Jtam-A4.dvi JOURNAL OF THEORETICAL AND APPLIED MECHANICS 53, 4, pp. 837-846, Warsaw 2015 DOI: 10.15632/jtam-pl.53.4.837 EXPERIMENTAL AND NUMERICAL INVESTIGATION OF FRICTION COEFFICIENT EFFECTS ON DEFECTS IN HORIZONTAL TUBE BENDING PROCESS Jalal Taheri Kahnamouei Memorial University of NewFoundland, Canada; e-mail: jtk441@mun.ca A.M. Fattahi Department of Mechanical Engineering, Tabriz Branch, Islamic Azad university, Tabriz, Iran Theaimof this paper is to investigatedefects in a thin-walled tube bendingprocess (without using mandrel and booster) and effects of friction between the dies and tube on wrinkles. In the tube bending process, there are several effective parameters such as wall thickness, outer diameter-to-wall thickness ratio, centerline bending radius-to-outer diameter ratio and friction coefficient. Any mismatch in the selection of the process parameters would cause defects inducing undesirable variations in wall thickness and cross-section distortion. In this work, firstly, tubes with several wall thickness values are bent, and the final depths of wrinkling and wall thickness change are reviewed. Then, to study the process numerically, numerical simulations are carried out. Then, a series of experimental tests are carried out to verify the simulation results. A comparison between numerical and experimental results shows a reasonable agreement. Finally, in order to obtain a suitable friction condition, the effects of friction coefficients on defects are studied. For this purpose, a series of simulations has been carried out. It shows that at a certain friction coefficient, a minimum wrinkling depth can be observed and variations in the friction coefficient between the dies and tube has no effective influence on wall thinning and thickening. Keywords: tube bending, wrinkle, simulation, friction, wall thickness change 1. Introduction Recently, curved thin-walled tubular elements have been attracting more applications in au- tomobile, aerospace, and oil industries. Tube bending is used as a hydroforming process and application that needs high strength/weight ratio products (Manabe andAmino, 2002; Koc and Altan, 2001). There are some parameters to be controlled in order to reduce the defect. For example, geometry parameters and friction conditions aremodified to restrain instability of the process. In the tube forming process, wrinkling can be avoided, but the wall thickness change is almost inevitable. From among the considerable number of studies dealing with the wrinkling and thinning of wall thickness in tube bending, only few studies have focused on methods to reduce the defect. Tang (2000) employed plastic-deformation theory to investigate the plastic deformation in pipe and tubebending and also explained the seven phenomena in tubebending, also mentioning their practical formulas. An experimental sample was also tested to illustrate that the results of the formulae are very similar to the experimental results. Yang et al. (2006) investigated the effect of friction on the cross-section quality of thin-walled tube NC bending. His results showed that the effects of frictions between all dies and tube on wall thinning are smaller than their effects on section distortion.Therefore, in order to improve the section quality, frictions between mandrel, wiper and tube should be decreased, but the frictions between the pressure die, bending die and tube should increase. Gaoa and Strano (2004) made a research 838 J. Taheri Kahnamouei et al. on the effect of friction on the quality of tube pre-bending and hydroforming. In that paper, process variables such as the friction coefficient, tube material and pre-bent tube radius were analyzed. It was found that a lower friction coefficient can reduce thinning in the pre-bending process, and that a large pre-bending radius is beneficial to both pre-bending and subsequent hydroforming. Zeng and Li (2002) introduced a tube push-bending process combining axial forces and internal pressure.Moreover, they also made research on effects of the internal pressure, friction condition andpushdistance on the tubedeformationpush-bendingprocess.Yang andLin (2004) studied thewrinkling in the tubebendingprocess,where the effects of bendingangle, geometrical dimensions, material properties and the original radius and strength coefficient of tubes on the minimumbending radiuswere analyzed. The role of the fillingmaterial on defects in the of thin- -walled tubebendingprocesswas reportedas anumerical and experimental studybySedighi and Taheri Kahnamouei (2014). That paper investigated approaches to avoid common defects such as the wrinkling, cross section distortion and wall thickness variation in the bending process of a thin-walled tube. So, a series of experimental tests was carried out by filling the tube with melted lead and different types of rubbers. They showed that wrinkle initiation and cross section distortion can be avoided with a lost core made of low temperature melting