Available online at www.sciencedirect.

com

ScienceDirect
Procedia Manufacturing 10 (2017) 791 – 803

45th SME North American Manufacturing Research Conference, NAMRC 45, LA, USA

Experimental Optimization of Fused Deposition Modelling

Processing Parameters: a Design-for-Manufacturing Approach
Ala’aldin Alafaghania, Ala Qattawia *, Buraaq Alrawia, Arturo Guzmana
a
Mechanical Engineering, University of California, Merced, Merced, California, 95343, USA

Abstract

The Additive Manufacturing (AM) technology initially was developed as a rapid prototyping tool for visualization and validation
of designs. The recent development of AM technologies, such as Fused Deposition Modelling (FDM), is driving it from rapid
prototyping to rapid manufacturing. However, building end-user functional parts using FDM proved to be a challenging task. The
difficulty arises from the large number of processing parameters that affect the final part design such as: building direction,
extrusion temperature, layer height, infill pattern and more. The processing parameters of FDM influence the quality of the parts
and their functionality. In addition, a more systemic understanding is required to elaborate on the impact of the FDM processing
parameters on the final part’s mechanical properties, dimensional accuracy and building time. The presented paper provides an
experimental study to investigate the independent effect of each processing parameter on the mechanical properties and
dimensional accuracy repeatability of FDM parts. A total of 18 test specimen samples were printed using varying processing
parameters. In order to investigate the repeatability and resulted tolerances, the dimensions of these specimens were measured
and compared with a 3D CAD model. The presented work utilizes a tensile test per ASTM D638 standards to obtain the
mechanical properties of each fabricated sample. In addition, the work provides a Finite Element Analysis (FEA) model for AM
parts. The work suggested to simulate their behavior under mechanical loads for future investigation on the coupled effects of
processing parameters.

©©2017
2017Published by Elsevier
The Authors. B.V. by
Published This is an open
Elsevier B.V.access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-reviewunder
Peer-review underresponsibility
responsibility of the
of the Scientific
organizing Committee
committee of 45th
of the NAMRI/SME.
SME North American Manufacturing Research Conference

Keywords: Fused Deposition Modeling; Design for Manufacturing; Optimization of Processing Parameters; 3D printing; Mechanical Properties;
Finite Element Analysis.

* Corresponding author. Tel.: +0-000-000-0000 ; fax: +0-000-000-0000 .

E-mail address: aqattawi@ucmerced.edu

2351-9789 © 2017 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of the 45th SME North American Manufacturing Research Conference
doi:10.1016/j.promfg.2017.07.079
792 Ala’aldin Alafaghani et al. / Procedia Manufacturing 10 (2017) 791 – 803

1. Introduction

AM is the generic term for the collective advanced manufacturing technologies that build parts layer by layer.
The layers are produced by adding material instead of removing it as opposed to subtractive manufacturing such as
machining. The material addition or fusion is controlled by G-codes generated directly from 3D CAD models. FDM,
one of the AM technologies, builds parts layer by layer by heating a thermoplastic filament to a semi-liquid state and
extruding it through a small nozzle per 3D CAD models usually in STL format as shown in Figure 1. The filament is
usually of circular cross section with specific diameters for each FDM system. The most widely used diameters are
either 1.75 mm or 3.0 mm.
Due to the nature of FDM process, many advantages arise, such as the design freedom to produce complex
shapes without the need to invest in dies and molds, the ability to produce internal features, which is impossible
using traditional manufacturing techniques. FDM enables the reduction of the number of assemblies by producing
consolidated complex parts [1]. More advantage of FDM can be reaped through the supply chain by reducing the
lead time and the need for storage and transportation, especially in applications where high customization is
necessary [2]. On the other hand, FDM technology has challenges; such as producing parts with anisotropic
mechanical properties, staircase effect at curves, coarse surface finish, the need for supports for overhanging regions
and more. To overcome these challenges, many researchers focus on refining the quality of FDM parts. Techniques
to improve the quality of AM or FDM parts, in particular, vary between chemical treatment [3]–[6], machining [7]
[8], heat treatment [9], and optimization of processing parameters.

