INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING Vol. 19, No. 1, pp.

137-142 JANUARY 2018 / 137

REGULAR PAPER DOI: 10.1007/s12541-018-0016-0

ISSN 2234-7593 (Print) / 2005-4602 (Online)

Effect of Fabrication Parameters on Surface Roughness

of FDM Parts

Min Kyung Kim1, In Hwan Lee2,#, and Ho-Chan Kim3

1 T&R Biofab Co., Ltd., 237, Sangidaehak-ro, Siheung-si, Gyeonggi-do, 15073, Republic of Korea
2 School of Mechanical Engineering, Chungbuk National University, 1, Chungdae-ro, Seowon-gu, Cheongju-si, Chungcheongbuk-do, 28644, Republic of Korea
3 Department of Mechanical and Automotive Engineering, Andong National University, 1375, Gyeongdong-ro, Andong-si, Gyeongsangbuk-do, 36729, Republic of Korea
# Corresponding Author / E-mail: anxanx@chungbuk.ac.kr, TEL: +82-43-261-3161

KEYWORDS: 3D Printing, FDM, Surface roughness, Extrusion, 3-dimensional circuit device

The FDM (Fused Deposition Modeling) technology is widely used due to its low process cost and good mechanical properties.
However, fabricated parts have relatively inferior surface roughness compared to liquid material type process such as SL
(Stereolithography). In this research, effects of fabrication parameters such as the gap between nozzle and substrate, inflow speed of
filament material and heating moving speed of nozzle on the FDM-fabricated line figuration was investigated experimentally. The
extruded line figurations such as width, thickness and cross-sectional shapes were examined. An empirical formula of the line
fabrication for fabrication parameters was made based on the experimental results. Moreover, effect of line fabrication distance on
the surface roughness was studied.

Manuscript received: April 15, 2016 / Revised: September 22, 2017 / Accepted: November 22, 2017
This paper was presented at ISGMA2016

1. Introduction inexpensive equipment and a simple process. But its surface roughness
is improper as a substrate for liquid conductive material dispensing.
Additive manufacturing (or 3D printing) technology is classified The staircase at the vertical plane is mainly influence on the surface
according to its material type or process. Material type is classified as roughness due to the layer-by-layer manufacturing of fabrication
solid, liquid and powder.1 Among solid material type additive system. Moreover, in FDM system, the layer thickness could not be
manufacturing processes, the FDM (Fused Deposition Modeling) small (less then tens of micrometers) because the diameter of extruded
process is widely used due to its low process cost and good mechanical filament material has hundreds of micrometers. Hence the staircase at
properties. However, fabricated parts have relatively inferior surface the vertical surface is remarkable. In this regards, various studies on the
roughness compared to liquid material type process such as vertical surface rough improvement for the FDM process were
Stereolithography (SL). reported.10-17 However, compare to other additive manufacturing
A multi-material Additive manufacturing technology is being focused processes, the surface roughness at the layer surface (normal to the
as next-generation Additive manufacturing technology.2-5 Especially, 3- layering direction) of FDM fabricated part is relatively high, too. But
dimensional circuit device fabrication (3DCD) technology based on there were little studies which focused on the surface roughness at layer
additive manufacturing and liquid material dispensing is expected for surface.
manufacturing an electronic device without PCB. In 3DCD technology, In this research, effects of fabrication parameters such as the gap
the 3D printed part takes roles as structure of an electronic device as well between nozzle and substrate, inflow speed of filament material and
as electrical insulation material. Mainly two additive manufacturing moving speed of heating nozzle on the FDM-fabricated line figuration
processes, SL and FDM, were used for the 3DCD technology.6-9 was investigated experimentally. The extruded line figurations such as
The surface roughness is adequate to dispense liquid conductive width, thickness and cross-sectional shapes were examined. An
material for circuits in the SL-based 3DCD technology. However, this empirical formula of the line fabrication for fabrication parameters was
technology requires expensive equipment and a complex process. The made based on the experimental results. Moreover, effect of line
FDM-based 3DCD technology, on the other hand, requires relatively fabrication distance on the surface roughness was studied, too.

