http://jst.tnu.edu.vn; Email: jst@tnu.edu.vn 13
INFLUENCE OF AIR GAPS ON THE STRENGTH AND BEHAVIOR OF
POLYCARBONATE IN FUSED DEPOSITION MODELING
Phan Quang The* University of Technology - TNU
ABSTRACT
Fused deposition modeling (FDM) is widely applied in producing thermoplastic parts of complicated geometry and very effective in small lots. It is well understood that the strength of the product is determined by the strength of the bonding in layer and between layers. This paper presents an experimental research to analyze the influence of air gaps on the strength and behavior of PC produced by FDM through tensile and compressive tests in directions of bonding. A change in air gap was evident to cause a modification of the material resulting in changes of the strength and behavior of the model.
Keywords: FDM, auto adhesion, polycarbonate behavior, uniaxial tests.
Received: 22/10/2018; Revised: 12/11/2018; Approved: 28/12/2018
SỰ ẢNH HƯỞNG CỦA KHE HỞ ĐƯỜNG IN ĐẾN ĐỘ BỀN VÀ ĐẶC TÍNH CƠ HỌC CỦA POLYCARBONATE TRONG IN 3D- FDM
Phan Quang Thế Trường Đại học Kỹ thuật Công nghiệp - ĐH Thái Nguyên
TÓM TẮT
Fused deposition modeling (FDM) là một dạng của in 3D được ứng dụng rộng rãi trong chế tạo các chi tiết nhựa nhiệt có hình dáng phức tạp, rất hiệu quả với sản xuất đơn chiếc. Độ bền của chi tiết được quyết định bởi độ bền liên kết của các sợi trong cùng một lớp và giữa các lớp. Bài báo này trình bày một nghiên cứu thực nghiệm nhằm phân tích ảnh hưởng của khoảng cách giữa các sợi in đến độ bền và đặc tính của PC trong in 3D thông qua thí nghiệm kéo và nén theo phương liên kết giữa các sợi in. Sự thay đổi của khoảng cách giữa các sợi in là nguyên nhân gây ra sự thay đổi cấu trúc dẫn tới thay đổi độ bền và đặc tính của sản phẩm in.
Từ khóa: In 3D, FDM, sự bám dính, đặc tính vật liệu PC, kéo nén đúng tâm.
Ngày nhận bài: 22/10/2018; Hoàn thiện: 12/11/2018; Duyệt dăng: 28/12/2018
(*) Corresponding author: Email: phanqthe@tnut.edu.vn
http://jst.tnu.edu.vn; Email: jst@tnu.edu.vn 14
INTRODUCTION
FDM is a type of additive manufacturing technology that applies the principle of layer by layer with the uses of solid polymer filament as a sort of work material. The material then is being heated to flow temperature and being extruded out of printed head under the action of force caused by material feed system before deposited on the machine table to form roads and layers and eventually solidified and cooled within envelop environment. The FDM product can have good designed shapes of accuracy of minimum layer high as close as 0.1 mm and wall thickness of 0.8 mm [1, 2, 3]. Moreover, the more choice in thermoplastic such as PC, ABS, Nylon, ULTEM, etc. is available, the more options for FDM products is in practical purpose. However, even FDM has many advantages of capacity either imitate very complex structure in modeling or functional in industrial and medical application, its limitation exits in build time consumption.
Small diameters of extrusion head lead to longer build time; feed rate and plotting speed are limited due to reciprocating motions [4].
The strength of the FDM part depends on both the strength of the bonding between roads, layers and density of voids in the part.
To increase the strength of the part, different layering strategies are applied and in some patterns, are proved to gain better strength [1- 5]. A model of bonding between roads based on the 2D contact area with the concept of the bond growth or bond length in relation with diffusion phenomena and heat transfer process occurring in the FDM working chamber to determine bond strength is proposed [6, 7]. Longmei Li and co-workers [7] did an experiment to show the effect of post tempering on bond growth. It was evident that the longer time of tempering at a certain temperature of 200oC, 220oC and 240oC for ABS, the higher degree of bond growth is.
Thermal driven mechanism is the main factor to determine the strength of bonding. It is clear that the higher extrusion temperature that closes to the flow temperature at the roads’ contact the better bond strength. This is because the temperature facilitates the chemical effects at the interface of the roads of contact in aspect of diffusion [6, 7].
Temperature is therefore a very important parameter in FDM. To determine temperature, some thermal models are built in which a simplified 2D thermal model and a lumped capacity model are widely applied.
