Conference PaperPDF Available

Effect of process Parameters on Dimensional accuracy of Down-facing surfaces in Selective Laser Melting of Ti6Al4V

Authors:

Abstract and Figures

Additive Manufacturing (AM) technologies have rapidly spread to cover a wide range of industrial applications, such as aerospace, automotive and biomedical. This is due to their inherent capability to produce parts with complex geometries in short lead times, which shows great potential to strengthen the profitability of supply chains. However, one constraint on the potential growth of AM to be a genuine competitor against existing conventional manufacturing technologies is the limited accuracy and process uncertainty that need addressing to improve AM predictability and precision. This is especially true for powder bed techniques such as SLM, where the as-built quality of produced parts does not satisfy current manufacturing standards and tolerances. In particular, the SLM process faces several challenges such as the shrinkage and distortion due to the buildup of residual stresses, as well as dross formations. A number of research studies have attempted to improve the precision of the SLM process. For instance, the effect of the scanning strategy on distortions and residual stresses have been examined. Different methods have been developed to model the SLM process to enable the prediction and compensation for distortions. Looking at the literature, although there have been noticeable reported studies on the influence of the process conditions of the SLM on the accuracy of the generated parts, only few attempts were made to optimize the process. Particularly to identify the proper processing window and optimal process parameters that would enable the production of highly precise products with tight tolerances and high quality. However, this imposes a systematic experimental study to comprehensively acquire the effect of the process parameters on precision matters such as dimensional accuracy, warpage and deflection. In this context, the aim of the present research work is to experimentally investigate and correlate the effects of different build parameters on the dimensional accuracy of SLM, especially parts containing different overhang angles.
Content may be subject to copyright.
EFFECT OF PROCESS PARAMETERS ON DIMENSIONAL ACCURACY
OF DOWN-FACING SURFACES IN SELECTIVE LASER MELTING OF
TI6AL4V
Amal Charles1, Ahmed Elkaseer1, Tobias Mueller1, Lore Thijs2, Veit Hagenmeyer1,
Steffen Scholz1,3
1Institute for Automation and applied Informatics (IAI)
Karlsruhe Institute of Technology
Karlsruhe, Baden-Württemberg, Germany
2Direct Metal Printing Engineering
3D Systems
Leuven, Flemish Brabant, Belgium
3Karlsruhe Nano Micro Facility
Hermann-von-Helmholtz-Platz 1
Eggenstein-Leopoldshafen, Germany
INTRODUCTION
Additive Manufacturing (AM) technologies have
rapidly spread to cover a wide range of industrial
applications, such as aerospace, automotive and
biomedical [1]. This is due to their inherent
capability to produce parts with complex
geometries in short lead times, which shows
great potential to strengthen the profitability of
supply chains [2]. However, one constraint on the
potential growth of AM to be a genuine competitor
against existing conventional manufacturing
technologies is the limited accuracy and process
uncertainty that need addressing to improve AM
predictability and precision.
This is especially true for powder bed techniques
such as SLM, where the as-built quality of
produced parts does not satisfy current
manufacturing standards and tolerances. In
particular, the SLM process faces several
challenges such as the shrinkage and distortion
due to the buildup of residual stresses, as well as
dross formations [3].
A number of research studies have attempted to
improve the precision of the SLM process. For
instance, the effect of the scanning strategy on
distortions and residual stresses have been
examined [4]. Different methods have been
developed to model the SLM process to enable
the prediction and compensation for distortions
[5].
Looking at the literature, although there have
been noticeable reported studies on the influence
of the process conditions of the SLM on the
accuracy of the generated parts, only few
attempts were made to optimize the process [6].
Particularly to identify the proper processing
window and optimal process parameters that
would enable the production of highly precise
products with tight tolerances and high quality.
However, this imposes a systematic experimental
study to comprehensively acquire the effect of the
process parameters on precision matters such as
dimensional accuracy, warpage and deflection.
In this context, the aim of the present research
work is to experimentally investigate and
correlate the effects of different build parameters
on the dimensional accuracy of SLM, especially
parts containing different overhang angles.
EXPERIMENTS
Test Piece
Separate test pieces containing different angles
of overhangs (45°, 35° and 25°) were designed to
enable the testing of the dimensional accuracy
and they were built with a layer thickness of 60
µm. The test pieces and their dimensions can be
seen in Figure.1.