Figure 1: FDM process schematic

Optimization of processing parameters has been the focus in the literature because reducing post processing is one
of AM’s goals. The optimization of processing parameters can be divided into experimental investigation and
simulation modeling. Notable experimental methods and designs of experiments were used to optimize the
processing parameters such as Taguchi method [10]–[12], full factorial designs [13]–[15], ANOVA [16], [17]
bacteria forging technique [18] and fuzzy logic [19]. In these studies, a variety of processing parameters sets were
investigated. Some papers focus on only one processing parameter such as building direction [20] and investigates it
thoroughly, while other papers focus on 3 or 4 parameters at the same time and study their coupled effects such as in
[15], [18] and [21] where the effect of layer height, building direction, raster angle and more parameters are
investigated at the same time. In the second approach, usually 2 or 3 levels are investigated for each parameter.
On the other hand, for simulation modeling FEA is the most used method. Domingo-Espin et al. [22] simulated
the effect of building direction on mechanical behavior using modeling FDM parts as orthotropic material. Farbman
& McCoy [23], simulated the effect of infill pattern on mechanical properties by adding the infill features in the
simulated CAD model for parts with low infill percentage, where the infill patterns are visible with the naked eye.
 Ala’aldin Alafaghani et al. / Procedia Manufacturing 10 (2017) 791 – 803 793

Despite there is many published work on optimizing the processing parameters for AM, very few discuss the
overall effect of the processing parameters, particularly the influence of the extrusion temperature on the mechanical
properties. In addition, the body of knowledge lacks the understanding for design optimization for FDM parts for
high repeatability and dimensional accuracy. Thus, a systematic way of investigating the influence of the processing
parameters is needed. This work focuses on investigating the effect of each parameter separately at more levels than
previous work does to get a clearer understanding of those effects. Also, this work establishes a new approach to
simulate the mechanical behavior of FDM parts by using the relative density of the printed parts to generating voids
or porosities to modify the model used in FEA of FDM parts and compares it with another simulation method and
the experimental data.

2. Methodology

2.1. Sample preparation

The specimens used in this study to evaluate the dimensional accuracy, repeatability, and mechanical properties
are modelled based on American Society for Testing and Materials ASTM D638 type IV standards for plastic tensile
testing. Figure 2 illustrates the dimensions used to create the CAD model. All the specimens were printed using PLA
filaments and a Makerbot Replicator 2X 3D printer.

Figure 2: Created specimens’ CAD model, dimensions in mm

The effects of six processing parameters on the mechanical properties, dimensional accuracy, and repeatability
were investigated independently. Thus, only one processing parameter’s value was changed at a time in each
fabricated sample. The investigated processing parameters were the building direction, printing speed, extrusion
temperature, layer height, infill percent and infill patterns. The building direction indicates the normal vector to the
printed layers using the axis in Figure 2 as a reference. For example, fabricating samples with building direction Z
means that the printed layers are parallel to the x-y plane. Infill percentage indicates how full or hollow the printed
part is, where zero percent is a shell and 100% is a solid. The printing speed controls how fast the nozzle of the 3D
printer moves while extruding the heated filament, whereas the extrusion temperature controls the temperature of the
extruding nozzle. The layer height controls the thickness of each layer. Finally, the infill patterns control how the
nozzle fills and raster across the infill layers. Three infill patterns were used in this study; those are linear, diamond
and hexagonal as shown in Figure 3, where Diamond F has the same shape as the Diamond infill pattern, but using
faster printing G codes due to the difference in partitioning of the pattern itself. However, this faster printing pattern
puts higher loads on the extruder for changing its directions abruptly. Each of these processing parameters was
investigated at a four-level scale as shown in Table 1. except for building direction, where there were three
directions only.
794 Ala’aldin Alafaghani et al. / Procedia Manufacturing 10 (2017) 791 – 803

Figure 3: Infill patterns shape schematic

Table 1: The Values of varying Processing Parameters.