The two models show prediction of temperature through the conduction between layers and along a road, respectively and of course the effect of convection from the build environment temperature on the value of temperature versus time dependent. A direct connection between bond growths due to diffusion at a temperature above flow temperature is addressed. The effect of history temperature on the cooling process of the part is mentioned and determined to by experiment through Sun and co-workers [7].
It is very clear that auto-adhesion of the roads in the part has been investigated in micro- scale of 2D through the area of contact between them under the influence of heat generated during the process with the concept of bond growth [8- 12]. However, two cases need to be classified: Firstly, roads are in contacts in a layer results in the strength of the part depending on both contact in a layer and between layers (0 and negative air gaps).
Secondly, roads are not in contacts (positive air gaps) in a layer to form a porous structure are the reasons for the strength of the part depending on only contacts between layers.
The material for all of the test is PC because of the relative strength, the variety in functional purpose and the reasonable price compared to others. This study concentrate on the behavior of PC model produced by 0 and negative air gaps and its effects on the strength of the FDM part.
http://jst.tnu.edu.vn; Email: jst@tnu.edu.vn 15 EXPERIMENTAL PROCEDURE
Machine and manufacturing parameters The machine used for producing specimens was Stratasys Fortus 400 MC, made in USA.
Nozzle T16 (slide height: 0.010 in/0.254 mm) was used for depositing model and T12 (slide height: 0.007 in/0.178 mm) for support. The material used for this experiment was polycarbonate PC in the form of string with diameter of 0.06 in (1.77 mm) and support of 0.07 in (1.80 mm). The hardness of the string may be sufficient to create a pressure in the nozzle for the flow of model out. Width of the road: 0.02 in (0.51 mm) and Height of the road: 0.01 in (0.25 mm); maximum extrusion temperature: 345C for model and 240C for support, maximum oven temperature: 145C by default.
Manufacturing parameters were custom selected as raster orientation: parallel to the width of the tensile specimen; Air gaps were 0.000 in; -0.001 in (- 0.025 mm); -0.002 in (- 0.051 mm) and -0.003 in (-0.076 mm).
Mass of the specimen is then weighted on OHAUS - PA214 scale with the accuracy of 10-4 gram (2.2046226218e-7 lb).
Tensile and compressive test was carried out on the ADMET universal testing machines, eXpert2600. Default setting: position adjusted (or rate): 0.125 in/min (3.175 mm/min) until position is 1.0000 in (25.4 mm), sampling at 100 samples/sec, load cell: 2250.0 lb (1020 kg).
Failure interface of tensile test is analyzed in microstructure by using Scanning Electron
Microscopy (SEM) – FEI Quanta 600 (SEM).
Setting: low vacuum mode, 40 Pascals using water vapor.
Specimen preparation
The specimens for the tensile tests are designed following the standard ASTM D638 test specimen [11] as shown in Fig. 1(a) and top view of raster orientation Fig. 1(b) simulated in Insight software – Stratasys FDM system package. The tensile test was carried out directly from the specimen shown in Figure 1 to determine the strength between roads in layers. The purpose of the test is to determine the strength between roads so the tensile test for the second direction (parallel to the roads) can be ignored.
The compressive tests were carried out on cubic specimen of 7x7x7 in mm (0.28x0.28x0.28 in inches) cut from larger specimen, then ground to get appropriate sizes and shape. The compressive tests were carried out in two directions. The strength between layers is determined in the perpendicular direction (Per) to the interface between layers. The strength between roads in layers can be determined by the direction parallel to the interface and perpendicular to the roads (Pa). Polycarbonate at room temperature is a brittle material in glass phase and its compressive strength expects to be stronger than the tensile one. The tensile and compressive tests are carried out in both directions shown in Figure 2.
Figure 1. (a) Specimen for the tensile test, (b) raster orientation from top view in Insight software
(a
)
(b
)
http://jst.tnu.edu.vn; Email: jst@tnu.edu.vn 16
Figure 2. Directions of testing for FDM products RESULTS AND DISCUSSION
Four groups of five samples were manufactured with different air gaps of 0.000 in; -0.001 in (- 0.025 mm); -0.002 in (-0.051 mm) and -0.003 in (-0.076 mm). Other build parameters were kept by default from the machine’s recommendation. Results of the tests are analyzed in aspects of material behaviors, modulus, tensile and compressive strengths through mechanical testing and the nature of roads in contact through SEM.
Results in mechanical testing Variation of dimensions and density
It is evident from the data experiment that the dimensions of the specimen increased with the absolute air gap values toward negative in which the height of the specimen was increased largest, the width increased much less and eventually the length increased negligible. All changes in dimension and weight are shown in Table 1.
