calculations and carrying out
validation of the developed 3D models
with the optimised steel sections
In the Monte Carlo analysis, multiple simulations
have been performed using randomly generated
values based on the specified range (Table 1).
From the Pearson correlation of the parametric
structural model after 900 simulations, it has
been observed that within the tested sample the
web thickness appears to have the highest
negative correlation with the efficiency ratio (-
0.8354) (Fig 1) and the highest positive
correlation with the structural weight (0.6338). On
the other hand, the width of the section appears
to have the highest negative correlation with the
slenderness ratio Lamz (-0.9253), whereas the
depth of the section has the highest negative
correlation with the slenderness ratio Lamy (-
0.9385).
The Genetic Algorithm’s (GA) conceptual theory,
developed by John Holland in the 1960’s and
1970’s (Holland, 1992) mimics the evolutionary
processes in nature by populations, reproduction
and heredity, with the inherent ability for the
designer to alter several parameters within the
method such as population size, crossover
technique and mutation rate. The steel beams’
parameters represent the four genes of the GA,
which comprise of the algorithm’s inputs. In
particular, the section’s Depth d and Width B, as
well as the flange’s Thickness T and the Web’s
Thickness t are allowed to vary between the
ranges displayed in the Table 1.
The fitness function is linked to the overall weight
of the structure, and it appears to have a
constant declining trend while the algorithm is
searching the entire design space for the
optimum solution. A constraint method that
penalises the infeasible solutions has been
implemented: a constant penalty to the solutions
that violate the Eurocode’s slenderness
constraints is being applied. The penalty function
for the minimisation problem with m constraints is
shown in Eqn 1.
   
 (1)
The effectiveness of the penalisation functionality
can be seen in Fig 2, where the efficiency ratio
sporadically exceeds the maximum value of 1
during the early generations, and the algorithm,
after approximately 350 generations, converges
to the maximum ratio of 0.999, which is
equivalent to a minimum structural weight of
380kg. In that way the algorithm develops an
“intelligent” behaviour by penalising the solutions
with efficiency ratio greater than 1 and gradually
eliminating them from the final population. The
obtained section details from the GA are: Depth d
= 152.6 mm, Width B = 90.4mm, Web t = 4.2mm,
Flange T = 4.1mm collectively forming a decent
estimation of the design problem. The reduction
in weight compared to the next available
standardised UB 152x89x16 section is 21%. In
regards to the equivalent LCA environmental
indicators the largest reduction of approximately
20.7% has been observed in the Global Warming
Potential (GWP) measurements, whereas the
smallest reduction has been detected for the
Acidification Potential (AP) with 19% reduction
compared to the UB152x89x16 section.
Potential for application of results
Preliminary validation trials of the model have
been conducted within Price & Myers. The
proposed framework could be utilised as a design
tool to inform early stage solutions that improve
the environmental and structural indicators of
steel design systems. The implementation of BIM
platform allows the system to automatically
update the 3D geometry based on the custom
steel sections. In addition the global search
engine of the evolutionary algorithm is an
expandable model and the conflicting nature of
structural engineering problems can be further
analysed. Cost-efficiency and value-engineering
parameters can also be implemented within the
fitness function of the configuration and a multi-
objective examination of the structural system will
allow designers to assess the Pareto front for
trade-offs between potential solutions that not
only reduce environmental impacts but also
reduce construction-fabrication costs.
References
Ibn-Mohammed, T., Greenough, R., Taylor, S.,
Ozawa-Meida, L., & Acquaye, A. (2013).
Operational vs. embodied emissions in
buildingsA review of current trends. Energy
and Buildings 66 , 232-245.
Holland, J. (1992). Genetic Algorithms -
Computer programs that "evolve" in ways that
resemble natural selection can solve complex
problems even their creators do not fully
understand. Scientific American .
Funding body
Engineering and Physical Sciences Research
Council (EPSRC)
Price & Myers
Further information
Stathis Eleftheriadis
(E: efstathios.eleftheriadis.13@ucl.ac.uk)
Professor Dejan Mumovic
(E: d.mumovic@ucl.ac.uk)
Dr Paul Greening
(E: paul.greening@ucl.ac.uk)
Table 1 Steel beam sections specified ranges for
genetic algorithm
Design
Inputs
Genes
Minimum
(mm)
Maximum
(mm)
Depth d
1
130
175
Width B
2
90
125
Flange
thickness T
3
4
6.5
Web
thickness t
4
4
6.5
Figure 1 Correlation between efficiency ratio and
Web thickness
Figure 2 Optimisation results with penalisation
structural weight
ResearchGate has not been able to resolve any citations for this publication.
Article
Global awareness of environmental impacts such as climate change and depletion of ozone layer has increased significantly in the last few years and the implication for emissions reductions in buildings are widely acknowledged. The goal, therefore, is to design and construct buildings with minimum environmental impacts. Lifecycle emissions resulting from buildings consist of two components: operational and embodied emissions. A great deal of effort has been put into reducing the former as it is assumed that it is higher than the latter. However, studies have revealed the growing significance of embodied emissions in buildings but its importance is often underestimated in lifecycle emissions analysis. This paper takes a retrospective approach to critically review the relationship between embodied and operational emissions over the lifecycle of buildings. This is done to highlight and demonstrate the increasing proportion of embodied emissions that is one consequence of efforts to decrease operational emissions. The paper draws on a wide array of issues, including complications concerning embodied emissions computation and also discusses the benefits that come with its consideration. The implication of neglecting embodied emissions and the need for an urgent policy framework within the current climate of energy and climate change policies are also discussed.