Levels Building Infill Print Speed Extrusion Layer Height Infill Pattern
Direction Percent Temperature

0 (Reference) Z 100 90 mm/s 185 C0 0.30 mm (1) Diamond F

1 X 20 70 mm/s 175 C0 0.10 mm (2) Diamond
2 Y 50 120 mm/s 180 C0 0.25 mm (3) Linear
3 Z 80 170 mm/s 205 C0 0.40 mm (4) Hexagonal

The values of the processing parameters listed in Table 1 were used to set an experiment plan to fabricate a total
of 18 samples. Table 2 lists the exact processing parameters for each sample. The bolded values indicate which
processing parameter is focused on for that sample. For each sample, three specimens were fabricated to investigate
the repeatability effect in the resulted mechanical proprieties and dimensional accuracy. Since only one parameter is
varied at a time the combined effect of multiple parameters at the same time is not in the scope of this work.
However, it is certainly set as the future direction.

Table 2 Sample processing parameters specification

Sample Building Infill Percent Print Speed Extrusion Layer Height Infill Pattern
Direction Temperature

1 X 100 90 mm/s 185 C0 0.30 mm Diamond (F)

2 Y 100 90 mm/s 185 C0 0.30 mm Diamond (F)
3 Z 100 90 mm/s 185 C0 0.30 mm Diamond (F)
4 Z 20 90 mm/s 185 C0 0.30 mm Diamond (F)
5 Z 50 90 mm/s 185 C0 0.30 mm Diamond (F)
0
6 Z 80 90 mm/s 185 C 0.30 mm Diamond (F)
7 Z 100 170 mm/s 185 C0 0.30 mm Diamond (F)
8 Z 100 120 mm/s 185 C0 0.30 mm Diamond (F)
9 Z 100 70 mm/s 185 C0 0.30 mm Diamond (F)
10 Z 100 90 mm/s 175 C0 0.30 mm Diamond (F)
11 Z 100 90 mm/s 180 C0 0.30 mm Diamond (F)
12 Z 100 90 mm/s 205 C0 0.30 mm Diamond (F)
13 Z 100 90 mm/s 185 C0 0.40 mm Diamond (F)
0
14 Z 100 90 mm/s 185 C 0.25 mm Diamond (F)
15 Z 100 90 mm/s 185 C0 0.10 mm Diamond (F)
16 Z 100 90 mm/s 185 C0 0.30 mm Diamond
17 Z 100 90 mm/s 185 C0 0.30 mm Linear
18 Z 100 90 mm/s 185 C0 0.30 mm Hexagonal
 Ala’aldin Alafaghani et al. / Procedia Manufacturing 10 (2017) 791 – 803 795

2.2. Influence of processing parameters on dimensional accuracy repeatability

To evaluate the effect of processing parameters over the dimensional accuracy and repeatability all the printed
specimens were measured and compared to designed CAD model. In total, the study included 9 measurements for
each specimen, which incorporated the overall length (OL) of the specimen, the total width (OW), the thickness (T)
and width (W) of the reduced section as shown in Figure 3. The length was measured using a Vernier calliper. The
rest of the dimensions were measured using a micrometer. The three measurements of width W1, W2 and W3 were
averaged to a single value W. The rest of measurements were averaged similarly.

Figure 4: Measurement locations of the FDM specimens

2.3. Mechanical Properties

To evaluate the mechanical properties of the FDM parts, the specimens were tested per ASTM D638-15
standards on an Instron 3369 universal testing machine. The speed of the testing was controlled using an extension
speed of 5 mm/min at room temperature. The real-time data recorded during the tests were load, extension, strain
(using an extensometer) and time. Using the recorded data, the following mechanical properties were calculated:
Young’s Modulus of elasticity, 0.1% yield stress, maximum tensile strength and the ductility. The stresses and
mechanical properties calculated are based on the actual dimensions of each specimen considering the dimensional
errors.

2.4. Finite Element Analysis

In order to make a working model for further investigation, this study establishes a new FEA approach that
utilizes FDM parts’ apparent density, as FDM parts have gaps and porosities even at 100% infill as shown in the
Scanning Electron Microscope (SEM) image shown in Figure 5(a) The specimen mass was measured by weighing
them, and their volume was measured by water displacement method. Using the mass and volume the apparent
density was calculated.
Since sample 3 was fabricated with all reference processing parameters, it was used to compare the results of the
rest of the specimens. In addition, sample 3 experimental data were used in the FEA as a reference data. In terms of
material properties, the mechanical properties of the original PLA filament, which was measured by the previous
tensile test. The summary of the measurement results for both sample 3 and PLA filament is shown in Table 3. The
filament diameter used through this study was 1.75 mm.