metal like lead or tin. Jiang et al. (2011) used a three-dimensional finite element method to investigate deformation behavior of medium-strength TA18 high-pressure tubes during NC bending with differentbendingradii.Thearticle showed that if amandrel is used, the thickening ratio increases from the initial bending section to the bending section. In the present study, firstly, the wrinkle phenomenon and the wall thickness change in a thin-walled tube are studied using theoretical, experimental and FEmethods. In order to verify the finite element analysis results, some experiments have been done and comparisons drawn between the experimental and numerical results in Section 2. In Section 3, FEM analysis of effects of friction between the dies and tube on these defects is studied. Finally, in Section 4 the results are presented and discussed. The results show that at a certain friction coefficient a minimumwrinkling depth can be observed and variations in the friction coefficient between the dies and tube have no effective influence on the wall thinning and thickening. 2. Methodology 2.1. Analytical approach Changes inwall thickness is related tomany factors, suchasmaterial, size and shape,bending and wall factors, bending method, tooling and bending operation. When a tube is bent, then tensile or compressive stresses cause wall thinning or thickening. Thewrinkling in the inner wall of the bending radius is one of the common defects in the tube bending process. It has a direct relation with the outer diameter-to-wall thickness ratio and the centerline bending radius-to- outer diameter ratio (Fig. 1). Fig. 1. Example of a wrinkled bending section of a tube and its parameters (Yang and Lin, 2004) Experimental and numerical investigation of friction coefficient effects... 839 Based on the known energy principle, the critical moment of wrinkling onset is when the internal energy of thewrinkled shell (U) is equal to thework doneby the external forces (T), and the wrinkling happens if the external forces are larger than the internal energy of the wrinkled shell (Yang and Lin, 2004). According to the wrinkling wave function proposed by Wang and Cao (2001) and Yang and Lin (2004), the depth of wrinkling in the normal direction w can be characterized with the following function w = w0 √ Rcosθ R0 ( 1−cos 2πmϕ φ1−φ0 ) w0 = √ r0R0(φ1−φ0) πm (2.1) where r0 is the tube diameter, R0 is the bending radius, m is the wave number along the circumferential direction of the tube, ϕ is the curve coordinate in the tube bending direction which changes from φ1 to φ2. When a tube is bent, two typical stress zones can be defined. One is the tension zone at the extrados of the bend; the other is the compressive zone at the intrados of the bend. These cause tube thickening at the intrados and thinning at the extrados, respectively, as shown in Fig. 2. Fig. 2. Cross section of tube after bending Based on plastic deformation theory, equations (2.2) give the rate of wall thinning and wall thickening for the extrados and intrados of the bend which has been proposed by Tang (2000) t0 max = ( 1− r0 4R0 ) t ti max = ( 1+ 2R0r0+3r20 8R20 ) t (2.2) where t is tube wall thickness. 2.2. Material properties and geometry Auniaxial tension test is used to obtainmechanical properties of steel (USt37) and is listed in Table 1. The hardening behavior can be described by equation σ = K(ε)n. The Coulomb friction model is used in the simulation process. Table 1.Mechanical properties of the tube Poisson’s ratio 0.3 Maximum elongation [%] 44.2 Elasticity modulus [GPa] 210 Yield stress [MPa] 270 K 345 n 0.05 The tooling parameters are shown in Table 2. Circular tubes of diameter r0 = 50mm, wall thickness of t0 =1.5, 1.25, 0.9mm, bending radius 150mm and bending angle 45◦ were used in the experiments. 840 J. Taheri Kahnamouei et al. Table 2. Tooling geometry dimensions Tooling parameter Length [mm] Dimension [mm] Bend die 300(D) 52 Rotary die 150 52 2.3. Experimental setup The experiment is performed to provide a general concept of the tubebendingprocess and to verify the FEmodeling. For this purpose, a hydraulic horizontal tube bendingmachine has been used. The horizontal bending is widely used for bending tubes, particularly for tight bending radii and thin wall tubes. The dies setup on the horizontal bendingmachine is shown in Fig. 3. Thework piece is held between the bend die and rotary dies. The bend die strokes linearly and, synchronically, rotary dies are actuated to rotate on the tube, and the work piece bends to any requested angle. Fig. 3. Sketch of the tube horizontal bending 2.4. The FE model A 3D finite element model is built using ANSYS software. The tube is modeled by shell 143 which is well suited to model nonlinear, flat or wrapped, thin to moderately thick shell structures. Three dimensional rigid elements for the dies models have been used. Shell 143 is 4-node 3D space shell element and has six degrees of freedom at each node: translations