FIGURE 1. Front view of test pieces
Build Parameters
The down-facing area of an overhang is
especially susceptible to deformations and defect
formation such as dross due to overheating of
loose powder, therefore this down-facing area is
the subject of the present research work. The
most significant process parameters (laser
power, scan speed and scan spacing) [6] were
varied and applied to examine their effects on the
down-facing area, where a hexagonal cell
scanning strategy was applied. A depiction of the
down-facing area can be seen in Figure.2. The
remainder of the part (middle area) were built
using 3D Systems recommended parameters for
a 60 µm layer thickness.
FIGURE 2. Depiction of the down-facing area
Design of Experiment
The chosen parameters are seen in Table.1.
TABLE 1. Selected parameters and their levels
Testpiece
Laser
power
Scan
speed
Scan
spacing
1
90
465
60
2
90
465
90
3
90
1235
60
4
90
1235
90
5
210
465
60
6
210
465
90
7
210
1235
60
8
210
1235
90
9
50
850
75
10
250
850
75
Additive Manufacturing Tests
The test pieces were pre-processed in the
3DXpert software and the production was
implemented in a 3D Systems ProX® DMP 320
and using LaserForm Ti gr23 (A) powder. As
common practice, the test pieces were heat-
treated before removal from the build plate to
minimize any deformations and warpage after
printing during base plate removal. Three built
pieces, exemplifying the three different test
angles can be seen in Figure.3.
FIGURE 3. Built test pieces after removal from
build plate
Measurement and Evaluation
An optical microscope was used to take high
definition pictures of the test pieces in order to
evaluate the dimensional accuracy. The images
were stitched in order to generate one complete
image for each sample that can then be used for
the evaluation. An example of one of the stitched
test pieces can be seen in Figure 4.
FIGURE 4. Stitched image of a 45° test piece
Since the down-facing area is being investigated,
the thickness was chosen as the quality to be
investigated as it is directly affected by the
process parameters used to print the down-facing
surface.
Measurement was done using the MATLAB
image processing toolbox. The process of the
measurement is as follows. First, the cross
section of the as-build sample were scanned
using an optical microscope as seen in Fig. 5a, to
obtain as clear a definition as possible of the
edges representing the boundaries of the part to
be characterized. Second, the Matlab image
processing software was employed to process
the captured pictures and convert them into black
and white images using a thresholding technique,
as depicted in Fig. 5b. Subsequently, an edge
detection technique was applied to define the
segment of the part where its thickness will be
evaluated. By conducting this segmentation, a
matrix of points (Xi, Yi, 0 or 1) is generated, where
Xi and Yi are the coordinates of each pixel and
Middle part of
the test piece
the third binary number gives information whether
this point belongs to the segment of the part to be
measured or not, Fig. 5c. Finally, this data matrix
is read to calculate the thickness of the part at
different heights as can be seen in Fig. 5d.
The thickness measurement values for all test
pieces were then analyzed to determine the error.
Which then made it possible to calculate the error
percentage. Minimizing the error percentage
would directly improve the as-built dimensional
accuracy, therefore it is the parameter chosen for
optimization in the work reported in this paper.
In this research, the analysis of the thickness of
the as-built part was restricted to the thickness
close to the edge, however, in future work this will
be extended to consider part thicknesses at
different positions over the as-build parts. Also,
mechanical measurement using caliper or
another measurment technique will be
undertaken which may give a more ‘bulk’ result,
that can give information on the local dross
formations.
FIGURE 5. Image processing-based
measurement techniques using MATLAB (a)
optical micrograph of the part to be characterized,
(b) black and white conversion of the original
image using predefined threshold, (c) edge
detection of the part, and (d) thickness
measurements at different heights
The measured errors, error percentages and the
standard deviation of the measured thickness
values of the test pieces with 35° and 45°
overhangs are presented in table 2. The
comparisons between two process parameters
for the 35° and 45° down-facing surfaces are
presented in the following section. The following
graphs were made by averaging the error
percentages in order to obtain the plot points for
the two process parameters. Only two
parameters are compared at a time. Future work
will see all parameters being considered during
the multi objective optimization. Results
presented herein are for the 35° and 45° down-
facing surfaces.
TABLE 2. Error and error percentages of test
pieces
Error for 35°
Error for 45°
mm
%
Std.
devi
ation
mm
%
Std.
devi
ation
0.20
11.99
0.05
NA
NA
NA
0.11
6.34
0.02
0.19
9.00
0.05
0.10
5.58
0.09
0.07
3.05
0.04
-0.07
-3.85
0.07
0.26
12.40
0.03
0.77
45.17
0.04
0.39
18.56
0.03
0.70
41.03
0.04
0.56
26.42
0.02
0.35
20.29
0.06
0.26
12.40
0.02
0.27
15.76
0.04
0.27
12.83
0.06
-0.09
-4.98
0.08
-0.03
-1.41
0.05
0.51
29.71
0.04
0.35
16.65
0.05
RESULTS
Effect of Laser power and Scan speed
Figure 6 presents the effect of laser power and
scan speed on the obtainable error percentage as
an indicator of the dimensional inaccuracies.
