- Email: [email protected]
Effect of flow forming on mechanical properties of high density polyethylene pipes
Journal of Manufacturing Processes 19 (2015) 155–162
Contents lists available at ScienceDirect
Journal of Manufacturing Processes journal homepage: www.elsevier.com/locate/manpro
Technical Paper
Effect of flow forming on mechanical properties of high density polyethylene pipes Amin Abedini ∗ , Payam Rahimlou, Taghi Asiabi, Samrand Rash Ahmadi, Taher Azdast Mechanical engineering department, Urmia University, Urmia, Iran
a r t i c l e
i n f o
Article history: Received 23 September 2014 Received in revised form 25 June 2015 Accepted 25 June 2015 Keywords: Flow forming HDPE Mechanical properties Taguchi method ANOVA
a b s t r a c t Flow forming is a single rotary-contact-point cold-forming process, in which the thickness of a tubular pre-form is reduced while its length increases without any change in internal diameter. Although metal flow forming has been studied by many researchers and industries, there does not appear to be any published study on the flow forming of polyethylene pipes. The main purpose of this study was to find out the applicability of flow forming on polyethylene pipes and the effects of process parameters on mechanical properties of polyethylene flow formed parts. The experiments were carried out using HDPE80 tubular pre-forms. Thickness reduction ratio, feed rate and rotation speed of mandrel were considered as variables. An L9 orthogonal array of Taguchi method was applied to carry out the experiments. Tensile and impact tests were performed to examine the failure behavior of specimens. Scanning electron microscopy (SEM) was implemented to find out the relation between mechanical properties and microstructure of the material. Stress at break, yield Stress, percentage of elongation at break and impact endurance were measured as indicators of mechanical properties. It was found that mechanical properties of HDPE increase significantly during flow forming. The solid state deformation makes material more oriented and anisotropic that leads to an enhancement in mechanical properties of HDPE. Through SEM it was found that the material experiences a transformation from an isotropic spherulite structure into an anisotropic lamellar-fibrous structure at large deformation. © 2015 The Society of Manufacturing Engineers. Published by Elsevier Ltd. All rights reserved.
1. Introduction Flow forming is a single contact-point cold-forming process, in which the thickness of a pre-form is reduced whilst its length increases without any change in the internal diameter. Flow forming is a simple, chip-less and cost-effective forming process to produce tubular products with a high degree of mechanical and dimensional properties [1,2]. In flow forming a pre-form is fitted to a rotating mandrel. An external force is applied to the outer diameter of pre-form by roller(s); the rollers move along the axis of mandrel to reduce the wall thickness of the preform. The length of part is increased while its thickness decreases due to the law of conservation of volume. The axial movement of rollers is called feed. There are two modes of flow forming associated with the process, backward and forward flow forming [3]. In forward flow forming material flows ahead of the
∗ Corresponding author. Tel.: +98 9187628856. E-mail address: st [email protected] (A. Abedini).
rollers in the same direction as the rollers feed. In backward flow forming material flows in the opposite direction of rollers feed, Figs. 1 and 2 show the modes of flow forming schematically [3]. Ameliorating the mechanical properties, sound finished surface, high dimensional accuracy, simple tool design, low cost and high productivity are some advantages of flow forming. [4,5]. Many experimental, analytical and numerical studies have been presented about metal flow forming. Davidson et al. [6] experimentally studied the effect of flow forming parameters on the surface quality of flow formed aluminum tubes. Roy et al. [7] introduced an analytical model to predict the shape of roller/work piece interface. Molladavoudi et al. [4] established an experimental study to examine the effects of thickness reduction on mechanical properties of aluminum tubes. Although metal flow forming has been studied by many researchers and industries, there does not appear to be any published study on the flow forming of polyethylene pipes. This study was performed to examine the effect of flow forming on mechanical properties of polyethylene pipes.
http://dx.doi.org/10.1016/j.jmapro.2015.06.014 1526-6125/© 2015 The Society of Manufacturing Engineers. Published by Elsevier Ltd. All rights reserved.
156
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162
Fig. 3. Flow forming machine.
Fig. 1. Backward Flow forming [3].
2. Experimental Fig. 4. Details of roller geometry.
2.1. Equipment A single roller backward flow forming machine was used to conduct the experiments. The machine was an NC lathe which was converted to a flow forming machine. A set of roller and mandrel was designed and manufactured to mount on the cross slide of the lathe. A steel mandrel with a male lug was secured by the chuck of lathe on one end and by the tailstock on the other end. Throughout the process the rotation speed of mandrel was applied in RPM (rotation per minute) and feed rate of roller in mm/rev. To apply the thickness reduction an equation was considered as thickness reduction ratio, Eq. (1) represents this ratio. Where to is initial wall thickness of pre-form and tf is the wall thickness of flow formed part. Thickness reduction ratio =
to − tf to
(1)
Table 1 Roller geometrical details. Roller radius
Attack angle
Tip radius
Smooth angle
120 mm
30◦
2.5 mm
6◦
Fig. 3 shows the general set-up of the flow forming machine during forming. Fig. 4 shows the details of roller geometry schematically, Table 1 represents the dimensions of roller. To avoid heating effects during flow forming an intense flow (0.5 l/s) of coolant was used to cool down the parts (roller and mandrel). 2.2. Material The material used for the pipes, was a copolymer of polyethylene HDPE80 in black color with a blue line along the tube to show and magnify the probable torsion of material. Table 2 shows the technical properties of HDPE 80 [8]. The tube specimen had an external diameter of 63 (±0.1)mm, an internal diameter of 51 (±0.1)mm and a length of 200 (±0.1)mm. 2.3. Design of experiment An L9 orthogonal array of Taguchi method was used to conduct and analyze the experiments. Taguchi method is a simple, systematic and rapid design of experiment tool. By using Taguchi method it is possible to study a process with only a small number of experiments. Thickness reduction ratio, rotation speed and feed rate were considered as variable parameters. Three levels Table 2 Technical properties of HDPE 80 [8]. Unit
Fig. 2. Forward Flow forming [3].