in the nodal x, y, and z-directions and rotations about the nodal x, y, and z-axis. The geometry, node locations, and the coordinate system for this element are shown in Fig. 4 (ANSYSHelp, 2007). The Coulomb friction coefficient between the tube and dies are assumed to be equal to 0.15 (Trana, 2002) andall dies and tube geometric parameters are the sameas used in the experimen- tal setup. In the tube bending process, there are three contact surfaces between the tube/bend die, tube/rotary dies. The “surface-to-surface contact” method has been employed to describe the mechanical constraints for different contact pairs using CONTACT174 and TARGET169. Figure 5 shows a representative finite element model with the initial tube blank and the tool set. Experimental and numerical investigation of friction coefficient effects... 841 Fig. 4. Shell (143) geometry (ANSYS Help, 2007) Fig. 5. FEmodel for the horizontal bending process 3. Experimental verification of the numerical model In order to verify the finite element analysis results, some experiments have been done. InFig. 6, a comparison between the experimental and numerical results are presented. No wrinkling can be seen in the tube with 1.5mm thickness but in the tube with 1.25mm and 0.9mm thickness wrinkling is observed. Also thewrinkling depth in the experimental andFEmodeling results is compared inFig. 7. It can be seen that the amount of wrinkling in the tube is in direct correlation with the tube wall thickness. An increasing in the wall thickness decreases the possibility of wrinkling. From Fig. 6 and Fig. 7 it can be concluded that the FE results are reasonable and can be used for further investigation of the friction coefficients effect on wrinkling. FEM simulations and experimental results for the extrados wall thickness at the bend zone have been compared and they are shown in Fig. 8. It can be seen that the wall thickness at the outside of the bend is always decreased. Also by an increase in d/t ratio, the thinning value decreases. Themaximum thinning reduction in 1.5mm tube is equal to 6.5% and theminimum thinning reduction in 0.9mm tube is equal to 3%. It can also be observed from Fig. 8 that the wall thickness has a great influence on the thinningwhen the bending radius and tube diameters are fixed. Also experimental and FEM results for the intrados wall thickness at the bend zone have been compared and shown in Fig. 9. It can be seen that the thickness at the inside of the bend is increased, and the maximum thickness thickening reduction is equal to 10%. 842 J. Taheri Kahnamouei et al. Fig. 6. Verification of the FE results by experiments.Wall thickness: (a) 1.5mm, (b) 1.25mm, (c) 0.9mm Fig. 7. Comparison of FE and experimental results Fig. 8. Comparison of the thickness distribution in FE and experimental results at the extrados radius (without considering hardening and friction parameters) Experimental and numerical investigation of friction coefficient effects... 843 Fig. 9. Comparison of the thickness distributions in FE and experimental results at the intrados radius (without considering hardening and friction parameters) 4. Results and discussion The friction conditions between the bend die, rotary die and tube have a large effect on the wrinkling in thin-walled tube horizontal bending, especially for small R/D and large D/t. In order to study the effects of friction on defects after verification of the FE model, a set of runs have been implemented. There are two different cases of friction in the horizontal bending process. One refers to friction between the bend die and tube; another refers to friction between the tube and rotary dies. For this purpose, two series of friction conditions are employed. These runs include values of the friction coefficient between the benddie and tube as 0.05, 0.1, 0.2, 0.3, 0.4, 0.5 and friction between the rotary dies and tube as 0.05, 0.1, 0.2, and 0.3. In the following, the obtained results will be discussed for four different subsections separately. 4.1. The effect of friction between rotary dies and tube on wrinkling TheFEmodeling of the bending process for different amounts of friction between the rotary dies and tube have been carried out. Variations in the friction coefficient and wrinkling depths are shown in Fig. 10. It shows that the depth of wrinkling for all three thicknesses are decreased when the friction coefficient is decreased from 0.3 to 0.2 and the depth of the wrinkling is increased when the friction coefficient is reduced from 0.2 to 0.05. This shows that the least wrinkling depth will happen at a certain friction coefficient. Fig. 10. Variations in wrinkling depth vs. friction coefficient between the tube and rotary dies 4.2. The effect of friction between rotary dies and tube on the wall thickness change The effects of the friction coefficient between the rotary die and tube on the wall thickness are shown in Figs. 11 and 12. It can be found that it has no influence on the thinning