© KSPE and Springer 2018

 138 / JANUARY 2018 INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING Vol. 19, No. 1

Fig. 2 Photograph of FDM head (a) and dispensing head (b)

Fig. 1 FDM-based multi-material additive manufacturing system

2. Experimental Setups

The FDM-based multi-material additive manufacturing system was

developed to examine whether the fabrication parameters has an effect
on the surface roughness of the fabricated structure as shown in Fig. 1.
Liquid material dispensing system was also included as well as an Fig. 3 Line fabrication conditions in a FDM process
FDM extrusion system for the multi-material additive manufacturing
system. The aim of our research is to fabricate the 3-dimensional circuit
device based on the additive manufacturing technology. In this regards barrel. Therefore, liquid material could be dispensed through the
we integrate the FDM head for the fabrication of the structure and nozzle. Fig. 2(b) is a photograph of the dispensing head.
dispensing head for the fabrication of conductive lines. The brief
conductive line fabrication result is discussed at the end of this paper.
Further research on the 3-dimensional circuit device fabrication will be 3. Line Fabrication Characteristics of FDM System
preceded by our group. The head which attached on the z-axis was
composed of FDM head and liquid material dispensing head. The z- 3.1 Gap between nozzle and substrate
axis state was attached on the x-y stage system. The stages, FDM head In the FDM process, the molten polymer material which is extruded
and direct writing head were controlled by PC. X1, X2 and Y axis through FDM nozzle is pressed between the substrate and nozzle tip.
stages were connected to SGDV-5R5A05A (Yaskawa Co.) motion As shown in Fig. 3, the extruded-and-pressed material is cooled and
controller and Z axis stage was connected to MR-J3-10A (Mitsubish solid line of width (w) and thickness (t) is fabricated. Moreover, surface
Co,) motion controller. Each motion controllers received control signals is fabricated by placing the extruded-and-pressed lines with certain
from the control board (LX504, Comizoa Co.) which was installed in spacing. Therefore, the cross-sectional shape of a line effects on the
the PC. All the motions were programmed by G-code. roughness of a surface. In this regards, it is essential that to examine the
The FDM head (TPC Mechatronics Co., Fig. 2(a)) consisted of line fabrication characteristics for the fabrication conditions. We
stepping motor and nozzle. Control commends for FDM head such as assumed that the inflow speed of the filament material (Q), moving
stepping motor speed, nozzle temperature as well as heating bed speed of the nozzle (V) and gap between the substrate and nozzle (H)
temperature from the PC were sent to an Arduino Mega 2560 board are major fabrication conditions which affect the extruded line width
(Stark Robotics Co.). Then the FDM control board (Ramps 1.4, Big (w) and thickness (t).
Tree Tech. Co.) which is an open source control board received signals The gap between the substrate and the nozzle (H) and the fabricated
from Arduino Mega 2560 board and generated control voltages for the line thickness (t) can’t be identical. It is because, if the gap between the
FDM head and heating bed. Moreover, temperatures of the heating bed substrate and nozzle is very small, the extruded polymer material is
and FDM head were controlled by a Ramp 1.4 board which detects compressed between the nozzle and substrate and expanded when the
temperatures through thermistors. nozzle passes by. In this regards, experiments were conducted to
The dispensing head was controlled by pneumatic dispenser examine the thickness and width as well as the cross-sectional shape
(Accura8-DX, Iwashita Engineering Inc.) which received control according to the varying gap between the substrate and the nozzle.
signals from control board (LX-504). The pneumatic dispenser controls The moving speed of the nozzle and the inflow speed of the filament
air pressure from air compressor (KAC-25, Keyang Electronic material were set as 750 mm/min and 200 mm/min, respectively. The
Machinery Co.). The air pressure sent to barrel through silicone tube temperature of nozzle and substrate were 230oC and 65oC, respectively.
(AA10n, Iwashita Engineering Inc.). Nozzle was attached at the end of The diameter of PLA filament was 1.75 mm (White, TPC Mechatronics).
 INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING Vol. 19, No. 1 JANUARY 2018 / 139

Fig. 6 Extruded-pressed line widths for various inflow speed and

nozzle speed

Fig. 4 Photographs of cross-sectional image of a line observed by

optical microscope

Fig. 5 Microscopic image of line width at gap between nozzle and Fig. 7 Schematic drawing of extruded-and-compressed material
substrate of 0.05 mm between nozzle and substrate