Density Yield stress endurance Breaking elongation Elasticity modulus Hardness
Value 3
g/cm MPa % MPa Shore D
0.940 20–22 >600 700–1000 40
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162
157
Table 3 Variables and their levels. Parameter
Low
Medium
High
Thickness reduction (%) Feed rate (mm/rev) Angular speed (rpm)
20 0.16 90
40 0.32 180
60 0.48 360
Fig. 6. Dumb-bell shaped tensile specimens.
Table 5 Measured mechanical properties of normal specimen. Stress at break (MPa) Fig. 5. Flow formed pipes.
including High, Medium and low were specified for each variable. Table 3 shows variables and their levels. Using an L9 orthogonal array as experiments layout needs 9 experiments to be performed. Table 4 shows the layout of this study and the measured mechanical properties of each experiment. The capital letters L, M and H indicate Low, Medium and High level respectively. Fig. 5 shows the flow formed pipes according to L9 orthogonal array. Table 5 shows the measured mechanical properties for normal un-flow formed specimen.
Yield stress (MPa)
10.5
Percentage of elongation at break (%)
19
Impact endurance (J/m)
620
450
standard introduces standard test methods which used to determine the Charpy impact endurance of plastics. A digital Charpy impact tester was used to do the tests. Fig. 8 shows the notched specimens, the Charpy impact tester and a specimen after Charpy test.
2.4. Tensile test
2.6. Scanning Electron Microscopy (SEM)
Dumb-bell shaped specimens were removed from the flow formed pipes according to ISO 6259 3 type 1 standard, using a CNC mill. Fig. 6 shows the prepared tensile specimens. The axis of specimens was parallel to the longitudinal axis of pipes. Tensile tests were carried out on an automatic universal testing machine. The cross-head speed of machine was set at 50 mm/min. The yield Stress and stress at break were defined as the load at specified points divided by the initial cross-sectional area. The tests were carried out in 3 iterations and the arithmetic mean was used to represent the average of measured values for each experiment. Fig. 7 shows the output plot of tensile tests for experiments of L9 orthogonal array.
Rectangular specimens were cut from the flow formed pipes and immerged in liquid nitrogen. The frozen samples were broken along the longitudinal axis of pipes to expose their cross sections. A thin layer of gold (about 100 angstroms) was deposited on the surface of specimens using physical vapor deposition method to avoid the charging effect of surface. A KYKY-SBC12 sputter coater was used to coat the surfaces with gold. The cross sections were observed using a KYKY-EM3200 electron microscope. The images of microstructure were prepared by different magnifications to investigate changing of microstructure due to flow forming.
3. Result and discussion 2.5. Impact test Impact notched specimens were removed from the flow formed pipes according to ASTM 6110-10 standard using a CNC mill. This
The output results of tensile, impact tests and SEM images were analyzed to discover the effect of flow forming parameters on mechanical properties of high density polyethylene pipes.
Table 4 L9 array and measured mechanical properties. RUN
RPM
Feed rate
Thickness reduction
Stress at break (MPa)
Yield stress (MPa)
Percentage of elongation at break (%)
Impact endurance J/m
1 2 3 4 5 6 7 8 9
L L L M M M H H H
L M H L M H L M H
L M H M H L H L M
13 14 16 18 11 20 13 14 13
20.5 21.5 22.0 21.0 23.5 20.5 24.5 21.5 21.5
790 817 817 730 632 790 680 850 820
712 982 1650 1015 2105 1223 3996 1682 3368
158
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162
Fig. 7. Output plots of tensile test.
3.1. Stress at break 3.1.1. Main effects Fig. 9 shows the average of stress at break at each level of process parameters. The plot of thickness reduction ratio vs. stress at break shows a significant improvement in stress at break due to cold flow forming. High density polyethylene is a semi-crystalline polymer and in this type of polymers amorphous and crystalline regions exist simultaneously. The crystalline region is composed of spherulites connected by tie molecules. The amorphous exist in the form of entangled chains, loop chains and floating chains. The main feature in amorphous region is entanglements. Therefore, semi-crystalline polymers consist of an intersecting network of spherulites and entanglements [9]. In flow forming process
material flows in axial, radial and tangential directions through the gap between mandrel and roller but the axial flow is the most dominant. Thus, thickness reduction ratio has an important effect on the plastic deformation of material. During flow forming of semi-crystalline polymers two processes are believed to occur, the spherulitic structure deforms to an ellipsoid and the lamellar structure inside the ellipsoid deforms to a fibrous structure. Usually the change of microstructure shows that an original isotropic structure becomes lamina in shape at moderate deformation and fibril at large deformation. Fig. 10 shows this phenomenon schematically. In this way, the semi-crystalline polymers behave as composites and the increase in stress at break due to cold work can be associated with the reorientation of different regions of polymer.
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162
159
Table 6 ANOVA results for stress at break. Source
DF
SS
MS
P
F0
Percentage
Rank
T. R ratio Feed rate Rotation speed Residual error Total
2 2 2 2 8
42.88 13.55 11.55 0.22 68.22
21.44 6.77 5.77
0.051 0.573 0.515
5.08 0.61 0.74
62.8 19.8 16.9
1 2 3
Table 7 Average and differences between levels for stress at break. Fig. 8. Notched specimens, the Charpy impact tester and a broken specimen.
18 Breakage stress (MPa)
17 16 15
Thickness reduction
14
FEED
13 RPM
12 11 L
M
H
Level Fig. 9. Average of Stress at break at each level of process parameters.