Table 3 Physical properties of sample 3 and PLA material

Sample Volume Mass Apparent Density Yield stress Load at Strain at yield Young’s
[cm3] [g] [g/cm3] [MPa] yield [N] [mm/mm] modulus [MPa]

Sample 3 5.05 5.91 1.15 40.09 877.9 0.01442 3144.19

PLA Filament 3.19 3.99 1.25 49.72 111.52 0.01401 3787.08
796 Ala’aldin Alafaghani et al. / Procedia Manufacturing 10 (2017) 791 – 803

To model FDM parts in FEA, the created CAD model should compensate for the voids or gaps. The difference in
densities was used to calculate the material discontinuities due to layers’ fusion in the CAD model for FEA. The
shape of discontinuities was chosen to be of a diamond shape of lengths 0.2 mm x 0.2 mm and perpendicular to each
other, which corresponds to the 45o degrees raster angle and the diamond infill pattern. Although in the actual FDM
part the size of the discontinuities is much smaller, this size was used to reduce the computation time and reduce the
need for finer meshing. In this suggested approach, the number of holes per area (HPA) is introduced using Equation
1.

U FDM V
CAD u U Filament  U Filament u l 2 u T u ACAD u HPA / VCAD (1)

Where, U FDM is the apparent density of the FDM part, VCAD is the volume of the CAD original CAD model,
U Filamentis the density of the filament (Table 3). T is the thickness of the specimen which is 3.50 mm, l is the
length of the discontinuity and equal to 0.20 mm, ACAD is the area of the specimen in the X-Y plane as shown in
Figure 2. By substituting these values in Equation 1 the HPA was found to be 1.9998 holes /mm2. Thus about 300
holes are necessary in the gage length region of 25 mm length and 6 mm width. To make the model more accurate
the length l can be reduced and the HPA will be increased, however, that needs finer mesh and will increase the
simulation time.
In the FEA, symmetry about the Y-axis was used and only the top half of the specimen was modeled with the
holes as shown in Figure (b) to reduce the computation time. Fixed boundary condition was applied to one of the
grips’ sides while a distributed load was applied to the other grip sides equal to 877.9 N, which is the load at yield
for sample 3, divided over the area of the grip sides. The model utilized the mechanical properties of the PLA
filament, which was assumed to be isotropic with a Poisson’s ratio of 0.35, and fine mesh was applied at the gage
length area with the holes. The simulated model assumes uniform properties through the layers, which is
unfortunately not the actual case.
Another FEA case was conducted using the previously described model. However, without the compensation for
layers’ fusion effect (i.e. voids). In this case, the mechanical properties given to the model were equal to the PLA
filament’s mechanical properties multiplied by the density ratio between sample 3 and the filament, which is equal
to 3787.08 multiplied by 0.92. To verify the simulations, the average strain in the gage length region calculated from
the average displacement for two sections in the gage length divided over the original distance between them.

a b

Figure 5: a) SEM of tested FDM specimen. b) Simulation CAD model with material discontinuities.
 Ala’aldin Alafaghani et al. / Procedia Manufacturing 10 (2017) 791 – 803 797

3. Results and Analysis

3.1. Dimensional Accuracy and Repeatability

The three width measurements, W1, W2, and W3, were averaged into a single width value for each sample. The
same process was done for the overall width measurements OW1 and OW2 and the overall thickness measurements
OT1, OT2, and OT3. The results of these measurements are shown in Table 4. By comparing these measured values
with the design dimensions, i.e. OL= 115.0 mm for overall length, OW=19.00 mm for overall width, w=6.00 mm
for reduced section width and T=3.50 mm for thickness, dimensional error Figure was created using error Equation
2.

Error Measured Value  Design Value (2)

Table 4 Samples measurements averaged results, dimensions in mm.