Error percentages are seen to be higher at the
higher power value of 210W and decrease in
error percentage as it can be seen at a high scan
speed. At the laser power of 90W both 35° and
45° down-facing surfaces show very similar error
percentages at a low scan speed. However, at a
higher scan speed the 35° surface shows a more
significant decrease in error percentage whereas
the difference is minimal for the 45° overhang at
the same speed.
Table 3 depicts the average standard deviations
for the test pieces at the measured power and
scanning speed. The results indicate that for the
35° overhangs the slower speed of 465mm/s
produces a more uniform part than the high
speed resulting in lower standard deviations at
(a)
(b)
(c)
(d)
1.828 mm
1.759 mm
1.793 mm
1.828 mm
1.867 mm
both level of laser power. While for the 45° the
relationship differs as at the lower power value,
standard deviation decreases while increasing
speed and for the higher power of 210W the
standard deviation increases while increasing
speed.
FIGURE 6. Relationship between laser power
and scan speed
TABLE 3. Standard deviations
Overhang
angle
laser
power
(W)
Std.
deviation
at 465
mm/s (mm)
Std. deviation
at 1235 mm/s
(mm)
35°
90
0.035
0.083
210
0.039
0.051
45°
90
0.046
0.033
210
0.025
0.038
Effect of Laser power and Scan spacing
When comparing laser power and scan spacing,
consistent with the previous table, the error
percentage is always seen to be higher for the
test pieces printed with the higher power of 210
W, as seen in figure 7. However, the influence of
increasing the scan spacing is seen to be
different between the 35° and 45° test pieces.
The error percentage decreases for the 35°
surface while increasing the scan spacing seems
to increase error for the 45° samples.
FIGURE 7. Relationship between laser power
and scan spacing
Table 5 depicts the average standard deviations
at the measured laser power and scan spacing.
The results indicate that at the 35° overhang the
standard deviation is lower at the 90µm than at
the 60µm scan spacing. While for the 45° down-
facing surface, at 90W, the standard deviation
only shows a small increase at the 90µm scan
spacing. While at the 210W, the standard
deviation shows a larger increase at the 90µm
scan spacing from the 60µm. While for both
angles, at the 60µm scan spacing the standard
deviation is consistently lower for the higher
power.
TABLE 4. Standard deviations
Overhang
angle
laser
power
(W)
Std.
deviation at
60µm (mm)
Std.
deviation at
90µm (mm)
35°
90
0.071
0.048
210
0.047
0.042
45°
90
0.036
0.038
210
0.025
0.038
Effect of Scan speed and Scan spacing
Observations from figure 8, while comparing scan
speed and scan spacing show that the varying of
these two parameters have different effects on
the accuracy depending on the angle of the
overhang. For a 35° overhang, the low speed and
low scan spacing has a large error but the error
reduces drastically when the scan spacing is
increased. Though the reduction is also present
for the high-speed built 35° surface, the degree of
reduction is lesser. The 45° surface shows a
small decrease when increasing scan spacing but
the error is reduced even more for a higher speed
build. These results indicate that the interaction
0
5
10
15
20
25
30
35
40
45
50
465 1235
Error percentage (%)
Scan Speed (mm/s)
Laser
power
210W at
35°
Laser
power
90W at
35°
Laser
power
210W at
45°
Laser
power
90W at
45°
0
5
10
15
20
25
30
35
60 90
Error Percentage (%)
Scan spacing (µm)
Laser
power
90W at
45°
Laser
power
210W
at 45°
Laser
power
90W at
35°
Laser
power
210W
at 35°
effect of scan spacing and scan speed have
different intensities at different angles.
FIGURE 8. Relationship between scan speed
and scan spacing
Table 6 depicts the average standard deviations
for the measured scan speed and scan spacing.
For the 35° angle, the 90µm scan spacing
showed lower deviations than the 60µm while the
1235 mm/s scan speed showed higher deviations
than the 465mm/s scan speed. While for the 45°,
at 465 mm/s, the higher scan spacing only shows
a small increase in standard deviation. While the
difference in standard deviation is higher at the
higher speed.