The inner diameter and outer diameter of the nozzle were 0.4 mm and 0.1 mm. Fig. 6 is measured widths of fabricated lines for various inflow
0.8 mm, respectively. The thickness and width as well as cross- speeds and nozzle speeds. The extruded-and-pressed line widths
sectional shape were observed in x-z plane in Fig. 3 by changing the decreased as the inflow speed decreased and nozzle speed increased.
gap between the substrate and the nozzle as from 0.1~1.0 mm. The minimum line width of 348 µm was observed at inflow speed and
Fig. 4 is photographs of the cross-sectional image observed by nozzle speed of 20 mm/min and 1500 mm/min, respectively. However,
optical microscope (OSM-U. Dong Won Co.) for various gaps between as seen in Fig. 7, it is possible that the extruded line thickness will be
the substrate and nozzle (H). Maximum thickness and widths were non-uniform if the width of extruded filament material is larger than the
measured using image processing software (IT PLUS-4.0, SomeTech outer diameter of the nozzle. In this regard, we set a valid fabrication
Vision Inc.). As seen in Fig. 4, the width was decreased and the thickness condition range as shown in Fig. 6 where the width is less than nozzle
was increased as the gap between nozzle and substrate increased. diameter of 0.8 mm.
Moreover, when observing the cross-sectional shape, the thickness was Fig. 8 is line width and thickness for various nozzle moving speeds
even as the gap between nozzle and substrate was decreased. From in valid condition ranges. The inflow speed was set as 20 mm/min, and
these results, we can conclude that the thickness tends to even over the the distance between nozzle and substrate was 0.1 mm. As seen in Fig.
cross-section as the gap between nozzle and substrate is small. That is, 8, the thickness was almost identical of 120 µm at any nozzle moving
a line with even thickness can help to fabricate a surface having good speed although the width changed according to the nozzle moving speed.
surface roughness. However, the width of the line is not uniform However, the thickness was larger than the gap between nozzle and
(458~558 µm) when the gap between nozzle and substrate is below 0.1 substrate of 100 µm. It is because the material suffered compress load
mm as shown in Fig. 5. It was because the material extruded improperly between nozzle and substrate and the load was released when the nozzle
through the narrow gap between nozzle and substrate. passed by. This phenomenon is known as Swell.18 In this regard, the layer
thickness can be different from the gap between nozzle and substrate.
3.2 Width and thickness of the extruded polymer material From these results, we can conclude that the line width could be
Further experiments were conducted to examine the thickness and changed during FDM process with identical hardware. That is, lines of
width of the fabricated line according to the varying inflow speed (Q) wide and narrow width can be fabricated as needed in a process.
and nozzle speed (V). Lines were fabricated changing the inflow speed Consequently, the fabrication speed, as well as the resolution, can be
and nozzle speed as 20~200 mm/min and 150~1500 mm/min, increased. In this regard, we generated a formula with fabrication
respectively. While the gap between nozzle and substrate was fixed as conditions as well as line widths using Matlab R2011b. In Curve Fitting
 140 / JANUARY 2018 INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING Vol. 19, No. 1

Fig. 8 Line width and thickness in valid condition range for nozzle
moving speed
Fig. 10 Schematic drawing of the surface fabrication in the FDM
system

Fig. 9 3-dimensional graph of derived polynomial

Fig. 11 Surface thickness for the fabrication distance change

Tool of Matlab, an n-th polynomial is expressed as Eq. (1).
n+1
n–i–i
y= ∑ pi x (1) were 20 mm/min and 1500 mm/min, respectively. The gap between
i=1
nozzle and substrate was set as 0.1 mm. The line width was 348 µm by
Where, n is order of polynomial and p is coefficient. The polynomial these fabrication conditions as represented in Fig. 8.
of maximum R-square value can be induced Using Eq. (1). The nozzle Cross-sectional shape was investigated to determine the thickness
moving speed (V) and inflow speed (Q) were used as input values variation using an optical microscope and image processing. Fig. 11
(independent variables) while the gap between nozzle and substrate was summarizes deviations of measured surface thicknesses for the fabrication
fixed as 0.1 mm. The dependent variable was width of the extruded- distance change. As seen in Fig. 11, the deviations decreased as the
pressed filament material. Eq. (2) is derived equations for the width of fabrication distance increased. That is, as the fabrication distance
filament material (f(Q, V)) as function of inflow speed (Q) and nozzle increases, a better surface roughness can achieved. Furthermore, the
speed (V). Therefore, we can change the fabricated line width in a best surface roughness (2 µm of deviation) was measured at fabrication
process by changing the inflow speed and nozzle speed. Fig. 9 is a 3- distance of 0.35 mm. Moreover, as the fabrication distance is larger than
dimensional graph of derived polynomial. line width, for example at 0.36 mm, the surface reveals discontinuity.
2 In this regard, we can conclude that a surface of best roughness will be
f ( Q, V ) = 774.8 + 22.94Q – 0.8841V – 0.09726Q
(2) fabricated when the fabrication distance is same as the line width.
2
– 0.004035QV + 0.0002639V Fig. 12 is cross-sectional shape comparisons of commercialized 3D
printers with our experiment result (δ = 0.35 mm). Dimension 1200es
sst (Stratasys Inc.), FB-9600 (TPC Mechatronics) and 3DP-110F
4. Surface Fabrication and Application (HyVISION Systems) were used as commercialized 3D printers. All
the fabrications, commercialized systems as well as experiments, were
Lines are fabricated side-by-side to form a surface in the FDM with same inner diameter of nozzle (0.4 mm) and filament diameter
system. In this regard, as seen in Fig. 10, it is conjectured that the (1.75 mm). Thickness deviations of Dimension 1200es sst and 3DP-
fabrication distance (δ) between lines is mainly influenced on the 110F were 24 µm and 23 µm, respectively. Rapid peak and valleys were
surface roughness. To examine the effect of fabrication distance on the observed in both specimens. Moreover, for FB-9600, discontinuities
surfaces roughness, flat surfaces were fabricated by changing fabrication were observed.
distances from 0.3 to 0.35 mm. The inflow speed and nozzle speed Fig. 13 shows the result of measuring the surface roughness using
 INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING Vol. 19, No. 1 JANUARY 2018 / 141