Fig. 11 shows the micrographs of specimens. The un-flow formed sample has an isotropic structure with roughly equal void-like spaces. At low level of thickness reduction ratio the microstructure has no clear transformation but the size of voidlike spaces has increased. At medium level of thickness reduction ratio, the increase of the size of void-like spaces is more visible and their orientations are along the axis of specimen where the most significant flow of material occurs. At high level of thickness reduction ratio, the transition from a spherulitic structure to a lamellar fibrous structure is clearly evident. Fig. 12 shows the fibril structure of highly flow formed sample with magnification of 1000×. The Plot of rotation speed vs. stress at break has an increasing trend, stress at break increases with increase in rotation speed. A possible explanation is that when the rotation speed increases the uniformity of deformation increases that leads to a higher anisotropy in texture of the material. The uniformity in texture makes material more orientated and causes a higher stress at break. According to Tsunekawa et al. [10] the cold forming of an isotropic semi-crystalline polymer usually produces a fibrous material with a high anisotropy of physical properties such as elastic modulus and tensile strength. According to Ingram and Peterlin [11] this anisotropy is closely connected with the molecular orientation
Level
RPM
Feed rate
T.R ratio
L M H (Delta) Rank
13 15 15.67 2.67 3
16 14 13 3 2
12 14.33 17.33 5.33 1
in the crystalline and amorphous regions. The amount of fibrous region in polymer depends on the amount of cold work applied to the material. Therefore, it seems that an increase in rotation speed increases the amount of re-orientation of fibrous region that lead to a higher tensile strength. The plot of stress at break vs. feed rate has a decreasing trend. The high level of feed rate gives the lowest stress at break. It seems at the higher feed rate roller neglects the formation of material and slides over the tubes. This behavior causes difference in mechanical properties of internal layers and external layers hence, yielding process occurs inhomogeneously and the material fails prematurely. 3.1.2. ANOVA results Table 6 shows the ANOVA results for average of stress at break as response function. According to Table 6, thickness reduction ratio is the most significant factor that affects the stress at break. The percentage of contribution of 62.8 for thickness reduction ratio is totally significant. Table 7 shows average of the stress at break at each level and differences between levels of parameters. As it can be observed from Table 7 the difference between the value of stress at break at high level and low level of thickness reduction ratio is totally significant. This means that the amount of deformation is a key factor that affects the yield behavior of HDPE polymer. 3.2. Yield stress 3.2.1. Main effects Fig. 13 shows the average of yield Stress at each level of process parameters. The plot of thickness reduction ratio vs. yield
Fig. 10. Schematic diagram of the flow formed spherulites.
160
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162
Fig. 11. Micrographs of specimens with different thickness reduction ratios.
Fig. 12. Fibril structure of highly flow formed sample.
23.5
Yield Stress (MPa)
23 22.5 Thickness reduction
22 21.5
Feed
21 RPM
20.5 20 L
M
H
Level Fig. 13. Average of yield Stress at each level of process parameters.
Stress shows a significant rise in yield Stress. It seems that the re-orientation of chains is the most important phenomenon that affects the yield Stress. It has been recognized that the yield behavior of semi-crystalline polymers is associated with plastic deformation of crystal parts [12] and the morphology transformation from original isotropic structure to an oriented one [13]. Solid state deformation is one of the fundamental processes to induce polymer to achieve high mechanical properties [14]. It is a
useful method to enhance the mechanical properties of polymers by inducing a high degree of orientation during solid state deformation of semi-crystalline polymers [15]. Therefore, as Fig. 12 indicates, it can be concluded that polymer chain orientation plays a vital role in controlling the yield behavior of a semi-crystalline polymer. The plot of rotation speed vs. yield stress also shows an increasing trend. It seems that higher rotation speed applies more cold deformation to the material and makes the material more oriented. As it is mentioned re-orientation of chains is the main factor associated with high tensile strength. The plot of yield stress vs. Feed rate shows that at higher levels of feed rate, the yield Stress decreases. This can be related to the imperfect deformation due to high feed rate. The high feed rate applies load only on the outer layers of material, it can scratch the surface of material and break the polymers chains. These imperfections are suitable places for creating stress concentration that leads to a premature failure.
3.2.2. ANOVA results Table 8 shows the ANOVA results for average of yield Stress as response function. According to Table 8, thickness reduction ratio is
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162 Table 8 ANOVA results for yield stress.
161
Table 10 ANOVA results for percentage of elongation at break.
Source
DF
SS
MS
P
F0
Percentage
Rank
Source
DF
T. R ratio Rotation speed Feed rate Residual error Total
2 2 2 2 8
10.50 2.17 1.17 0.66 14.50
5.25 1.08 0.58
0.021 0.615 0.778
7.87 0.53 0.26
74.21 14.94 8.05
1 2 3
T. R ratio Rotation speed Feed rate Residual error Total
2 2 2 2 8
Table 9 Average and differences between levels.
SS 20,551 13,444 7264 13,737 54,996
MS 10,275 6722 3632
P
F0
Percentage
Rank
0.246 0.431 0.654
1.79 0.97 0.46
37.37 24.24 13.21
1 2 3
Table 11 Average and differences between levels.
Level
RPM
Feed rate
T.R ratio
Level
RPM
Feed rate
T.R ratio
L M H (Delta) Rank
21.33 21.67 22.50 1.17 2
22 22.17 21.33 0.83 3
20.83 21.33 23.33 2.5 1
L M H (Delta) Rank
808.0 760.0 713.3 94.7 2
723.3 792.3 765.7 69.0 3
826.7 739.0 715.7 111.0 1
840 820 Elongaon (%)
800 780
Thickness reduction
760
Feed
740
Impact endurance (J/m)
3400 2900
Feed
1900
RPM
1400
RPM
720
Thickness reduction
2400
900 L
700 L
M
M Level
H
H
Level Fig. 14. Average of percentage of elongation at break at each level of process parameters.
the most significant factor that affects the yield Stress. The percentage of contribution of 74.21 for thickness reduction ratio is totally significant. Table 9 shows average of the yield stress at each level and differences between levels of parameters. As Table 9 indicates thickness reduction ratio has a predominate effect on yield stress of flow formed HDPE.
Fig. 15. Average of Impact endurance at each level of process parameters. Table 12 ANOVA results for impact endurance. Source
DF SS
Rotation speed T. R ratio Feed rate Residual error Total
2 2 2 2 8
MS
618,1002 30,90501 287,0940 14,35470 371,692 18,5846 757,312 10,180,946
P
F0
Percentage Rank
0.061 4.64 60.71 0.370 1.18 28.20 0.894 0.11 3.65
1 2 3
3.4. Impact endurance 3.3. Percentage of elongation at break 3.3.1. Main effects Fig. 14 shows the average percentage of elongation at break at each level of process parameters. In all experiments of L9 orthogonal array the percentage of elongation at break increases in comparison to normal HDPE but the plot of thickness reduction ratio and rotation speed show a significant downward slope. As it was mentioned before the high amount of deformation makes material more oriented and produces greater molecular orientation and compaction of material that leads to a ductile to brittle transformation. It seems that the predominant element of an oriented polymer is the presence of microfibrils. The microfibrils are connected to many tie molecules through the amorphous region and make a fibrous stiffness in the material. 3.3.2. ANOVA results Table 10 shows the ANOVA results for average of percentage of elongation at break as response function. According to Table 10, thickness reduction ratio is the most significant factor that affects the percentage of elongation at break. Table 11 shows the average percentage of elongation at break at each level and the differences between levels of parameters.