and thickening of the tube wall. 844 J. Taheri Kahnamouei et al. Fig. 11. Variations in wall thinning vs. friction coefficient between the tube and rotary dies (without considering hardening and friction parameters) Fig. 12. Variations in wall thickening vs. friction coefficient between the tube and rotary dies (without considering hardening and friction parameters) 4.3. The effect of friction between bend die and tube on the wrinkling Variations in the friction coefficient andwrinkling depths for different thicknesses (1.5, 1.25, 0.9mm) are presented in Fig. 13. It shows that for all three thicknesses, the depth of wrinkling is decreased when the friction coefficient is reduced from 0.5 to 0.2. But when the friction coefficient is reduced from 0.2 to 0.05, the depth of wrinkling increases. It is found that the minimumwrinkling depth takes place at friction coefficient of 0.2. Fig. 13. Variations in wrinkling depth vs. friction coefficient between the tube and bend die 4.4. The effect of friction between bend die and tube on the wall thickness change The effects of the friction between the bend die and tube on the wall thickness are shown in Figs. 14 and 15. Same as in the previous Subsection, it can be found that friction between the bend die and tube has no influence on the thinning and thickening of the tube wall. Experimental and numerical investigation of friction coefficient effects... 845 Fig. 14. Variations in wall thinning vs. friction coefficient between the tube and bend die (without considering hardening and friction parameters) Fig. 15. Variations in wall thickening vs. friction coefficient between the tube and bend die (without considering hardening and friction parameters) According to the results given in these figures, it can be concluded that friction between the tube, bend die and rotary dies have a significant and important effect on the wrinkling, and have no influence on the thinning and thickening of tube wall thickness in the tube horizontal bending process. 5. Conclusion A 3D FE model has been created to study the effect of friction on defects in the bending process.TheFE results have beenverifiedby experimental tests and they are in good agreement. According to the results of the analysis, it can be concluded that: • Theminimumwrinkling depth occurs at a certain value of the friction coefficient. • Friction conditions may affect the balance of internal energy between the wrinkled shell and the work done by the external forces. So, there should be a certain friction coefficient value for which a stable state exists between the two mentioned energies. • Extremely low or high friction conditions are detrimental for the tube bending process. A perfect tubular part can be obtained at a suitable friction condition. • Variations in the friction coefficient between the dies and tube have no influence on the thinning and thickening of the tube wall. Finally, it should be noticed that there are other effective parameters in the tube bending such as bending speed, bending radius. Future works should address these factors in detail. 846 J. Taheri Kahnamouei et al. References 1. ANSYS Inc., “ANSYSHelp”, Release 11.0, Documentation, Copyright 2007 2. Gaoa L., Strano M., 2004, FEM analysis of tube pre-bending and hydroforming, Journal of Materials Processing Technology, 151, 294-297 3. Jiang Z., Zhan M., Yang H., Xu X., Li G., 2011, Deformation behavior of medium-strength TA18 high-pressure tubes during NC bending with different bending radii, Chinese Journal of Aeronautics, 24, 657-664 4. Koc M., Altan T., 2001, an overall review of the tube hydroforming (THF) technology, Journal of Materials Processing Technology, 108, 384-393 5. Manabe K., Amino M., 2002, Effects of process parameters andmaterial properties on deforma- tion process in tube hydroforming, Journal of Materials Processing Technology, 123, 285-291 6. SedighiM., TaheriKahnamouei J., 2014,Role of fillingmaterial on defects of thin-walled tube bending process, Journal of Theoretical and Applied Mechanics, 52, 1, 227-233 7. TangN.C., 2000, Plastic-deformationanalysis in tube bending, International Journal of Pressure Vessels and Piping, 77, 751-759 8. TranaK., 2002, Finite element simulation of the tube hydroforming process-bending, performing and hydroforming, Journal of Materials Processing Technology, 127, 401-408 9. Wang X., Cao J., 2001,Wrinkling limit in tube bending,Trans. ASME, 123, 430-435 10. Yang H., Lin Y., 2004, Wrinkling analysis for forming limit of tube bending processes, Journal of Materials Processing Technology, 152, 363-369 11. Yanh H., Gu R.-J., Zhan M., Li H., 2006, Effect of frictions on cross se ction quality of thin- walled tube NC bending,Transactions of Nonferrous Metals Society of China, 16, 878-886 12. Zeng Y., Li Z., 2002, Experimental research on the tube push-bending process, Journal of Mate- rials Processing Technology, 122, 237-240 Manuscript received December 2, 2013; accepted for print April 17, 2015