Fig. 12 Cross-sectional shape comparisons of commercialized 3D

printers with experiment result

Fig. 14 Conductive materials dispensing on FDM surfaces

means dispensing and extrusion direction is rectangular. The fabrication

conditions for FDM surface were the same as for those of Fig. 12. The
conductive material (ELCOAT-P100, Cans Co.) dispensed at the speed
of 10 mm/sec. As seen in Fig. 14, dispensed line widths on improved
surface are almost identical in all directions. However, for Dimension
1200es sst, there is almost no deviation in width in 0° direction. But,
in width in 90° direction, width was drastically changed. In this regard,
an electrical circuit can be fabricated in any direction on improved
surface.

Fig. 13 Surface roughness measurement results

5. Conclusion

a laser surface measuring device (Nanoview, NanoSystem Co., Ltd.). In this research, the FDM-based multi-material additive manufacturing
As can be seen in Fig. 13, some part of the specimens made with (AM) system was developed. Line fabrication experiments were
commercially available 3D printers have surface features that are conducted using FDM head to examine whether the fabrication
outside the measurement range of the measuring device (dotted circles). parameters has an effect on the surface roughness of the fabricated
Therefore, the appropriate Ra value could not be obtained. However, the structure. From the experimental research, it was found that the
surface shape of the experimental result was within the measurement thickness of the line tends to even over the cross-section as the gap
range of the device and the Ra value was 1.38 µm. From these results, between nozzle and substrate is small. The line widths decreased as the
it can be concluded that a better surface roughness can be achieved inflow speed and nozzle speed increased. Moreover, the line thickness
only adjusting fabrication parameters such as the gap between nozzle will be non-uniform if the width of extruded line is larger than the outer
and substrate, inflow speed of filament and nozzle speed without diameter of the nozzle. The thickness of the line is larger than the gap
changing hardware configurations. between nozzle and substrate due to the residual stress in the extruded
A multi-material 3D printing technology is being focused as next- line. Based on these results, an empirical formula for the line fabrication
generation 3D printing technology. Especially, 3-dimensional circuit was made. A surface of best roughness will be fabricated when the
device fabrication (3DCD) technology based on 3D printing and liquid fabrication distance is same as the line width. As an application, an
material dispensing is expected for manufacturing electronic devices electrically conductive liquid material was uniformly dispensed on the
without PCB.5-8 In this regard, liquid conductive material, which takes improved surface, successfully.
the role of electric connection among electronic elements, was dispensed
on improved surface. Moreover, for comparison, same conductive
material was dispensed on a surface fabricated using a commercialized ACKNOWLEDGEMENT
FDM system (Dimension 1200es), too.
The conductive materials were dispensed in two directions as seen This study was supported by the Development of advanced 3D
in Fig. 14. The 0° direction means that conductive material dispensing printing technology for the realistic artificial hand funded by KIST-
and polymer material extrusion direction are the same. The 90° direction ETRI projects. This study was also supported by the National Research
 142 / JANUARY 2018 INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING Vol. 19, No. 1

Foundation of Korea Grant funded by the Korean Government (NRF- Journal of Materials Processing Technology, Vol. 132, No. 1, pp.
2016R1D1A1B03936016). 323-331, 2003.