3.4.1. Main effects Fig. 15 shows the average impact endurance at each level of process parameters. The Charpy test measures the impact energy or the energy absorbed prior to fracture. The impact behavior of semi-crystalline polymers is complicated because of the presence of the crystalline and the amorphous phases at the same time. Cold state flow forming transforms spherulites into a fibrous structure. As it is shown in Figs. 11 and 12 the fibrous structure consists of oriented lamellae in crystalline phase connected to many tie molecules within the amorphous phase. It seems that three-dimensional entanglement occurs in amorphous regions. Three-dimensional entanglement forms a network by intertwining of neighboring polymer chains. As Fig. 15 indicates at high levels of deformation the material shows a better impact endurance, a reasonable explanation can be the presence of re-orientation, fibrous structure and entanglement of polymer chains. 3.4.2. ANOVA results Table 12 shows the ANOVA results for average impact endurance as response function. According to Table 12, rotation speed is the most significant factor that affects the impact endurance. The percentage of
162
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162
Table 13 Average and differences between levels. Level
RPM
Feed rate
T.R ratio
L M H (Delta) Rank
1115 1448 3015 1901 1
1908 1590 2080 491 3
1206 1788 2584 1378 2
contribution of 60.71 is totally considerable. The effect of feed rate is trivial. Table 13 shows the average impact endurance at each level and differences between levels of parameters. 4. Conclusion The effect flow forming parameters on the mechanical properties of high density polyethylene (HDPE) pipes were critically analyzed using Taguchi method. Thickness reduction ratio, feed rate and rotation speed of mandrel were considered as variables. Stress at break, yield Stress, percentage of elongation at break and impact endurance were measured as indicators of mechanical properties. Scanning electron microscopy was performed to find the relation between mechanical properties and microstructure of the material. It was found that mechanical properties of HDPE increase considerably during flow forming. The solid state deformation makes the material more oriented and anisotropic that leads to an enhancement in mechanical properties of HDPE. High density polyethylene is a semi-crystalline polymer and its microstructure consists of amorphous and crystalline regions simultaneously. The crystalline region is composed of spherulites connected by tie molecules, the amorphous regions exist in the form of entangled chains, loop chains and floating chains. During flow forming the spherulitic structure deforms to an ellipsoid and the lamellar structure inside the ellipsoid deforms to a fibrous structure. The amount of fibrous region in polymer depends on the amount of cold work applied to the material. Through scanning electron microscopy, it was found that an original isotropic structure becomes lamina in shape at moderate deformation and fibril at large deformation. Because the
material experiences a transformation from spherulite structure into a fibrous structure at large deformation, the semi-crystalline polymers behave as composites and the amelioration of mechanical properties due to cold work can be associated with the reorientation of different regions of polymer. A high degree of anisotropy in texture of the material was evident at large deformation. Through ANOVA, it was found that thickness reduction ratio is the most significant parameter affecting the mechanical properties of HDPE. References [1] Podder B, Mondal C, Kumar KR, Yadav DR. Effect of preform heat treatment on the flow formability and mechanical properties of AISI4340 steel. J Mater Des 2012;37:174–81. [2] Xia QX, Cheng XQ, Hu Y, Ruan F. Finite element simulation and experimental investigation on the forming forces of 3D non-axisymmetrical tubes spinning. Int J Mech Sci 2006;48:726–35. [3] Wong CC, Dean TA, Lin J. A review of spinning, shear forming and flow forming processes. Int J Mach Tools Manuf 2003;43:1419–35. [4] Molladavoudi HR, Djavanroodi F. Experimental study of thickness reduction effects on mechanical properties and spinning accuracy of aluminum 7075-O, during flow forming. Int J Adv Manuf Technol 2011;52:949–57. [5] Joseph Davidson M, Balasubramanian K, Tagore GRN. Experimental investigation on flow forming of AA6061 alloy—a Taguchi approach. J Mater Process Technol 2008;200:283–7. [6] Joseph Davidson M, Balasubramanian K, Tagore GRN. An experimental study on the quality of flow-formed AA6061 tubes. J Mater Process Technol 2008;203:321–5. [7] Roy MJ, Maijer DM, Klassen RJ, Wood JT, Schost E. Analytical solution of the tooling/work piece contact interface shape during a flow forming operation. J Mater Process Technol 2010;210:1976–85. [8] Tec Industries Inc Polyethylene pipes and fittings supplier, Technical catalog, http://www.tecindustries.com.my/pe gas supply pipe.html, 02/18/2015. [9] Sizhu Wu. The effect of molecular orientation on the mechanical behavior of cold-rolled semi-crystalline polymers. Hung Hom, Hon Kong: Hong Kong Polytechnic University; 2002 (Doctoral thesis submitted). [10] Tsunekawa Y, Oyane M, Kojima K. Effects of irradiation and rolling on the tensile behavior of polyethylene. J Polym Sci 1961;50:35–44. [11] Ingram I, Peterlin A. Electron diffraction and microscopy of deformed polyethylene spherulites. J Polym Sci—Polym Lett 1964;2:739–45. [12] Juska T, Harrison IR. A criterion for craze formation. Polym Eng Sci 1982;22:766–76. [13] Crist B, Fisher CJ, Howard PR. Mechanical properties of model polyethylene tensile elastic-modulus and yield stress. J Macromol 1989;22:1709–18. [14] Shaw MT. Cold forming of polymeric materials. Annu Rev Mater Sci 1980;10:19–42. [15] Katto A, Makayama K, Kanestsuna H. Infrared dichroism and visible—ultraviolet dichroism studies on rolled-drawn polypropylene and polyethylene sheets. J Macromol Sci-Phys B26 1987:281–306.
