IOP Conference Series: Materials Science and Engineering

PAPER • OPEN ACCESS

Research on Shape and Dimensional Accuracy of FDM Produced Parts

To cite this article: J Beniak et al 2019 IOP Conf. Ser.: Mater. Sci. Eng. 501 012030

View the article online for updates and enhancements.

This content was downloaded from IP address 191.101.147.110 on 10/04/2019 at 01:59

9th TSME-International Conference on Mechanical Engineering (TSME-ICoME 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 501 (2019) 012030 doi:10.1088/1757-899X/501/1/012030

Research on Shape and Dimensional Accuracy of FDM

Produced Parts

J Beniak1*, PKrižan1, Ľ Šooš1, and M Matúš 1

1
Slovak University of Technology in Bratislava, Faculty of Mechanical Engineering,
Nam. Slobody 17, 812 31 Bratislava, Slovak Republic

Abstract. Production of technical parts for Mechanical Engineering field is very important.
There are many conditions which the produced parts have to satisfy. Very important for
functional parts are accuracy of final parts. This accuracy could be evaluated in different ways.
For example it can be evaluated shape accuracy or dimensional accuracy. This paper is focused
to Fused Deposition Modelling (FDM) technology which is used for Additive Manufacturing
process. The parts in additive manufacturing are produced directly from virtual model and digital
file which could be obtained by direct modelling with CAD software of by 3D scanning.
Published paper deals realized research where the full factor experiment was prepared and
measured shape and dimensional accuracy of produced specimens. There are designed two types
of specimens with different shapes and dimensions. Digital models are processed with different
different 3D printer settings. Measured values are evaluated by statistical methods.

* Corresponding Author: e-mail: juraj.beniak@stuba.sk; tel. +421 905 593 953

1. Introduction
For technical practice is very important question of production precision of the machine parts in view
of their intended use. The use of additive production of additive manufacturing systems increase from
year to year in the forefront also in the production of parts and components for final use. That is why in
this area it is necessary to know the technology so much that it is possible to influence the preparation
and process of production of components through additive manufacturing systems in order to ensure the
required accuracy of the final part [1]. In the case of mechanical engineering practice there have to be
taken into the consideration dimensional accuracy of produced parts, what is the most common, but it
have to be counted with shape and positional tolerances. It is necessary to make sure that the produced
shapes have defined positions between important entities.
There are presented in this this paper the experiments focused to dimensional and shape and
positional precision measurement [8]. There are many ways how could be this precision measured. In
this paper is presented the measuring process with use of optical measuring instrument combined with
special software which is directly evaluate specified tolerance. It is new way how to make the
measurement process more effective and make it easier.
This tolerance evaluation is also connected with the specified factors which are selected as possible
factors which could influence accuracy of parts produced by additive production techniques. In this case
is the production process focused to FDM (Fused Deposition Modeling) technology which work with
semi melted thermoplastic material which is deposited layer by layer to build required 3D part.

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
9th TSME-International Conference on Mechanical Engineering (TSME-ICoME 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 501 (2019) 012030 doi:10.1088/1757-899X/501/1/012030

2. Material and Devices

The main aim of this research is to specify what are the shape and positional tolerances and also
dimensional accuracy produced specimens by FDM technology. First of all there were specified the
factors which could be possible for influencing of produced parts. There are many factors which could
work with tolerance influence during the 3D printing [2, 3]. After experiences and literature research
we specified two main used factors.
Temperature necessary for plastic material melting and layer thickness. Each producer of filaments
state just range of suitable temperatures, not the exact temperature. This is because different 3D printers
work with different software and have different technical design, so there could be necessary different
temperature setting, but also settings of many others parameters. The temperature affect the consistence
or the flow of semi melted thermoplastic what will affect the diffluence of deposited fibres.
Layer thickness is affecting the dimensional accuracy, so we try to figure out if there is also influence
to shape and positional tolerances [4]. Each 3D printer allows to set layer thickness and it is easy to
change and affect the final quality of produced parts.
Each of this two factors have two levels. The specified values are set to marginal possible values.
For PLA plastic, which have been selected as very easy used production material, the producer state the
printing temperature from 180°C to 220°C. The layer thickness is set to 0,1 and 0,2mm.
Using of this selected factors and their specified levels, there have been prepared full factor
experiment, so combination of each levels each factors together. Respecting Design of Experiment
(DOE) there is prepared full factor experiment (Table 1).
From each of this combination are produced 5 specimens to ensure data samples for measurement
and proper statistical evaluation. Each specimen or each measurement have been repeated 5 times on
each specimen. So we collected 25 values for each measured parameter, what is really significant set of
values.
Table 1. Design of Experiment (DOE)