Specimen Number Overall Length Overall Width Reduced Section Width Thickness

CAD Model 115.00 19.00 6.00 3.50

1 115.08 18.98 5.95 4.12
2 114.58 18.73 6.55 3.72
3 114.51 19.29 6.44 3.93
4 114.58 19.21 6.36 3.74
5 114.51 19.24 6.28 3.70
6 114.67 19.29 6.42 3.72
7 114.59 19.24 6.35 3.74
8 114.66 19.21 6.30 3.75
9 114.76 19.27 6.31 3.71
10 114.64 19.17 6.27 3.56
11 114.61 19.35 6.33 3.72
12 114.74 19.30 6.46 3.95
13 114.68 19.48 6.48 3.84
14 114.48 19.17 6.18 3.59
15 114.99 18.81 6.27 3.56
16 114.69 19.34 6.36 3.74
17 114.66 19.29 6.43 3.71
18 114.64 19.31 6.38 3.74

From Figure 6-a it is clear that the building direction affects the dimensional accuracy significantly. The first
observation is that most of the errors have positive values, which indicates that the printer tends to create bigger
parts than designed. Although it is hard to draw patterns of how building direction affects the error in dimensions, it
is clear that the dimensional accuracy is affected by the building direction. That might be explained by the
difference in positioning accuracy of the FDM printer’s extruder, where it is listed by the manufacturer that the
positioning resolution of the extruder for the Makerbot 2X is 11 microns in the layer plane and 2.5 microns in the
building direction. In addition to that the height of the part must be an integer number of layer thicknesses, whereas
for the plane dimensions, it is more related to positioning accuracy of the extruder and the nozzle diameter. Infill
percentage and infill patterns have little to no influence on the dimensional error, as shown in Figure 6-b and 6-d.
However ,using the diamond fast infill pattern has resulted in higher error in thickness which is the most sensitive
dimension. Looking at Figure 6-c, it can be seen that the errors in all the measured dimensions were constant except
798 Ala’aldin Alafaghani et al. / Procedia Manufacturing 10 (2017) 791 – 803

at printing speed 90 mm/s. At 90 mm/s speed the thickness and the width at the reduced section had an increased
error from 0.31 to 0.44mm approximately for the reduced width and from 0.21 to 0.43 for the thickness since that
increase in error only occurred at one printing speed thus giving no clear trends. There is no clear explanation for
this phenomenon except it might be due to a special heat transfer or temperature gradient developed when printing at
that speed. However, it is worth mentioning that typical printing speeds range from 40mm/s to 80mm/s. Extrusion
temperature has a significant effect on the accuracy; as the extrusion temperature increases the error increases as
shown in Figure 6-e, which might be due to the increased fluidity of the extruded filament at higher temperature
which allows the extruded filament to flow out of the nozzle control thus increasing the error. The significance of
layer height on dimensional accuracy and resolution is well known [24]. From Figure 6-f, it can be seen that smaller
layer height in general yields lower error values. In addition, it can be seen that when the layer height was equal to
0.25 mm the error was small in the thickness dimension even though the layer height is relatively big, which can be
explained because T=3.50 mm an integer multiple of 0.25 mm. That explains the jump in error when the layer is
slightly increased to 0.30 mm. It should be mentioned that using absolute changes in the dimensions might be better
when comparing multiple dimensions at the same time as it can be seen in Figure 6 that error in the overall length
and its changes are negligible when compared to the errors of the other dimensions if the error was described as
percentage error.

Figure 6: The dimensional error [%] caused by (a) building direction, (b) infill precentage, (c) printing speed, (d) infill patterns (refer to Table 1),
(e) extrusion temperature, (f) layer height.

3.2. Mechanical properties result and analysis

For this analysis, the same approach for investigating the dimensional accuracy was followed. The sample
averaged results of the tensile testing are shown in Table 5. The indicated results were used to create Figure 7.
 Ala’aldin Alafaghani et al. / Procedia Manufacturing 10 (2017) 791 – 803 799