Contents lists available at ScienceDirect
Journal of Manufacturing Processes journal homepage: www.elsevier.com/locate/manpro
Technical Paper
Effect of flow forming on mechanical properties of high density polyethylene pipes Amin Abedini ∗ , Payam Rahimlou, Taghi Asiabi, Samrand Rash Ahmadi, Taher Azdast Mechanical engineering department, Urmia University, Urmia, Iran
a r t i c l e
i n f o
Article history: Received 23 September 2014 Received in revised form 25 June 2015 Accepted 25 June 2015 Keywords: Flow forming HDPE Mechanical properties Taguchi method ANOVA
a b s t r a c t Flow forming is a single rotary-contact-point cold-forming process, in which the thickness of a tubular pre-form is reduced while its length increases without any change in internal diameter. Although metal flow forming has been studied by many researchers and industries, there does not appear to be any published study on the flow forming of polyethylene pipes. The main purpose of this study was to find out the applicability of flow forming on polyethylene pipes and the effects of process parameters on mechanical properties of polyethylene flow formed parts. The experiments were carried out using HDPE80 tubular pre-forms. Thickness reduction ratio, feed rate and rotation speed of mandrel were considered as variables. An L9 orthogonal array of Taguchi method was applied to carry out the experiments. Tensile and impact tests were performed to examine the failure behavior of specimens. Scanning electron microscopy (SEM) was implemented to find out the relation between mechanical properties and microstructure of the material. Stress at break, yield Stress, percentage of elongation at break and impact endurance were measured as indicators of mechanical properties. It was found that mechanical properties of HDPE increase significantly during flow forming. The solid state deformation makes material more oriented and anisotropic that leads to an enhancement in mechanical properties of HDPE. Through SEM it was found that the material experiences a transformation from an isotropic spherulite structure into an anisotropic lamellar-fibrous structure at large deformation. © 2015 The Society of Manufacturing Engineers. Published by Elsevier Ltd. All rights reserved.
1. Introduction Flow forming is a single contact-point cold-forming process, in which the thickness of a pre-form is reduced whilst its length increases without any change in the internal diameter. Flow forming is a simple, chip-less and cost-effective forming process to produce tubular products with a high degree of mechanical and dimensional properties [1,2]. In flow forming a pre-form is fitted to a rotating mandrel. An external force is applied to the outer diameter of pre-form by roller(s); the rollers move along the axis of mandrel to reduce the wall thickness of the preform. The length of part is increased while its thickness decreases due to the law of conservation of volume. The axial movement of rollers is called feed. There are two modes of flow forming associated with the process, backward and forward flow forming [3]. In forward flow forming material flows ahead of the
∗ Corresponding author. Tel.: +98 9187628856. E-mail address: st [email protected] (A. Abedini).
rollers in the same direction as the rollers feed. In backward flow forming material flows in the opposite direction of rollers feed, Figs. 1 and 2 show the modes of flow forming schematically [3]. Ameliorating the mechanical properties, sound finished surface, high dimensional accuracy, simple tool design, low cost and high productivity are some advantages of flow forming. [4,5]. Many experimental, analytical and numerical studies have been presented about metal flow forming. Davidson et al. [6] experimentally studied the effect of flow forming parameters on the surface quality of flow formed aluminum tubes. Roy et al. [7] introduced an analytical model to predict the shape of roller/work piece interface. Molladavoudi et al. [4] established an experimental study to examine the effects of thickness reduction on mechanical properties of aluminum tubes. Although metal flow forming has been studied by many researchers and industries, there does not appear to be any published study on the flow forming of polyethylene pipes. This study was performed to examine the effect of flow forming on mechanical properties of polyethylene pipes.
http://dx.doi.org/10.1016/j.jmapro.2015.06.014 1526-6125/© 2015 The Society of Manufacturing Engineers. Published by Elsevier Ltd. All rights reserved.
156
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162
Fig. 3. Flow forming machine.
Fig. 1. Backward Flow forming [3].
2. Experimental Fig. 4. Details of roller geometry.
2.1. Equipment A single roller backward flow forming machine was used to conduct the experiments. The machine was an NC lathe which was converted to a flow forming machine. A set of roller and mandrel was designed and manufactured to mount on the cross slide of the lathe. A steel mandrel with a male lug was secured by the chuck of lathe on one end and by the tailstock on the other end. Throughout the process the rotation speed of mandrel was applied in RPM (rotation per minute) and feed rate of roller in mm/rev. To apply the thickness reduction an equation was considered as thickness reduction ratio, Eq. (1) represents this ratio. Where to is initial wall thickness of pre-form and tf is the wall thickness of flow formed part. Thickness reduction ratio =
to − tf to
(1)
Table 1 Roller geometrical details. Roller radius
Attack angle
Tip radius
Smooth angle
120 mm
30◦
2.5 mm
6◦
Fig. 3 shows the general set-up of the flow forming machine during forming. Fig. 4 shows the details of roller geometry schematically, Table 1 represents the dimensions of roller. To avoid heating effects during flow forming an intense flow (0.5 l/s) of coolant was used to cool down the parts (roller and mandrel). 2.2. Material The material used for the pipes, was a copolymer of polyethylene HDPE80 in black color with a blue line along the tube to show and magnify the probable torsion of material. Table 2 shows the technical properties of HDPE 80 [8]. The tube specimen had an external diameter of 63 (±0.1)mm, an internal diameter of 51 (±0.1)mm and a length of 200 (±0.1)mm. 2.3. Design of experiment An L9 orthogonal array of Taguchi method was used to conduct and analyze the experiments. Taguchi method is a simple, systematic and rapid design of experiment tool. By using Taguchi method it is possible to study a process with only a small number of experiments. Thickness reduction ratio, rotation speed and feed rate were considered as variable parameters. Three levels Table 2 Technical properties of HDPE 80 [8]. Unit
Fig. 2. Forward Flow forming [3].
Density Yield stress endurance Breaking elongation Elasticity modulus Hardness
Value 3
g/cm MPa % MPa Shore D
0.940 20–22 >600 700–1000 40
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162
157
Table 3 Variables and their levels. Parameter
Low
Medium
High
Thickness reduction (%) Feed rate (mm/rev) Angular speed (rpm)
20 0.16 90
40 0.32 180
60 0.48 360
Fig. 6. Dumb-bell shaped tensile specimens.