Specimen Layer Thickness Printing Temperature

(Experiment) (mm) (°C)
1 0,1 220
2 0,2 220
3 0,1 180
4 0,2 180

The specimens are presented on Figure 1. There are two types of designed models, just to measure the
both rounded and angular shapes, to test the possibility of selected 3D printer. As it is visible, the objects
have also the holes with appropriate shape. On the Figure 1 are also presented the dimensions of each
model. The printed height of models is the same with value 40mm.

Figure 1. Designed 3D models selected for tolerance measurement.

Preparation of specimens are realized on FDM 3D printer Profi3DMaker (Figure 2 left) with following
conditions:

2
9th TSME-International Conference on Mechanical Engineering (TSME-ICoME 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 501 (2019) 012030 doi:10.1088/1757-899X/501/1/012030

 material of specimens – PLA  density of top layers – 100%

 printing speed – 50 mm.s-1  number of wall contours – 2
 nozzle diameter – 0,3 mm  infill density – 5%
 number of bottom layers – 3  infill structure – rectangular grid
 density of bottom layers – 100%  temperature of built platform - 50°C
 number of top layers – 3
Produced parts are subsequently measured. As is mentioned above, for measurement is selected
sophisticated optical measuring device SmartScope MVP 200 (Figure 2 right).

Figure 2. Profi3Dmaker 3D printer (left) [6] , Smartscope MVP 200 measuring instrument (right) [7],

Device MVP 200 work with measuring software Measure-X. This software automatically evaluate the
final tolerance, it is just necessary to specify the appropriate lines or surfaces for correct measurement.
The following Figure 3 and Figure 4 present labeling of measured tolerances. The dimension (for
example “a” means the width of tolerance zone in which are the measured points.

Figure 3. Measured precision for trapezium.

3
9th TSME-International Conference on Mechanical Engineering (TSME-ICoME 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 501 (2019) 012030 doi:10.1088/1757-899X/501/1/012030

Figure 4. Measured precision for cylinder.

3. Measurement and evaluation of dimensional precision

The whole measurement is divided to measurement of dimensional precision and shape precision. In the
case of trapezium there are measured dimension “m” and dimension “n” Measured values are presented
in Table 2. Average values are calculated based on the equation (1). For statistical evaluation are also
necessary at least standard deviation calculated by equation (2). Difference between theoretical designed
value and average measured value are presented also on the Figure 5.

Table 2. Average values of trapezium dimensions

Exp. Dimension Difference m SD Dimension n Difference SD

m (mm) (mm) (mm) (mm) n (mm) (mm)
1 49,86 -0,14 0,0749 39,87 -0,27 0,01573
2 49,955 -0,045 0,3991 39,96 -0,039 0,06515
3 49,8 -0,2 0,0618 39,85 -0,35 0,03554
4 49,96 -0,04 0,03977 39,83 -0,021 0,02344

The following is the formal mathematical formula for the arithmetic mean (a fancy name for the
1
average). 𝑋̅ = 𝑛 ∗ ∑𝑛𝑖=1 𝑥𝑖 (1)