Table 1. Variation of the average height and mass of the specimen following the change of air gaps
Air gaps (in) 0.000 - 0.001 - 0.002 - 0.003
Density change (%) 0.00 3.40 2.90 -8.00
By calculation, it shown that with an increase of air gap values of (-0.001 in/-0.0025mm), the density of the specimen is increased by 3.4% and then reduced by 8% compared with that of air gap of 0.000 in (0.000 mm).
The voids between roads are illustrated in Figure 3.
The effect of voids is the model can be evaluated through density only it is impossible to include them in any sections through the model due it is widely scattered in both shapes and sizes. The area of the cross section is still evaluate as solid materials in both tensile and compressive tests.
Tensile strength
In fact, the tensile test was used to evaluate the strength between roads in the same layer.
In the tensile test, all the specimens were suddenly broken with a soft explosion as
shown in Figure 4. Each type of specimen was tested for three time. The average maximum load was recorded and then calculated in the form of stress as the tensile strength. The result is shown in Table 2 according to the change in density of the specimens. The relation between the average maximum tensile loads and air gap, and the relation between the tensile strength and air gaps are sketched as shown in Figure 4. All the values are calculated in relative to that of the values of air gap of (0.000 in/ 0.000 mm).
It shown that the law of change is the same for both loads and tensile strength. However, the difference increases toward air gap of - 0.003 (-0.076 mm) in.
It is very interesting to note that the minimum load was 287.5 lbf (1279 N) corresponding to the air gap of 0.000 in. The maximum load
http://jst.tnu.edu.vn; Email: jst@tnu.edu.vn 17 was 657.95 lbf (2927 N) corresponding to the
air gaps of -0.002 in (-0.051 mm), 2.29 times compared with that of the air gap of 0.000 in (0.000 mm). The increase of the load is linearly proportional to air gaps form value of 0.000 in (0.000 mm) to -0.002 in (-0.051 mm)
and inversely proportional from -0.002 in (- 0.051 mm) to -0.003 in (-0.076 mm). The law is approximately the same with the tensile strength one except the values are lower due to increasing of the area of the specimen’s cross section according to the changes of air gaps.
(a) (b)
Figure 3. Optical micrographs showing the voids in the cross section perpendicular to the roads (a) air gap of 0.000 in and (b) air gap of -0.002 in.
Table 2. Values of relative density and maximum tensile loads and strength in the relation with air gap’s values
Air gaps (mm) 0.000 - 0.025 - 0.051 - 0.076
Tensile loads (N) 1279 2056 2927 2784
Area (mm2) 46.45 47.09 49.67 59.35
Stresses (N/mm2) 27.53 43.66 58.93 46.91
Stress increased (%) 0.00 58.59 114.05 70.39
Figure 4. Relation between the maximum tensile loads and air gaps of the specimens
Compressive strength
In contrast to the tensile tests, all the specimens in compressive tests were not broken even the overload value of the machine reaches at (2200 lb/ 997 kg) as shown in Fig 6(a). A transition point between the convex part and concave part of the graph is used to evaluate the compressive strength of the specimens. The cross section was 7mm x 7mm (0.28x0.28 in2). An attempt was made to see the failure with the use of a smaller specimen of 3.5mm x 3.5mm (0.14x0.14x0.14)
http://jst.tnu.edu.vn; Email: jst@tnu.edu.vn 18
in3 and air gap of 0.000 in/ 0.000mm. There was a point of the compressive failure with a very small explosion as shown in Fig 6(b). It will be discussed in details in next section.
Table 3. Values of relative density and compressive strength corresponding to the transition point in the graph in the relation with air gap’s values
Air gaps (mm) 0.000 - 0.025 - 0.051 - 0.076
Pa compressive strength (N/mm2) 63.07 70.24 78.95 77.73
Relative change % 0.00 11.36 25.17 23.24
Per compressive strength (N/mm2) 65.67 75.93 82.05 79.46
Relative change % 0.00 15.62 24.94 20.99
Figure 5. Relation between the compressive loads at the transition point and air gaps of the samples.
Similarity to the tensile strength, the relations between the compressive strength and air gap are plotted in Figure 5. The compressive strength here were recorded in not only between the roads in a layer, parallel (Pa) but also between roads in different layer, perpendicular (Per). The tendency of change is similar to that of the tensile strength but the values are very close to each other. However, the compressive strength at failure are much higher than that of the tensile strength as shown in Figure 7(b). FDM models can suffer very high compressive strength without failure.