TABLE 5. Standard deviations
Overhang
angle
Scan
speed
(mm/s)
Std.
deviation
at 60µm
(mm)
Std.
deviation at
90µm (mm)
35°
465
0.041
0.032
1235
0.077
0.057
45°
465
0.031
0.032
1235
0.027
0.043
DISCUSSIONS
First results show that one of the dominant effects
affecting the dimensional accuracy is the laser
power. Higher laser powers consistently showed
high values of error percentage. This can be
explained intuitively, as a higher energy
absorption by the powder leads to the formation
of larger melt pools. This coupled with the
presence of only loose powder below the melt
pool leads to an increased presence of partially
melted powder on the surface of the overhang,
leading to high dimensional inaccuracies.
45° down-facing surfaces generally exhibit self-
supporting behavior to an extent. Therefore, to a
certain degree they are able to dissipate heat into
the build platform. However due to a larger
amount of loose powder below the 35° test
pieces, the heat does not escape, as a result an
overheated zone where more powder is melted
causing more deformation and dross formation.
This can be seen as the 35° overhangs
consistently showed greater error percentages
than the 45° overhangs.
One thing to keep in mind is that large melt pools
generally have a greater wettability and can
therefore easily fuse together with adjacent melt
pools, which could cause the formation of large
uniform dross. This kind of dross can manifest as
smooth flat surfaces that appear to have good
surface quality. Therefore, the study of
dimensional accuracy must be coupled together
with other surface profile studies such as the
surface roughness.
The effect of changing the scan spacing is
interesting as it has a different effect depending
on the angle of the overhang. For the 45°, at both
the high and low laser power, increasing the scan
spacing led to larger error percentages, while the
error percentages decreased for the 35°
overhang. Moreover, for the 35° surfaces, built
with a low scan speed, increasing the scan
spacing showed a drastic decrease in the error
percentage, while the effect was not as
pronounced for the parts built with a higher scan
speed. On the other hand, for the 45° test piece,
the low speed sample showed a miniscule
decrease in error percentage and the high-speed
part actually saw an increase in error percentage.
Therefore, the effect of scan spacing is also
geometry dependent.
Conducting an investigation into parameters such
as line energy can give insight into the size of the
melt pools formed, leading to a better
understanding of the scan spacing parameter,
which allows us to learn more about the optimum
overlap between melt pools and as a result will
enable better decision making on optimum
scanning space.
Dross formation can cause local variations on the
dimensions of down-facing surfaces thereby
affecting the thickness. In this case, investigating
the standard deviations gives an idea on the
uniformity of a surface and therefore information
on the stability/instability of the process can be
procured. Tables 4, 5 and 6 depict the process
interactions that show the variations in thickness
0
5
10
15
20
25
30
35
60 90
Error Percentage (%)
Scan spacing (µm)
Scan speed
465mm/s
at 45°
Scan speed
1235mm/s
at 45°
Scan speed
465mm/s
at 35°
Scan speed
1235
mm/s at
35°
measurements. Thereby giving an idea on the
combinations that give rise to unstable
processes. This information is valuable while
conducting an optimization study where the
development of a precision process is the goal.
Repeatedly the SLM process has shown that
apart from its non-linearity, there exist complex
interactions between parameters. In addition,
uncertainties, such as the effect of middle
parameters on the down-facing surface quality,
exist. This adds to its complexity and
unpredictability, making it a prime candidate for
process modelling and optimization.
CONCLUSIONS
A beginning into the process modelling and
optimization of the SLM process is shown in this
paper. The first process of data acquisition is
currently underway with some initial results and
discussions presented herein.
1. Laser power is a dominant factor and
further work must focus on including the
effect of factors such as normalized
enthalpy on the obtainable quality of the
as-build part.
2. Higher laser power consistently showed
higher values of error percentages.
3. 35° down-facing surfaces consistently
showed higher error percentages than
the 45° surfaces as expected. This is due
to ability of 45° overhangs to provide a
degree of self-support and therefore
possesses a better ability to dissipate
heat into the build platform.
4. The effect of scan spacing is also
geometry dependent and factors such as
line energy must be investigated for a
better understanding.
5. The dimensional accuracy study,
coupled with other surface quality studies
such as roughness is required for a true
understanding of the process.
6. Finally, including information on the
variation of data points is important for
the process optimization of the SLM
process when precision is the goal.
ACKNOWLEDGEMENTS
This work was undertaken in the H2020-MSCA-
ITN-2016 project PAM2, Precision Additive Metal
Manufacturing, which is funded by The EU
Framework Programme for Research and
Innovation - Grant Agreement No 721383. This
work was implemented under the STN
programme, part of the Helmholtz association. In
addition, the support by the Karlsruhe Nano Micro
Facility (KNMF-LMP, http://www.knmf.kit.edu/) a
Helmholtz research infrastructure at KIT, is
gratefully acknowledged.