Table 5 Measured mechanical properties of normal specimen. Stress at break (MPa) Fig. 5. Flow formed pipes.
including High, Medium and low were specified for each variable. Table 3 shows variables and their levels. Using an L9 orthogonal array as experiments layout needs 9 experiments to be performed. Table 4 shows the layout of this study and the measured mechanical properties of each experiment. The capital letters L, M and H indicate Low, Medium and High level respectively. Fig. 5 shows the flow formed pipes according to L9 orthogonal array. Table 5 shows the measured mechanical properties for normal un-flow formed specimen.
Yield stress (MPa)
10.5
Percentage of elongation at break (%)
19
Impact endurance (J/m)
620
450
standard introduces standard test methods which used to determine the Charpy impact endurance of plastics. A digital Charpy impact tester was used to do the tests. Fig. 8 shows the notched specimens, the Charpy impact tester and a specimen after Charpy test.
2.4. Tensile test
2.6. Scanning Electron Microscopy (SEM)
Dumb-bell shaped specimens were removed from the flow formed pipes according to ISO 6259 3 type 1 standard, using a CNC mill. Fig. 6 shows the prepared tensile specimens. The axis of specimens was parallel to the longitudinal axis of pipes. Tensile tests were carried out on an automatic universal testing machine. The cross-head speed of machine was set at 50 mm/min. The yield Stress and stress at break were defined as the load at specified points divided by the initial cross-sectional area. The tests were carried out in 3 iterations and the arithmetic mean was used to represent the average of measured values for each experiment. Fig. 7 shows the output plot of tensile tests for experiments of L9 orthogonal array.
Rectangular specimens were cut from the flow formed pipes and immerged in liquid nitrogen. The frozen samples were broken along the longitudinal axis of pipes to expose their cross sections. A thin layer of gold (about 100 angstroms) was deposited on the surface of specimens using physical vapor deposition method to avoid the charging effect of surface. A KYKY-SBC12 sputter coater was used to coat the surfaces with gold. The cross sections were observed using a KYKY-EM3200 electron microscope. The images of microstructure were prepared by different magnifications to investigate changing of microstructure due to flow forming.
3. Result and discussion 2.5. Impact test Impact notched specimens were removed from the flow formed pipes according to ASTM 6110-10 standard using a CNC mill. This
The output results of tensile, impact tests and SEM images were analyzed to discover the effect of flow forming parameters on mechanical properties of high density polyethylene pipes.
Table 4 L9 array and measured mechanical properties. RUN
RPM
Feed rate
Thickness reduction
Stress at break (MPa)
Yield stress (MPa)
Percentage of elongation at break (%)
Impact endurance J/m
1 2 3 4 5 6 7 8 9
L L L M M M H H H
L M H L M H L M H
L M H M H L H L M
13 14 16 18 11 20 13 14 13
20.5 21.5 22.0 21.0 23.5 20.5 24.5 21.5 21.5
790 817 817 730 632 790 680 850 820
712 982 1650 1015 2105 1223 3996 1682 3368
158
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162
Fig. 7. Output plots of tensile test.
3.1. Stress at break 3.1.1. Main effects Fig. 9 shows the average of stress at break at each level of process parameters. The plot of thickness reduction ratio vs. stress at break shows a significant improvement in stress at break due to cold flow forming. High density polyethylene is a semi-crystalline polymer and in this type of polymers amorphous and crystalline regions exist simultaneously. The crystalline region is composed of spherulites connected by tie molecules. The amorphous exist in the form of entangled chains, loop chains and floating chains. The main feature in amorphous region is entanglements. Therefore, semi-crystalline polymers consist of an intersecting network of spherulites and entanglements [9]. In flow forming process
material flows in axial, radial and tangential directions through the gap between mandrel and roller but the axial flow is the most dominant. Thus, thickness reduction ratio has an important effect on the plastic deformation of material. During flow forming of semi-crystalline polymers two processes are believed to occur, the spherulitic structure deforms to an ellipsoid and the lamellar structure inside the ellipsoid deforms to a fibrous structure. Usually the change of microstructure shows that an original isotropic structure becomes lamina in shape at moderate deformation and fibril at large deformation. Fig. 10 shows this phenomenon schematically. In this way, the semi-crystalline polymers behave as composites and the increase in stress at break due to cold work can be associated with the reorientation of different regions of polymer.
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162
159
Table 6 ANOVA results for stress at break. Source
DF
SS
MS
P
F0
Percentage
Rank
T. R ratio Feed rate Rotation speed Residual error Total
2 2 2 2 8
42.88 13.55 11.55 0.22 68.22
21.44 6.77 5.77
0.051 0.573 0.515
5.08 0.61 0.74
62.8 19.8 16.9
1 2 3
Table 7 Average and differences between levels for stress at break. Fig. 8. Notched specimens, the Charpy impact tester and a broken specimen.
18 Breakage stress (MPa)
17 16 15
Thickness reduction
14
FEED
13 RPM
12 11 L
M
H
Level Fig. 9. Average of Stress at break at each level of process parameters.