𝑋̅ = average (or arithmetic mean)

n = the number of terms (e.g., the number of items or numbers being averaged)
xi = the value of each individual item in the list of numbers being averaged
Standard deviation values are calculated based on following equation:
∑(𝑥𝑖 −𝑋̅)2
𝑆𝐷 = √ 𝑛−1
(2)

4
9th TSME-International Conference on Mechanical Engineering (TSME-ICoME 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 501 (2019) 012030 doi:10.1088/1757-899X/501/1/012030

Measured values are also evaluated by ANOVA statistical method [5]. The results from ANOVA
analysis are as follow:
Dimension “m”
Thickness F(1,96) = 6,75 p<0,010847 SS=0,02 MSe=0,01
Temperature F(1,96) = 119 p<0,000001 SS=0,38 MSe=0,01
Thickness*Temperature F(1,96) = 6,48 p<0,012475 SS=0,02 MSe=0,01
Dimension “n”
Thickness F(1,96) = 24,1 p<0,000004 SS=0,04 MSe=0,01
Temperature F(1,96) = 85,2 p<0,000001 SS=0,13 MSe=0,01
Thickness*Temperature F(1,96) = 51,4 p<0,000001 SS=0,08 MSe=0,01

Figure 5. Measured difference of dimensions m and n.

Measured values from cylindrical specimens are presented in Table 3. Difference between theoretical
designed value and average measured value are presented also on the Figure 6.

Table 3. Average values of cylinder dimensions

Exp. Diameter Difference SD Diameter ɸ5 Difference SD

ɸ4 (mm) ɸ4 (mm) (mm) (mm) ɸ5 (mm) (mm)
1 40,27 +0,27 0,1675 24,44 -0,44 0,1363
2 40,25 +0,25 0,1336 24,5 -0,5 0,1841
3 39,85 -0,15 0,0433 24,82 -0,18 0,0209
4 39,84 -0,16 0,0265 24,87 -0,13 0,0263

Measured values are also evaluated by ANOVA statistical method. The results from ANOVA analysis
are as follow:
Dimension “ɸ4”
Thickness F(1,96) = 0,020; p<0,888327; SS=0,01; MSe=0,01
Temperature F(1,96) = 364 p<0,000001; SS=4,41; MSe=0,01
Thickness*Temperature F(1,96) = 0,286; p<0,593873; SS=0,01; MSe=0,01
Dimension “ɸ5”
Thickness F(1,96) = 0,002 p<0,965637; SS=0,01; MSe=0,01
Temperature F(1,96) = 264 p<0,000001; SS=3,54; MSe=0,01
Thickness*Temperature F(1,96) = 5,69 p<0,018994; SS=0,08; MSe=0,01

5
9th TSME-International Conference on Mechanical Engineering (TSME-ICoME 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 501 (2019) 012030 doi:10.1088/1757-899X/501/1/012030

Figure 6. Measured difference of dimensions ɸ4 and ɸ5..

From this results we can see that the temperature as one of the factors is significant for dimensional
precision. This is clear from the ANOVA analysis.

4. Measurement and evaluation of shape and positional tolerances

Based on the specified shape and positional tolerances from Figures 3 and 4, there was measured this
tolerances. In the Table 4 are average values of tolerances specified on trapezium specimens. Average
values of tolerances specified on trapezium specimens are in the Table 5.

Table 4. Average values of tolerances for trapezium specimens

(Exp) Straight Straight Paralell Paralell Inclin Inclin Perpend Flatness

a (mm) c (mm) i(B) j(B) f(A) h(B) p(B) l (mm)
(mm) (mm) (mm) (mm) (mm)
1 0,042 0,022 0,047 0,037 0,071 0,12 0,062 0,021
2 0,062 0,039 0,051 0,063 0,076 0,12 0,061 0,038
3 0,041 0,022 0,023 0,037 0,029 0,077 0,028 0,025
4 0,062 0,027 0,029 0,036 0,054 0,11 0,045 0,049