REFERENCES
[1] K. Brans, “3D Printing, A Maturing
Technology”, IFAC Proc., 2013; 46; 468-472.
[2] N. Guo, Leu Ming C, “Additive manufacturing:
technology, applications and research needs”,
Fron. of Mech. Eng., 2013; 8; 215-243.
[3] Wang Di, Mai Shuzhen, Xiao Dongming, Yang
Yongqiang,“Surface quality of the curved
overhanging structure manufactured from 316-L
stainless steel by SLM”, The Int. Jou. Of Adv.
Manu. Tech., 86; 781-792
[4] Stacey D. Bagg, Lindsay M. Sochalski-
Kolbus, Jeffrey R. Bunn, “ The effect of laser scan
strategy on distortion and residual stresses if
arches made with Selective Laser Melting”, ASPE
Summer topical meeting 2016, Raleigh, NC,
United States.
[5] Shukri Afazov, Annestacy Okioga, Adam
Holloway, Willem Denmark, Andrew
Triantaphyllou, Sean-Anthony Smith, Liam
Bradley-Smith, “ A methodology for precision
additive manufacturing through compensation”
Prec. Eng, 2017;50;269-274.
[6] R. Mertens, K. Kempen, S. Clijsters, J.-P.
Kruth, Production of AlSi10Mg parts with
downfacing areas by Selective Laser Melting”, 6th
PMI conf. 2014, Guimarães, Portugal.
... According to Tang et al. [74] and Fotovvati et al. [75], the thickness of the individual layer and the hatching distance possess a mild effect in terms of dimensional accuracy; nevertheless, a slightly improvement of the accuracy is observed in situations where these parameters are increased to a certain extent. Charles et al. [76] and Zhang et al. [77] declare that as the laser power increases the accuracy tends to decrease and in case of employing high scan speeds, the dimensional accuracy could be enhanced to some extent. Studies have shown that high scan speeds and low laser power create small melting pools resulting to weaker shrinkage effects, thus the dimensional accuracy is relatively higher. ...
Article
Full-text available
Metal additive manufacturing (AM) has been recently acknowledged as a method to produce metal parts with complex geometries and unique characteristics. In general, metal AM components are used in applications where enhanced mechanical properties are required; thus it is essential to optimize the parameters that affect the metal AM processes. The immense advantages of metal AM technologies in the fields of production and manufacturing placed these technologies in the forefront of scientific sectors relevant to engineering, biomedical science and electronics. In recent years, many studies have been published analyzing the parameters which influence metal AM techniques to achieve advanced overall quality and optimized mechanical behavior of the fabricated products. Existing research has individually studied the influence of various factors on metal AM methods such as feedstock material properties, build orientation, printing conditions, infill patterns and post-processing procedures. The present study aims to review the research carried out until today identifying and classifying all parameters affecting metal AM, based on their impact on the quality of the final product. This survey intents to establish a process map of the categories which influence metal AM components. Additional recommendations, research gaps and directions towards the improvement of metal AM are also provided.
... 24 In another study, the effects of different processing parameters on part dimensional accuracy and different protrusion angles in the SLM method using image processing techniques were investigated. 25 In this study, dimensional control of the manufacturing part is carried out with EOS M290 Direct Metal Additive Manufacturing (DMLS) device with Yb fiber laser; 400 W by using maraging steel material. Different geometric shapes are used in the drawing of the part to perform dimensional control correctly with the image processing method. ...
Article
In this study, a sample part is manufactured by using the Powder Bed Fusion (PBF) method with maraging steel metal powder, and the manufactured part is examined in terms of dimensional accuracy by image processing techniques. The sample =art is drawn with various geometric shapes so that dimensional and areal controlling with the image processing techniques can be performed accurately. The purpose of dimensional control is to choose the most suitable manufacturing parameters to ensure the closest production of the real dimensions of the part. With the detection of significant deviations in dimensional values by image processing techniques, the designs of the part should be changed according to these deviations. In this study, each layer of the manufactured sample part image is captured by a camera. Afterward, dimensional and areal percent error rates are determined by three different methods based on image processing techniques.
... In order to prevent excessive melting of powder particles and dross formation, adjusting processing parameters is commonly suggested [16,25,26]. Several studies lowered the energy density in the down-skin region [15,18,27] and identified the optimum conditions in the SLM process, e.g., a low laser velocity and power (65 W for aluminum [27] and 35 W for titanium [18]). ...