As expected the mechanical properties are highly affected by the processing parameters as Figure 7 shows.
Young’s modulus, tensile strength and the yield strength are lower using X building direction than using Y or Z
building directions and having maximum values using the later as can be seen from Figure 7-a. The result of the
building direction supports previous studies [25]. This indicates that the bonding or fusion between layers is weaker
than in the layers themselves. This can be interpreted by that consecutive extruded paths have better fusion due to
their mutual high temperature whereas consecutive layers usually have higher temperature gradient reducing the
adhesion between the layers and resulting in a weaker part. Although both Y and Z building direction have layers
parallel to the tension force, the difference in strength and stiffness could be because the Z building direction has
less number of layers, which interprets the higher strength in Z direction.
As expected the mechanical properties improved by increasing the infill percentage as shown in Figure 7-b.
Although, the infill percentage cannot be directly related to the weight percentage of the specimen to the fully filled
specimen due to the weight of the shell, higher infill percentages improved the mechanical properties by providing
more material to take the loads applied by the tensile machine.
Figure 7-c and 7-d show that the mechanical properties are not highly affected by the printing speed or infill
patterns. However, previous studies [26], [27] show that infill patterns affect the mechanical properties. The relative
independence of the mechanical properties from infill patterns can be explained by the fact that the parts with
varying infill patterns were printed with 100% infill percentage which is expected to reduce the effect of the
patterns. Thus, larger specimens with lower infill percentages are necessary to highlight the influence or infill
patterns.
The effect of extrusion temperature is shown in Figure 7-e. As the temperature increased the mechanical
properties increased, which is explained by the improved fusion within the extruded layer and between the layers.
Similarly, the figure shows that there is no improvement in increasing the extrusion temperature after a certain limit.

Table 5: Tensile testing results

Sample Number Young’s Modulus [MPa] Sy 0.001 offset Tensile Strength Ductility
[MPa] [MPa] [mm/mm]

1 2608.80 35.11 42.12 0.019010

2 2815.58 37.22 45.72 0.019190
3 3144.19 40.09 46.06 0.019633
4 2851.84 38.53 42.95 0.024780
5 2240.56 30.00 34.39 0.026593
6 2255.35 28.83 33.74 0.020033
7 2837.48 38.67 42.98 0.028873
8 2950.99 38.80 43.65 0.026807
9 2803.19 38.79 43.25 0.028623
10 1947.05 25.37 28.59 0.019040
11 2581.47 34.65 40.58 0.026373
12 3004.46 39.39 43.79 0.019667
13 3177.53 35.32 46.10 0.021803
14 2599.09 31.02 43.17 0.025797
15 2586.30 29.31 39.19 0.020920
16 2810.19 38.26 43.61 0.025470
17 2834.32 37.77 43.57 0.026600
18 2859.27 39.47 44.75 0.028100

In Figure 7-f the mechanical properties are shown to improve by increasing layer height. That supports that
having less layers provides stronger parts. However, the figure shows little improvement in the tensile strength and
Young’s modulus and a decline in the yield strength by increasing the layer height from 0.30 mm to 0.40 mm. For
ductility, no clear relationships can be observed. Also, the specimens in the same sample showed high repeatability
800 Ala’aldin Alafaghani et al. / Procedia Manufacturing 10 (2017) 791 – 803

in terms of yield strength, tensile strength and Young’s modulus, however, in terms of ductility they showed high
diversity.

3.3. FEA results and analysis

The results of the first FEA model with material discontinuities are shown in Figure 8. In Figure 8 stress
concentration around the holes is observed and that explains the breaking pattern near the raster lines seen in the
tested specimens. The maximum stress in the simulation 55.97 MPa is larger than the yield stress to the
experimental yield of the filament 46.90 MPa. The strain and displacement results in the gage region are close to
experiment with an average strain of 0.012715 with error 9.2434%. Figure 8 shows the results of the first simulation
with the Von Mises stresses and principle strains contours.
In the second simulation, the yield stress in the simulation 43.48 MPa was very close to the yield stress of sample
3 40.09 MPa. The average strain was 0.01167 compared to 0.01442 from the tensile test on sample 3 with error of
19.0%.
Both simulation offer comparable results. The second FEA model requires less computational time but does not
explain the breaking pattern of the tensile specimens, which can be observed in the first FEA model as shown in
Figure 9.

Figure 7: The resulting mechanical properties [MPa] caused by (a) building direction, (b) infill percentage, (c) printing speed, (d) infill patterns,
(e) extrusion temperature, and (f) layer height, where E is Young’s modulus and Sy is the yield strength and Ts is the tensile strength.
 Ala’aldin Alafaghani et al. / Procedia Manufacturing 10 (2017) 791 – 803 801

Figure 8: Von Mises stress distribution on the model with discontinuities with legends of Von Mises and principle strain.

Figure 9: Von Mises stress concentration around discontinuities.