Table 5. Average values of tolerances for cylindrical specimens

(Exp) Straight Paralell Paralell Perpend Perpend Round Round Concentric

a (mm) b(A) d(A) c(D) c(C) ɸ1 ɸ2 ɸ3(B)
(mm) (mm) (mm) (mm) (mm) (mm) (mm)
1 0,065 0,11 0,19 0,13 0,11 0,29 0,21 0,15
2 0,084 0,096 0,19 0,21 0,14 0,15 0,24 0,14
3 0,05 0,1 0,22 0,14 0,13 0,16 0,074 0,11
4 0,084 0,13 0,2 0,2 0,14 0,092 0,094 0,058

From this measured values stated in Table 4 and Table 5 we can see that statistically the worst results
are reached from the combination of factors temperature and layer thickness for experiment number 2
(0,2mm layer thickness and 220°C printing temperature) for trapezium specimens. We can also see that
the main change of tolerance is caused by change of temperature. It is possible to explain that the

6
9th TSME-International Conference on Mechanical Engineering (TSME-ICoME 2018) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 501 (2019) 012030 doi:10.1088/1757-899X/501/1/012030

deposited material is diffluent more to the sides and this cause bigger inaccuracy. This statement is also
based on ANOVA statistical analysis, which had similar results as in case of dimensional accuracy
measurement.

5. Conclusion
Based on the presented average vales from tolerance and dimension measurement and based on the
ANOVA analysis it is possible to state that the printing temperature have the significant influence for
shape and dimensional tolerance of FDM produced parts. This is because with higher temperature the
melted material is more liquid with easy flow. So the material flow to the side caused by higher liquid
and gravity, so the dimensions and tolerances are more inaccurate compare to specimens produced by
lower temperature. This statement is based also on the optical comparison of specimens. The specimens
with high printing temperature have worst surface quality what also lead to the worst tolerances.

Acknowledgments
The paper is a part of the research done within the project KEGA 051STU-4/2018 – Innovative methods
for education process improving in the field of additive manufacturing and computer-aided systems.

References
[1] Chua C K, Leong K F, Lim C S, Rapid Prototyping, Principles and Applications, Nanyang
Technological University, Singapore, World Scientific Co. Pte. Ltd, 2003, ISBN 981-238-117-1.
[2] DUC PHAM OBE FREngAND STEFAN DIMOV, Rapid Prototyping, A time compression tool,
Manufacturing Engineering Centre, Cardiff University.
[3] Beniak J, Križan P, Matúš M, Svátek M, Ecological PLA plastic used for FDM Rapid Prototyping
Technology. In SGEM 2015. 15th International Multidisciplinary Scientific Geoconference :
conference proceedings. Albena, Bulharsko, 18.-24. 6. 2015. 1. vyd. Sofia : STEF92 Technology,
2015, S. 117-123. ISSN 1314-2704. ISBN 978-619-7105-34-6.
[4] Anoop K. Sood, Raj K. Ohdar, Siba S. Mahapatra, Experimental investigation and empirical
modelling of FDM process for compressive strength improvement, In Journal of Advanced
Research, Volume 3, Issue 1, 2012, Pages 81-90, ISSN 2090-1232,
https://doi.org/10.1016/j.jare.2011.05.001.
[5] Mendonsa C, Naveen KV, Upadhyaya P, Shenoy V D, Influence of FDM Process Parameters on
Build Time Using Taguchi and ANOVA Approach, International Journal of Science and
Research (IJSR), ISSN (Online): 2319-7064
[6] Company information. Smartscope MVP 200. Available at: www.smartscope.com
[7] Printing Industry. Profi3DMaker. Available at: https://3dprintingindustry.com
[8] Svetlík J, Demeč P., Methods of Identifying the Workspace of Modular Serial Kinematic
Structures. In: Applied Mechanics and Materials : CECOL 2012: 3rd Central European
Conference on Logistics : Trnava, Slovakia, 28-30 November 2012. Vol. 309 (2013), p. 75-79. -
ISBN 978-303785636-9 - ISSN 1662-9336