Article
Full-text available
Powder-bed fusion additive manufacturing technology makes it possible to produce parts with complicated geometry and high accuracy. However, dimensional deviation caused by powder overmelting and dross formation is still a challenge for manufacturing thin channels. In this study, the origins of the overmelting of printed thin channels were analyzed and a concept called “melting cell” is proposed to describe and quantify the geometric error. Based on the geometrical relationship between the melting cell and target channel, a method for predicting and optimizing the final geometry of thin channels is outlined. In order to verify the method, geometries of thin horizontal circular channels in various sizes are studied as examples. The predicted results by the proposed method show a remarkable agreement with available experimental results. Moreover, a new egg-shaped compensated design, which is able to improve the dimensional accuracy of thin horizontal circular channels, is presented. The proposed method is simple yet very effective. It can be easily extended to the manufacturing of thin channels with various shapes, materials, and different powder bed fusion processes.
... The research reported in the literature can be divided into two major areas, namely PBF metal-and polymer AM-related. In metal AM, Charles et al. [7,8] assessed the impact of using different laser powers, scan speeds and hatch spacing on the generated surface roughness of the downfacing surfaces of the Selective Laser Melted (SLM) parts. Whip et al. [9] identified a strong correlation between surface average height (Sa), maximum valley depth (Sv) with laser powder in SLM. ...
Article
Full-text available
High Speed Sintering is an advanced powder bed fusion polymer Additive Manufacturing technique aimed at economical production of end-use parts in series manufacture. Surface finish is thus of high importance to end users. This study investigates the surface topography of High Speed Sintered parts produced using a range of different energy-related process parameters including sinter speed, lamp power and ink grey level. Areal surface texture was measured using Focus Variation microscopy and the sample porosity was systematically examined by the X-ray Computed Tomography technique. Surface topography was further characterised by Scanning Electron Microscopy, following which the samples were subject to tensile testing. Results showed that areal surface texture is strongly correlated with porosity, which can be further linked with mechanical properties. Certain texture parameters i.e. arithmetic mean height Sa, root-mean-square Sq and maximum valley depth Sv were identified as good indicators that can be used to compare porosity and/or mechanical properties between different samples, as well as distinguish up-, down-skins and side surfaces. Sa, Sq and Sv for up- and down-skins were found to correlate with the above energy-related process parameters. It was also revealed that skewness Ssk and kurtosis Sku are related to sphere-like protrusions, sub-surface porosity and re-entrant features. Energy input is the fundamental reason that causes varying porosity levels and consequently different surface topographies and mechanical properties, with a 10.07 µm and a 30.21% difference in Sa and porosity, respectively, between the ‘low’ and ‘high’ energy input.
... It should be mentioned that the build process parameters: laser power, scan speed and scan spacing affect the dimensional accuracy (Charles et al., 2018;Han et al., 2018) and density (Han et al., 2018) of SLM manufactured walls and struts. However, these effects were not considered in the current work that requires a future investigation for the optimum process parameter effects on the elastic behavior of hexagonal honeycombs. ...
Article
Purpose-Honeycomb cellular structures exhibit unique mechanical properties such as high specific strength, high specific stiffness, high energy absorption and good thermal and acoustic performance. This paper aims to use numerical modeling to investigate the effective elastic moduli, in-plane and out-of-plane, for thick-walled honeycombs manufactured using selective laser melting (SLM). Design/methodology/approach-Theoretical predictions were performed using homogenization on a sample scale domain equivalent to the as-manufactured dimensions. A Renishaw AM 250 machine was used to manufacture hexagonal honeycomb samples with wall thicknesses of 0.2 to 0.5 mm and a cell size of 3.97 mm using 304 L steel powder. The SLM-manufactured honeycombs and cylindrical test coupons were tested using flatwise and edgewise compression. Three-dimensional finite element and strain energy homogenization were conducted to determine the effective elastic properties, which were validated by the current experimental outcomes and compared to analytical models from the literature. Findings-Good agreement was found between the results of the effective Young's moduli ratios numerical modeling and experimental observations. In-plane effective elastic moduli were found to be more sensitive to geometrical irregularity compared to out-of-plane effective moduli, which was confirmed by the analytical models. Also, it was concluded that thick-walled SLM manufactured honeycombs have bending-dominated in-plane compressive behavior and a stretch-dominated out-of-plane compressive behavior, which matched well with the simulation and numerical models predictions. Originality/value-This work uses three-dimensional finite element and strain energy homogenization to evaluate the effective moduli of SLM manufactured honeycombs.