Behavior of Polycarbonate
Model’s behavior under tensile tests
Two examples of material behavior in tensile tests at room temperature with the rate 0.125 in/minute (3.175 mm/min) is shown in Figure 6. The specimen was suddenly broken at the maximum loads. The average maximum loads were recorded as 1278.86 N; 2055 N; 2922.48 N;
2780 N corresponding to displacements of the specimens at the failure 2.28 mm, 3.81 mm, 5.21 mm and 4.31 mm, respectively and the specimen’s air gaps of 0.000 in; -0.001 in (- 0.025 mm); - 0.002 in (-0.051 mm) and -0.003 in (-0.076 mm), respectively. It is clear that the highest value of the load, the highest displacement of the specimen is.
http://jst.tnu.edu.vn; Email: jst@tnu.edu.vn 19 Figure 6. Load and displacement under tensile test (rate 0.125 in/minute) in direction parallel to the
interface and perpendicular to road’s orientation (a) Air gap: 0.000 in (b) -0.002 in
Model’s behavior under compressive test
Figure 7. (a) Load and displacement under compressive tests, air gap: - 0.003 in in the direction perpendicular to the interface of the layers (b) Compressive failure, air gap: 0.000 in, sample’s size of
0.14x0.14x0.14 in3 (rate 0.125 in/minute) It is obviously from Figure 7 that the material
exhibits more brittle at the air gap of 0.000 in and the highest ductile at air gap of -0.002 in (-0.051 mm) and reduce ductility at the air gap of -0.003 in (-0.076 mm). The relation between the tensile load and the sample’s displacement is approximately linear at air gap of 0.000 in to a more curve at air gap of - 0.002 (-0.051 mm) in as shown in Figure 6. It appears that there is a connection between high tensile strength and high displacement of the FDM specimen and also the values of air gaps. The behavior of the model exhibits more ductile behavior, the higher tensile strength it is. It shows that there is a change in
material behavior with different air gap’s values. This can be explained by the principle of FDM layer by layer resulting in microstructure different from that of solid polymers.
The material behavior in compressive tests in two directions is shown in Figure 7.
According to Ravi-Chandar and Ma [6], a testing curve of stress-strain relation of polycarbonate under uniaxial compressive loading can be divided into 4 regimes in Figure 5 in their publication. In the 4th regime, the behavior of PC is approximately similar to that of plastic region in metals. A nonlinear
(a)
(b)
http://jst.tnu.edu.vn; Email: jst@tnu.edu.vn 20
stress – strain behavior is attributed to viscous effects. Non-uniform deformation appears when the barreling initiates coincident with the peak load. The response of the specimen beyond the peak load is not the constitutive behavior of the material. Polycarbonate exhibits intrinsic softening.
The behavior of FDM PC in this study is different from Solid PC of Ravi – Chandar and Ma. The relation between the load and the sample’s displacement is divided into 2 main regions. In the first region of convex curve in the graph, the relation follows a convex curve and then a concave one in the second region. A transition between the two curves is considered as a transition point of PC behavior in load-displacement graph. The sample becomes flatter with the increasing of the load, no failure for the designed specimens was detected even the load limit of the machine (2250.0 lbf) reached. The model’s behavior in the second region appears in the way similar to ductile materials in the second half of strain hardening of the ductile materials.
It is very interesting to note that a failure point in the load – displacement graph was detected with much smaller samples. The specimen failed immediately as shown in Figure 7(b) or there was a small drop of the load on the graph with a very soft exploded with other specimens. The specimen becomes flatter and flatter along with the fact that the load continues increasing to the machine’s limit. The failure’s behavior of the specimen exhibits a characteristic of FDM structure with void inside even in the specimen with air gap of (-0.003 in/-0.076 mm)
Polycarbonate is a thermal plastic existing in amorphous structure. At room temperature, it exists in the form of glass phase and is a brittle material. However, it appears that in compressive process, after the transition point
is reached, the material behavior changes to like plastic one for all the specimens of the different air gaps in the range of the study. Of course, the material behavior here may be not intrinsic softening as in solid state but the structure as a composite with voids in it is the main reason for this phenomenon.
Microstructure analysis
It is very clear that the negative air gap is characterized for level of material overlap between roads. The larger absolute value, the more overlap is. The negative air gap therefore produces contacts between roads not only between layers but also in each layer.
The surfaces generated at the failure region under tensile test was examine on SEM. The microstructure on the surfaces were different when different air gap’s values were applied.