13. Galantucci, L. M., Lavecchia, F., and Percoco, G., “Experimental

Study Aiming to Enhance the Surface Finish of Fused Deposition
REFERENCES
Modeled Parts,” CIRP Annals-Manufacturing Technology, Vol. 58,
No. 1, pp. 189-192, 2009.
1. Chua, C. K., Leong, K. F., and Lim, C. S., “Rapid Prototyping:
Principles and Applications,” World Scientific, 3rd Ed., 2010. 14. Ahn, D., Kweon, J.-H., Kwon, S., Song, J., and Lee, S.,
“Representation of Surface Roughness in Fused Deposition
2. Lee, J., Kim, H.-C., Choi, J.-W., and Lee, I. H., “A Review on 3D
Modeling,” Journal of Materials Processing Technology, Vol. 209,
Printed Smart Devices for 4D Printing,” Int. J. Precis. Eng. Manuf.,
No. 15, pp. 5593-5600, 2009.
Vol. 4, No. 3, pp. 373-383, 2017.
15. Rao, A. S., Dharap, M. A., Venkatesh, J. V., and Ojha, D.,
3. Choi, S. M., Jian, X., Park, I. B., and Lee, S. H., “Development of
“Investigation of Post Processing Techniques to Reduce the Surface
Hybrid-FDM Process Using Automatic Tool Changer for Multi-
Roughness of Fused Deposition Modeled Parts,” International Journal
Material Production and Post-Processing,” J. Korean Soc. Precis.
of Mechanical Engineering and Technology, Vol. 3, No. 3, pp. 531-
Eng., Vol. 33, No. 3, pp. 235-242, 2016.
544, 2012.
4. Kim, M.-S., Chu, W.-S., Kim, Y.-M., Avila, A. P. G., and Ahn, S.-H.,
16. Vahabli, E. and Rahmati, S., “Application of an RBF Neural Network
“Direct Metal Printing of 3A Electrical Circuit Using Rapid
for FDM Parts’ Surface Roughness Prediction for Enhancing Surface
Prototyping,” Int. J. Precis. Eng. Manuf., Vol. 10, No. 5, pp. 147-
Quality,” Int. J. Precis. Eng. Manuf., Vol. 17, No. 12, pp. 1589-1603,
150, 2009.
2016.
5. Bai, Y.-Q., Kim, M.-K., Lee, I. H., and Cho, H. Y., “Fabrication of a
17. Jo, K.-H., Jeong, Y.-S., Lee, J.-H., and Lee, S.-H., “A Study of Post-
Planar Spiral Antenna Using Direct Writing Technology,” Journal of
Processing Methods for Improving the Tightness of a Part Fabricated
Mechanical Science and Technology, Vol. 29, No. 6, pp. 2461-2465,
by Fused Deposition Modeling,” Int. J. Precis. Eng. Manuf., Vol. 17,
2015.
No. 11, pp. 1541-1546, 2016.
6. Lopes, A., Navarrete, M., Medina, F., Palmer, J., MacDonald, E.,
18. Kostic, M. M. and Reifschneider, L. G., “Design of Extrusion Dies”
and Wicker, R., “Expanding Rapid Prototyping for Electronic
Encylopedia of Chemical Processing, Vol. 10 pp. 633-649, 2006.
Systems Integration of Arbitrary Form,” Proc. of 17th Annual Solid
Freeform Fabrication Symposium, pp. 14-16, 2006.

7. Jang, S. H., Oh, S. T., Lee, I. H., Kim, H.-C., and Cho, H. Y., “3-
Dimensional Circuit Device Fabrication Process Using
Stereolithography and Direct Writing,” Int. J. Precis. Eng. Manuf.,
Vol. 16, No. 7, pp. 1361-1367, 2015.

8. Espalin, D., Muse, D. W., MacDonald, E., and Wicker, R. B., “3A
Printing Multifunctionality: Structures with Electronics,” The
International Journal of Advanced Manufacturing Technology, Vol.
72, No. 5-8, pp. 963-978, 2014.

9. Lee, I. H., Jang, S. H., Oh, S. T., Woo, S. G., and Kim, M. K.,
“Fabrication of 3-Dimensional Circuit Device based on 3D Printing
Technology,” Proc. of Korean Society of Mechanical Engineers
Spring Conference, pp. 159-160, 2016.

10. Vasudevarao, B., Natarajan, D. P., Henderson, M., and Razdan, A.,
“Sensitivity of RP Surface Finish to Process Parameter Variation,”
Proc. of Solid Freeform Fabrication Proceedings Symposium, pp.
251-258, 2000.

11. Anitha, R., Arunachalam, S., and Radhakrishnan, P., “Critical

Parameters Influencing the Quality of Prototypes in Fused Deposition
Modelling,” Journal of Materials Processing Technology, Vol. 118,
No. 1, pp. 385-388, 2001.

12. Pandey, P. M., Reddy, N. V., and Dhande, S. G., “Improvement of

Surface Finish by Staircase Machining in Fused Deposition Modeling,”