... It should be mentioned that the build process parameters: laser power, scan speed and scan spacing affect the dimensional accuracy (Charles et al., 2018;Han et al., 2018) and density (Han et al., 2018) of SLM manufactured walls and struts. However, these effects were not considered in the current work that requires a future investigation for the optimum process parameter effects on the elastic behavior of hexagonal honeycombs. ...
Article
Full-text available
Purpose Honeycomb cellular structures exhibit unique mechanical properties such as high specific strength, high specific stiffness, high energy absorption and good thermal and acoustic performance. This paper aims to use numerical modeling to investigate the effective elastic moduli, in-plane and out-of-plane, for thick-walled honeycombs manufactured using selective laser melting (SLM). Design/methodology/approach Theoretical predictions were performed using homogenization on a sample scale domain equivalent to the as-manufactured dimensions. A Renishaw AM 250 machine was used to manufacture hexagonal honeycomb samples with wall thicknesses of 0.2 to 0.5 mm and a cell size of 3.97 mm using 304 L steel powder. The SLM-manufactured honeycombs and cylindrical test coupons were tested using flatwise and edgewise compression. Three-dimensional finite element and strain energy homogenization were conducted to determine the effective elastic properties, which were validated by the current experimental outcomes and compared to analytical models from the literature. Findings Good agreement was found between the results of the effective Young’s moduli ratios numerical modeling and experimental observations. In-plane effective elastic moduli were found to be more sensitive to geometrical irregularity compared to out-of-plane effective moduli, which was confirmed by the analytical models. Also, it was concluded that thick-walled SLM manufactured honeycombs have bending-dominated in-plane compressive behavior and a stretch-dominated out-of-plane compressive behavior, which matched well with the simulation and numerical models predictions. Originality/value This work uses three-dimensional finite element and strain energy homogenization to evaluate the effective moduli of SLM manufactured honeycombs.
... To mitigate these problems the most common approach is to separate the sample in two zones, namely the middle zone which stands over the bulk metal of the previous layer and the downfacing zone which is printed directly above lose powder [9]. For each region the process parameters can be tuned in order to achieve the required properties (minimum density, low residual stresses, high surface quality and so on); in particular for the downfacing area a more stable weld track is generally desired [10,11,12]. ...
Conference Paper
Full-text available
In this study the accuracy of downfacing surfaces of small LPBF (Laser Powder Bed Fusion) fabricated features is investigated. Calibration geometries with downfacing angles ranging from 75° to 25° are printed in three different layer thicknesses to quantify the dimensional error due to dross formation and other melt pool phenomena in the downfacing area. Their dimension is measured with a digital microscope and compared with the CAD model: all geometries show a negative mismatch which is proportional to the downfacing angle, therefore a correction parameter is calculated. To validate the effectiveness of this offset, microchannels with diameter ranging from 1 mm to 0.6 mm were printed both with and without the correction. Their deviations from the nominal diameter were investigated through the use of micro Computed Tomography. Results show less deviation from the nominal cylindrical CAD when the offset is applied, while the dross formation remains almost unaffected.
... They found more than 1% inaccuracy in the AM parts. Charles et al. [16] used a 3D Systems ProX® DMP 320 machine to correlate different process parameters, including laser power, scan speed, and scan spacing to the dimensional accuracy of DMLS for Ti-6Al-4V ELI parts containing different overhanging angles. The effect of powder reuse is investigated in [17]. ...
Article
Full-text available
A comprehensive investigation of the size and geometry dependency of the dimensional accuracy of direct metal laser sintering (DMLS) for Ti-6Al-4V ELI is presented. For features such as walls, squares, tubes, and rods with different sizes, the percent error significantly increases with decreasing the feature size. The polynomial function a t-b is suggested to describe this size dependency of the dimensional error where a and b are parameters depending on the geometry, material, and DMLS process parameters. This function is used to successfully predict the dimensional error in DMLS of two spinal cages. Therefore, these functions can be used to account for these errors in DMLS by design change or by adjusting DMLS scaling factors. Furthermore, the inconsistency of the DMLS-manufactured dimensions within the feature is shown to be in the same range of the dimensional inconsistency for features located at different positions on the build platform, implying that the location of the feature on the build platform has a negligible effect on the dimensional accuracy. Finally, it is shown that the error in the position accuracy of DMLS-manufactured features is negligible when the size dependency of the dimensional features is considered in the measurements.
... Further results indicate that only looking at the roughness does not provide reliable information on the presence and formation of dross. Dimensional accuracy tests are also required and are the focus in [11]. Post-processing AM parts might need post-processing as the final surface roughness is often higher than what is allowed for most applications. ...