The voids in cross section are shown in Figure 3 with smaller voids corresponding to the increase of air gaps toward negative values. The failure of the specimen with air gap of 0.000 in is brittle fracture as shown in Figure 8(a) in combination with the largest areas of voids resulting in the lowest tensile strength. The type of failure of the specimen with air gap of (-0.001 in/-0.025 mm) is also brittle fracture but the areas of voids reduces considerably in combination of larger areas of contact as shown in Figure 8(b) producing higher tensile strength in comparison with the first case. It is very interesting to note that the areas of voids are reduced and the contact areas are also enlarge compared with the first two cases. However, necking or ductile behavior appears on the failure surface as shown in Figure 8(c). In other words, the behavior of the specimen with air gap of (- 0.002 in/-0.025mm) moves toward ductile resulting in the highest strength of the specimen. The failure surface of the specimen with air gap of (-0.003 in/-0.076 mm) is evident to have negligible voids and brittle fracture occurred in nearly the whole of the
http://jst.tnu.edu.vn; Email: jst@tnu.edu.vn 21 apparent area of contact. The largest area of
contact is the reason for high tensile strength
of the specimen than the first two cases.
Figure 8. Failure surface of the specimens of air gaps: (a) 0.000 in; (b) -0.001 in (c) -0.002 in and (d) - 0.003 in
Conditions created to produce the specimen with air gap of -0.002 in (-0.025mm) may be involved with the most ductile property, a precious property for maximum tensile strength this model gets.
CONCLUSIONS
In conclusions, the application of zero and negative air gaps results in changes of both the height and density of the specimen in this study up to 26% for height and 8% for density. The tensile strength at air gap of - 0.002 in (-0.025 mm) increased by 2.14 times compared with that of 0.000 in with only approximately 10% of mass increased. In compressive test, the transition point in the load-displacement graph is consider as the compressive strength. The law of change of compressive strength is similar to that of the
tensile strength but the differences in two directions parallel to the interface between layers are very small. Moreover, the differences in strength in relation of changes of air gaps are also small. The behavior of the model is evident to depend on the air gap.
Like ductile behavior is observed in the specimen with air gap of -0.002 in (-0.025 mm) and brittle behavior is detected in other air gaps selected in the study.
REFERENCE
1. Wendel. B, Rietzel. D, Kuhnlein. F, Feulner. R,
(a) (b)
(c) (d)
http://jst.tnu.edu.vn; Email: jst@tnu.edu.vn 22
Hulder. G and Schmachtenberg. E., “Additive Processing of Polymers – Review”, Macromolecular Materials and Engineering, 2008, Vol. 293, pp. 799-809.
2. Brian N. Turner, Robert Strong and Scott A.
Gold., “A review of melt extrusion additive manufacturing processes: I. Process design and modeling”, Rapid Prototyping Journal, 20/3 (2014), pp. 192–204.
3. Brian N. Turner, Robert Strong and Scott A.
Gold., “A review of melt extrusion additive manufacturing processes: II. “Materials, Dimensional Accuracy, and Surface Roughness”, Rapid Prototyping Journal, 20/3 (2015), pp. 250–
261.
4. Kulkarni. P and Dutta. D., “Deposition Strategies and Resulting Part Stiffness in Fused Deposition Modelling”, Journal of Manufacturing Science and Engineering, February 1999, Vol.
121, pp. 93-103.
5. Ang. K.C, Leong. K.F and Chua. C.K.,
“Investigation of the Mechanical Properties and Porosity Relationships in Fused Deposition Modelling – Fabricated Porous Structures”, Rapid Prototyping Journal, Vol. 12/2, 2006, pp.
100-105.
6. Ravi-Chandar, K and Ma. Z., “Inelastic Deformation in Polymers under Multiaxial Compression”, Mechanics of Time-Dependent Materials, Vol. 4, 2000, pp. 333-357.
7. Bellehumeur. C, Li. L, Sun. Q, GU. P.,
“Modelling of the Bond Formation between Polymer Filaments in the Fused Deposition Modelling Process”, Journal of Manufacturing Processes, 2004, Vol. 6/2, pp. 170-178.
8. Bellehumeur. C, Li. L, Sun. Q, GU. P.,
“Modelling of the Bond Formation between Polymer Filaments in the Fused Deposition Modelling Process”, Journal of Manufacturing Processes, 2004, Vol. 6/2, pp. 170-178.
9. Stark. M.S., Improving and Understanding Inter-Filament Bonding in 3D-Printed Polymers, Thesis Project, University of Tennessee, 2016.
10. Voyutskii.S.S., Autohesion and Adhesion of High Polymers, Interscience Publishers, New York, 1963.
11. Osswald Tim.A and Menges Georg., Material Science of Polymers for Engineers, third edition, Hanser Publishers, Munich, 2012.
12. Klein. R., Laser welding of Plastics, Wiley – VCH Verlag GmbH & CO. KGaA, 2011.