802 Ala’aldin Alafaghani et al. / Procedia Manufacturing 10 (2017) 791 – 803

4. Conclusion

This paper examines the influence of FDM processing parameters on the final parts characteristics. The study
investigates the effect of building directions, infill percentage, infill patterns, print speed, extrusion temperatures and
layer height independently on mechanical properties and dimensional accuracy. Furthermore, the study presents a
new developed approach to model FDM parts using FEA.
It was shown that the dimensional accuracy is affected by building direction extrusion temperature and layer
height more than infill percentage, infill pattern, and printing speed. Generally, to improve the dimensional accuracy
in a certain direction for FDM parts the critical dimension should be parallel the layer orientation instead of the
building direction, in addition to using lower extrusion temperature and layer height.
It was shown that the mechanical properties are influenced significantly building direction, extrusion
temperature, and layer height; and less significantly on infill patterns, for high infill percentages specimens, and
printing speed. To improve the mechanical properties, higher extrusion temperature and larger layer height are
needed in addition to appropriate building direction, that makes the layers and the load direction in the same plane.
To highlight the significance of infill patterns, it is necessary to use larger specimens with lower infill percentages to
provide bigger room for the infill patterns.
Using density ratio between the filament and the FDM part shows promising results using both approaches,
however in both cases the effect of the layers requires further improvements.
The future direction to build on this work is to investigate the effect of infill percentages using larger specimens
where there is more infill volume than shells, investigate the effect of infill patterns with lower infill percentages,
study the effects of other processing parameters such as cooling rate and environment conditions and investigate the
combined effect of changing multiple processing parameters at multiple levels. Simulate the model using lamia
material properties instead of isotropic.

References

[1] K. J. De Laurentis and C. Mavroidis, ĀRapid fabrication of a nonϋassembly robotic hand with embedded components,ā Assem. Autom.,
vol. 24, no. 4, pp. 394̢405, Dec. 2004.
[2] D. Thomas and S. Gilvert, “Costs and Cost Effectiveness of Additive Manufacturing,” US Dep. Commer., no. December, 2014.
[3] L. M. Galantucci, F. Lavecchia, and G. Percoco, “Experimental study aiming to enhance the surface finish of fused deposition modeled
parts,” CIRP Ann. - Manuf. Technol., vol. 58, no. 1, pp. 189–192, 2009.
[4] L. M. Galantucci, F. Lavecchia, and G. Percoco, “Quantitative analysis of a chemical treatment to reduce roughness of parts fabricated using
fused deposition modeling,” CIRP Ann. - Manuf. Technol., vol. 59, no. 1, pp. 247–250, 2010.
[5] A. Garg, A. Bhattacharya, and A. Batish, “Chemical vapor treatment of ABS parts built by FDM: Analysis of surface finish and mechanical
strength,” Int. J. Adv. Manuf. Technol., pp. 1–17, 2016.
[6] V. Tiwary, P. Arunkumar, A. S. Deshpande, and V. Khorate, “Studying the effect of chemical treatment and fused deposition modelling
process parameters on surface roughness to make acrylonitrile butadiene styrene patterns for investment casting process,” Int. J. Rapid
Manuf., vol. 5, no. 3–4, pp. 276–288, 2015.
[7] P. M. Pandey, N. Venkata Reddy, and S. G. Dhande, “Improvement of surface finish by staircase machining in fused deposition modeling,”
J. Mater. Process. Technol., vol. 132, no. 1, pp. 323–331, 2003.
[8] O. Kerbrat, P. Mognol, and J. Y. Hasco??t, “A new DFM approach to combine machining and additive manufacturing,” Comput. Ind., vol.
62, no. 7, pp. 684–692, 2011.
[9] J. F. Rodríguez, J. P. Thomas, and J. E. Renaud, “Mechanical behavior of acrylonitrile butadiene styrene (ABS) fused deposition materials.
Experimental investigation,” Rapid Prototyp. J., vol. 7, no. 3, pp. 148–158, Aug. 2001.
[10] R. Anitha, S. Arunachalam, and P. Radhakrishnan, “Critical parameters influencing the quality of prototypes in fused deposition modelling,”
J. Mater. Process. Technol., vol. 118, no. 1–3, pp. 385–388, 2001.
[11] J. S. Chohan and R. Singh, “Enhancing dimensional accuracy of FDM based biomedical implant replicas by statistically controlled vapor
smoothing process,” Prog. Addit. Manuf., vol. 1, no. 1, pp. 105–113, 2016.
[12] C. C. Wang, T. Lin, and S. Hu, “Optimizing the rapid prototyping process by integrating the Taguchi method with the Gray relational
analysis,” Rapid Prototyp. J., Apr. 2013.
[13] S.-H. Ahn, M. Montero, D. Odell, S. Roundy, and P. K. Wright, “Anisotropic material properties of fused deposition modeling ABS,” Rapid
Prototyp. J., vol. 8, no. 4, pp. 248–257, 2002.
[14] K. Chin Ang, K. Fai Leong, C. Kai Chua, and M. Chandrasekaran, ĀInvestigation of the mechanical properties and porosity relationships in
fused deposition modellingϋfabricated porous structures,ā Rapid Prototyp. J., vol. 12, no. 2, pp. 100̢105, Mar. 2006.
[15] G. C. Onwubolu and F. Rayegani, “Characterization and Optimization of Mechanical Properties of ABS Parts Manufactured by the Fused
Deposition Modelling Process,” Int. J. Manuf. Eng., vol. 2014, p. 13, 2014.
 Ala’aldin Alafaghani et al. / Procedia Manufacturing 10 (2017) 791 – 803 803