Conference Paper
Full-text available
PAM^2, which stands for Precision Additive Metal Manufacturing, is a European MSCA project in which 10 beneficiaries and 2 partners collaborate on improving the precision of metal Additive Manufacturing. Within this project, research is done for each process stage of AM, going from the design stage to modelling, fabricating, measuring and assessment. For each step we aim to progress the state of the art with a view on improving the final AM part precision and quality by implementing good precision engineering practice. In this article we will list PAM^2’s detailed objectives, the envisioned approach and the results achieved after 1,5 years of research.
Article
Full-text available
The efficiency of fabricating an overhanging structure by selective laser melting (SLM) is an important indicator of the performance of metallic parts. This is due to the fact that defects such as warpage and adherent dross may occur during fabrication of the curved surfaces of overhanging structures. In order to investigate the optimum conditions for fabrication of the curved surfaces of the overhanging structures, experiments were carried out using 316-L stainless steel powder. Initially, the almost 100 % dense parts were fabricated. Then, a model that has a circular curved surface along the Z axis was designed. For a given fabrication depth of 25 μm, several overhanging structures were produced when the laser scanning energy input ranges from 0.15 to 0.6 J/mm. Results show that the upper surface of the almost 100 % dense cube fluctuates like ripples and that the fabrication quality of the curved surface of the overhanging structure varies greatly depending on the energy input and the obliquity angle. For a given energy input of 0.2 J/mm, the obliquity angle for fabricating a totally overhanging surface is as low as 30°. The warpage and adherent dross grow with an increase in the energy input and a decrease in the obliquity angle. Warpage may accumulate, and the accumulated warpage of many layers significantly exceeds the predetermined thickness of the layer. All the four overhanging structures fabricated using varying energy inputs have the following four zones: no dross surface, dense-sinking transition surface, totally sinking surface, and forming failure surface. In the overhanging structures, fabricated with varying laser energy parameters, the angle corresponding to each region was different. The quality of the overhanging surface can be improved by reducing the laser energy. Additionally, a better overhanging surface can be obtained by increasing the obliquity angle. The variation trend of the roughness Rz was almost the same as that of Ra, but the variation range of Rz was much larger than that of Ra. Finally, a foldable abacus with several curved-surface overhanging structures was fabricated to verify the research results. Fundamental methods for controlling and optimizing the SLM-based direct fabrication of curved surfaces of overhanging structures are proposed in this paper, from the perspectives of crafting and design.
Article
Full-text available
Additive manufacturing (AM) technology has been researched and developed for more than 20 years. Rather than removing materials, AM processes make three-dimensional parts directly from CAD models by adding materials layer by layer, offering the beneficial ability to build parts with geometric and material complexities that could not be produced by subtractive manufacturing processes. Through intensive research over the past two decades, significant progress has been made in the development and commercialization of new and innovative AM processes, as well as numerous practical applications in aerospace, automotive, biomedical, energy and other fields. This paper reviews the main processes, materials and applications of the current AM technology and presents future research needs for this technology.
Article
Maturation of powder-bed Additive Manufacturing (AM) is essential for the business benefit the rapid adoption of AM offers to industry. One of the principal challenges in powder-bed AM is the mitigation of distortion due to material shrinkage and residual stresses induced during the build process. In order to address this, a new methodology for distortion compensation is developed and presented in this paper. The novelty of the methodology lies in the use of a mathematical model for pre-distorting the design geometry based on 3D optical scanning measurement data. The methodology has been applied to two industrial Inconel 718 components (a turbine blade and an impeller). It was experimentally demonstrated that distortion compensation is achievable using the proposed methodology. The results showed the compensation methodology reduced distortion from approximately ±300 μm to approximately ±65 μm for both components. In summary, the novel methodology can be used to deliver near-zero distorted parts for industry using powder-bed AM processes.
3D Printing, A Maturing Technology
  • K Brans
K. Brans, "3D Printing, A Maturing Technology", IFAC Proc., 2013; 46; 468-472.
The effect of laser scan strategy on distortion and residual stresses if arches made with Selective Laser Melting
  • D Stacey
  • Lindsay M Bagg
  • Jeffrey R Sochalski-Kolbus
  • Bunn
Stacey D. Bagg, Lindsay M. Sochalski-Kolbus, Jeffrey R. Bunn, " The effect of laser scan strategy on distortion and residual stresses if arches made with Selective Laser Melting", ASPE Summer topical meeting 2016, Raleigh, NC, United States.
Production of AlSi10Mg parts with downfacing areas by Selective Laser Melting
  • Kruth
Kruth, "Production of AlSi10Mg parts with downfacing areas by Selective Laser Melting", 6 th PMI conf. 2014, Guimarães, Portugal.