Fig. 11 shows the micrographs of specimens. The un-flow formed sample has an isotropic structure with roughly equal void-like spaces. At low level of thickness reduction ratio the microstructure has no clear transformation but the size of voidlike spaces has increased. At medium level of thickness reduction ratio, the increase of the size of void-like spaces is more visible and their orientations are along the axis of specimen where the most significant flow of material occurs. At high level of thickness reduction ratio, the transition from a spherulitic structure to a lamellar fibrous structure is clearly evident. Fig. 12 shows the fibril structure of highly flow formed sample with magnification of 1000×. The Plot of rotation speed vs. stress at break has an increasing trend, stress at break increases with increase in rotation speed. A possible explanation is that when the rotation speed increases the uniformity of deformation increases that leads to a higher anisotropy in texture of the material. The uniformity in texture makes material more orientated and causes a higher stress at break. According to Tsunekawa et al. [10] the cold forming of an isotropic semi-crystalline polymer usually produces a fibrous material with a high anisotropy of physical properties such as elastic modulus and tensile strength. According to Ingram and Peterlin [11] this anisotropy is closely connected with the molecular orientation
Level
RPM
Feed rate
T.R ratio
L M H (Delta) Rank
13 15 15.67 2.67 3
16 14 13 3 2
12 14.33 17.33 5.33 1
in the crystalline and amorphous regions. The amount of fibrous region in polymer depends on the amount of cold work applied to the material. Therefore, it seems that an increase in rotation speed increases the amount of re-orientation of fibrous region that lead to a higher tensile strength. The plot of stress at break vs. feed rate has a decreasing trend. The high level of feed rate gives the lowest stress at break. It seems at the higher feed rate roller neglects the formation of material and slides over the tubes. This behavior causes difference in mechanical properties of internal layers and external layers hence, yielding process occurs inhomogeneously and the material fails prematurely. 3.1.2. ANOVA results Table 6 shows the ANOVA results for average of stress at break as response function. According to Table 6, thickness reduction ratio is the most significant factor that affects the stress at break. The percentage of contribution of 62.8 for thickness reduction ratio is totally significant. Table 7 shows average of the stress at break at each level and differences between levels of parameters. As it can be observed from Table 7 the difference between the value of stress at break at high level and low level of thickness reduction ratio is totally significant. This means that the amount of deformation is a key factor that affects the yield behavior of HDPE polymer. 3.2. Yield stress 3.2.1. Main effects Fig. 13 shows the average of yield Stress at each level of process parameters. The plot of thickness reduction ratio vs. yield
Fig. 10. Schematic diagram of the flow formed spherulites.
160
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162
Fig. 11. Micrographs of specimens with different thickness reduction ratios.
Fig. 12. Fibril structure of highly flow formed sample.
23.5
Yield Stress (MPa)
23 22.5 Thickness reduction
22 21.5
Feed
21 RPM
20.5 20 L
M
H
Level Fig. 13. Average of yield Stress at each level of process parameters.
Stress shows a significant rise in yield Stress. It seems that the re-orientation of chains is the most important phenomenon that affects the yield Stress. It has been recognized that the yield behavior of semi-crystalline polymers is associated with plastic deformation of crystal parts [12] and the morphology transformation from original isotropic structure to an oriented one [13]. Solid state deformation is one of the fundamental processes to induce polymer to achieve high mechanical properties [14]. It is a
useful method to enhance the mechanical properties of polymers by inducing a high degree of orientation during solid state deformation of semi-crystalline polymers [15]. Therefore, as Fig. 12 indicates, it can be concluded that polymer chain orientation plays a vital role in controlling the yield behavior of a semi-crystalline polymer. The plot of rotation speed vs. yield stress also shows an increasing trend. It seems that higher rotation speed applies more cold deformation to the material and makes the material more oriented. As it is mentioned re-orientation of chains is the main factor associated with high tensile strength. The plot of yield stress vs. Feed rate shows that at higher levels of feed rate, the yield Stress decreases. This can be related to the imperfect deformation due to high feed rate. The high feed rate applies load only on the outer layers of material, it can scratch the surface of material and break the polymers chains. These imperfections are suitable places for creating stress concentration that leads to a premature failure.
3.2.2. ANOVA results Table 8 shows the ANOVA results for average of yield Stress as response function. According to Table 8, thickness reduction ratio is
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162 Table 8 ANOVA results for yield stress.
161
Table 10 ANOVA results for percentage of elongation at break.
Source
DF
SS
MS
P
F0
Percentage
Rank
Source
DF
T. R ratio Rotation speed Feed rate Residual error Total
2 2 2 2 8
10.50 2.17 1.17 0.66 14.50
5.25 1.08 0.58
0.021 0.615 0.778
7.87 0.53 0.26
74.21 14.94 8.05
1 2 3
T. R ratio Rotation speed Feed rate Residual error Total
2 2 2 2 8
Table 9 Average and differences between levels.
SS 20,551 13,444 7264 13,737 54,996
MS 10,275 6722 3632
P
F0
Percentage
Rank
0.246 0.431 0.654
1.79 0.97 0.46
37.37 24.24 13.21
1 2 3
Table 11 Average and differences between levels.
Level
RPM
Feed rate
T.R ratio
Level
RPM
Feed rate
T.R ratio
L M H (Delta) Rank
21.33 21.67 22.50 1.17 2
22 22.17 21.33 0.83 3
20.83 21.33 23.33 2.5 1
L M H (Delta) Rank
808.0 760.0 713.3 94.7 2
723.3 792.3 765.7 69.0 3
826.7 739.0 715.7 111.0 1
840 820 Elongaon (%)
800 780
Thickness reduction
760
Feed
740
Impact endurance (J/m)
3400 2900
Feed
1900
RPM
1400
RPM
720
Thickness reduction
2400
900 L
700 L
M
M Level
H
H
Level Fig. 14. Average of percentage of elongation at break at each level of process parameters.
the most significant factor that affects the yield Stress. The percentage of contribution of 74.21 for thickness reduction ratio is totally significant. Table 9 shows average of the yield stress at each level and differences between levels of parameters. As Table 9 indicates thickness reduction ratio has a predominate effect on yield stress of flow formed HDPE.
Fig. 15. Average of Impact endurance at each level of process parameters. Table 12 ANOVA results for impact endurance. Source
DF SS
Rotation speed T. R ratio Feed rate Residual error Total
2 2 2 2 8
MS
618,1002 30,90501 287,0940 14,35470 371,692 18,5846 757,312 10,180,946
P
F0
Percentage Rank
0.061 4.64 60.71 0.370 1.18 28.20 0.894 0.11 3.65
1 2 3
3.4. Impact endurance 3.3. Percentage of elongation at break 3.3.1. Main effects Fig. 14 shows the average percentage of elongation at break at each level of process parameters. In all experiments of L9 orthogonal array the percentage of elongation at break increases in comparison to normal HDPE but the plot of thickness reduction ratio and rotation speed show a significant downward slope. As it was mentioned before the high amount of deformation makes material more oriented and produces greater molecular orientation and compaction of material that leads to a ductile to brittle transformation. It seems that the predominant element of an oriented polymer is the presence of microfibrils. The microfibrils are connected to many tie molecules through the amorphous region and make a fibrous stiffness in the material. 3.3.2. ANOVA results Table 10 shows the ANOVA results for average of percentage of elongation at break as response function. According to Table 10, thickness reduction ratio is the most significant factor that affects the percentage of elongation at break. Table 11 shows the average percentage of elongation at break at each level and the differences between levels of parameters.