[16] B. H. Lee, J. Abdullah, and Z. A. Khan, “Optimization of rapid prototyping parameters for production of flexible ABS object,” J. Mater.
Process. Technol., vol. 169, no. 1, pp. 54–61, 2005.
[17]T. Nancharaiah, “Optimization of Process Parameters in FDM Process Using Design of Experiments,” Optimize, vol. 2, no. 1, pp. 100–102,
2011.
[18]S. K. Panda, “Optimization of Fused Deposition Modelling (FDM) Process Parameters Using Bacterial Foraging Technique,” Intell. Inf.
Manag., vol. 1, no. 2, pp. 89–97, 2009.
[19] Ranjeet Kumar Sahu, S.S. Mahapatra, Anoop Kumar Sood “A Study on Dimensional Accuracy of Fused Deposition Modeling (FDM)
Processed Parts using Fuzzy Logic ,” Journal for Manufacturing Science & Production , vol. 13. p. 183, 2013.
[20] K. Thrimurthulu, P. M. Pandey, and N. Venkata Reddy, “Optimum part deposition orientation in fused deposition modeling,” Int. J. Mach.
Tools Manuf., vol. 44, no. 6, pp. 585–594, 2004.
[21] K. P. K. Vishal N. Patel, “Parametric Optimization of The Process of Fused Deposition Modeling In Rapid Prototyping Technology- A
Review,” Int. J. Innov. Res. Sci. Technol., vol. 1, no. 7, pp. 80–82, 2014.
[22] M. Domingo-Espin, J. M. Puigoriol-Forcada, A. A. Garcia-Granada, J. Llumà, S. Borros, and G. Reyes, “Mechanical property
characterization and simulation of fused deposition modeling Polycarbonate parts,” Mater. Des., vol. 83, pp. 670–677, 2015.
[23] M. Farbman, Daniel, Chris, “Materials Testing of 3D Printed ABS and PLA Samples to Guide Mechanical Design,” pp. 1–12, 2016.
[24] P. M. Pandey, N. V. Reddy, and S. G. Dhande, “Real time adaptive slicing for fused deposition modelling,” Int. J. Mach. Tools Manuf., vol.
43, no. 1, pp. 61–71, 2003.
[25] O. S. Es-Said, J. Foyos, R. Noorani, M. Mendelson, R. Marloth, and B. A. Pregger, “Effect of Layer Orientation on Mechanical Properties of
Rapid Prototyped Samples,” Mater. Manuf. Process., vol. 15, no. 1, pp. 107–122, 2000.
[26] B. M. Tymrak, M. Kreiger, and J. M. Pearce, “Mechanical properties of components fabricated with open-source 3-D printers under realistic
environmental conditions,” 2014.
[27] A. K. Sood, R. K. Ohdar, and S. S. Mahapatra, “Parametric appraisal of mechanical property of fused deposition modelling processed parts,”
Materials and Design, vol. 31, no. 1. pp. 287–295, 2010.