3.4.1. Main effects Fig. 15 shows the average impact endurance at each level of process parameters. The Charpy test measures the impact energy or the energy absorbed prior to fracture. The impact behavior of semi-crystalline polymers is complicated because of the presence of the crystalline and the amorphous phases at the same time. Cold state flow forming transforms spherulites into a fibrous structure. As it is shown in Figs. 11 and 12 the fibrous structure consists of oriented lamellae in crystalline phase connected to many tie molecules within the amorphous phase. It seems that three-dimensional entanglement occurs in amorphous regions. Three-dimensional entanglement forms a network by intertwining of neighboring polymer chains. As Fig. 15 indicates at high levels of deformation the material shows a better impact endurance, a reasonable explanation can be the presence of re-orientation, fibrous structure and entanglement of polymer chains. 3.4.2. ANOVA results Table 12 shows the ANOVA results for average impact endurance as response function. According to Table 12, rotation speed is the most significant factor that affects the impact endurance. The percentage of
162
A. Abedini et al. / Journal of Manufacturing Processes 19 (2015) 155–162
Table 13 Average and differences between levels. Level
RPM
Feed rate
T.R ratio
L M H (Delta) Rank
1115 1448 3015 1901 1
1908 1590 2080 491 3
1206 1788 2584 1378 2
contribution of 60.71 is totally considerable. The effect of feed rate is trivial. Table 13 shows the average impact endurance at each level and differences between levels of parameters. 4. Conclusion The effect flow forming parameters on the mechanical properties of high density polyethylene (HDPE) pipes were critically analyzed using Taguchi method. Thickness reduction ratio, feed rate and rotation speed of mandrel were considered as variables. Stress at break, yield Stress, percentage of elongation at break and impact endurance were measured as indicators of mechanical properties. Scanning electron microscopy was performed to find the relation between mechanical properties and microstructure of the material. It was found that mechanical properties of HDPE increase considerably during flow forming. The solid state deformation makes the material more oriented and anisotropic that leads to an enhancement in mechanical properties of HDPE. High density polyethylene is a semi-crystalline polymer and its microstructure consists of amorphous and crystalline regions simultaneously. The crystalline region is composed of spherulites connected by tie molecules, the amorphous regions exist in the form of entangled chains, loop chains and floating chains. During flow forming the spherulitic structure deforms to an ellipsoid and the lamellar structure inside the ellipsoid deforms to a fibrous structure. The amount of fibrous region in polymer depends on the amount of cold work applied to the material. Through scanning electron microscopy, it was found that an original isotropic structure becomes lamina in shape at moderate deformation and fibril at large deformation. Because the
material experiences a transformation from spherulite structure into a fibrous structure at large deformation, the semi-crystalline polymers behave as composites and the amelioration of mechanical properties due to cold work can be associated with the reorientation of different regions of polymer. A high degree of anisotropy in texture of the material was evident at large deformation. Through ANOVA, it was found that thickness reduction ratio is the most significant parameter affecting the mechanical properties of HDPE. References [1] Podder B, Mondal C, Kumar KR, Yadav DR. Effect of preform heat treatment on the flow formability and mechanical properties of AISI4340 steel. J Mater Des 2012;37:174–81. [2] Xia QX, Cheng XQ, Hu Y, Ruan F. Finite element simulation and experimental investigation on the forming forces of 3D non-axisymmetrical tubes spinning. Int J Mech Sci 2006;48:726–35. [3] Wong CC, Dean TA, Lin J. A review of spinning, shear forming and flow forming processes. Int J Mach Tools Manuf 2003;43:1419–35. [4] Molladavoudi HR, Djavanroodi F. Experimental study of thickness reduction effects on mechanical properties and spinning accuracy of aluminum 7075-O, during flow forming. Int J Adv Manuf Technol 2011;52:949–57. [5] Joseph Davidson M, Balasubramanian K, Tagore GRN. Experimental investigation on flow forming of AA6061 alloy—a Taguchi approach. J Mater Process Technol 2008;200:283–7. [6] Joseph Davidson M, Balasubramanian K, Tagore GRN. An experimental study on the quality of flow-formed AA6061 tubes. J Mater Process Technol 2008;203:321–5. [7] Roy MJ, Maijer DM, Klassen RJ, Wood JT, Schost E. Analytical solution of the tooling/work piece contact interface shape during a flow forming operation. J Mater Process Technol 2010;210:1976–85. [8] Tec Industries Inc Polyethylene pipes and fittings supplier, Technical catalog, http://www.tecindustries.com.my/pe gas supply pipe.html, 02/18/2015. [9] Sizhu Wu. The effect of molecular orientation on the mechanical behavior of cold-rolled semi-crystalline polymers. Hung Hom, Hon Kong: Hong Kong Polytechnic University; 2002 (Doctoral thesis submitted). [10] Tsunekawa Y, Oyane M, Kojima K. Effects of irradiation and rolling on the tensile behavior of polyethylene. J Polym Sci 1961;50:35–44. [11] Ingram I, Peterlin A. Electron diffraction and microscopy of deformed polyethylene spherulites. J Polym Sci—Polym Lett 1964;2:739–45. [12] Juska T, Harrison IR. A criterion for craze formation. Polym Eng Sci 1982;22:766–76. [13] Crist B, Fisher CJ, Howard PR. Mechanical properties of model polyethylene tensile elastic-modulus and yield stress. J Macromol 1989;22:1709–18. [14] Shaw MT. Cold forming of polymeric materials. Annu Rev Mater Sci 1980;10:19–42. [15] Katto A, Makayama K, Kanestsuna H. Infrared dichroism and visible—ultraviolet dichroism studies on rolled-drawn polypropylene and polyethylene sheets. J Macromol Sci-Phys B26 1987:281–306.