HYDROFORMING OF TUBULAR MATERIALS AT VARIOUS TEMPERATURES
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Doctoral Degree of Philosophy in the Graduate
School of the Ohio State University
By
Yingyot Aue-u-lan, M.S
* * * *
The Ohio State University 2007
Dissertation Committee: Approved by
Professor Taylan Altan, Adviser
Professor Gary Kinzel ______Advisor Associate Professor Jerald Brevick Industrial and Systems Engineering Graduate Program
Copyright by Yingyot Aue-u-lan 2007
ABSTRACT
This dissertation research covered two main areas in tube hydroforming process.
The first was to develop the methodology to determine the flow stress directly from the
tube at room temperature. The hydraulic bulge test was selected for this purpose, because it emulates the real state of stress (biaxial state of stress) occurring during hydroforming.
Dimensions of the hydroformed tube were used to calculate the flow stress. The analytical model based on an incremental strain theory (non-proportional strain path) was used to predict the wall thickness at the apex of the dome and curvature radius. The thickness predictions were compared with the measured data. The agreement was good.
The application of the hydraulic bulge test was extended for use as a tool for a quality control of incoming tubular materials. The experiments were performed to investigate the variations in formability of the tubes due to the tube manufacturing processes (rolling process to produce a sheet and roll forming to bend the sheet to form the tube). Different criteria (maximum bulge height (h), strain hardening exponent (n) and maximum percentage thinning) were evaluated to determine the sensitivity of the material property variations to manufacturing processes. The maximum bulge height at the bursting pressure was found to be the most sensitive variable.
ii The second portion of this research was to develop a prototype tube hydroforming system that could be used to form lightweight alloy tubes (aluminum and magnesium
alloys) at elevated temperatures. The existing knowledge on process development,
especially in equipment and process designs, for forming these materials at the elevated
temperature was not sufficient. Therefore, a new design approach called “submerged
concept”, was developed to reduce the heating and filling time and maintain uniform
temperature in the tube during hydroforming.
The prototype tube hydroforming system was used to investigate the effect of the
tube extrusion processes (with mandrel –seamless and with porthole die –with seams) on
the quality of tubes. Seamless extruded tubes were studied extensively regarding the effect of the process parameters (forming temperatures and forming rates) on the formability and loading behavior (internal pressure). The tubes with seams were found to have defects at the welding line that caused fracture during hydroforming. The results indicated that formability increases with increasing temperature. The forming pressure dropped before the tube touched the die surface, indicating of strain softening. Tensile test was used to obtain the flow stress of the tubes at different temperatures (100, 150,
200 and 250oC) and strain rates (0.001, 0.01 and 0.1 /s). These flow stress data were used
in Finite Element simulations to predict process variables, i.e. pressure and axial feed
versus time. The comparison between the simulation and experimental results showed
reasonable agreement.
iii
DEDICATION
This work is dedicated to my parents and my family (sisters) for their
encouragements and supports.
iv
ACKNOWLEDGEMENT
I wish to thank my advisor, Taylan Altan, for intellectual support, encouragement, and enthusiasm, which made this thesis possible, and for his patience in correcting both my stylistic and scientific errors.
I also would like to thank my committee members, Prof. Gary Kinzel and Prof.
Jarald Brevick, for their support. I would like to thank the Tube Hydroforming
Consortium members of ERC/NSM and the Department of Energy (DOE) that supported
this study.
I would like to thank to Mr. David Guza and Prof. Naksoo Kim for their
suggestions and support. Also I would like to thank to my colleagues; Jon Ander Esnaola,
Christopher Muelberg, Karan Shah, Shrinivas Patil, Dr. Gracious Ngaile for their technical support and the colleagues at the ERC/NSM (Euyene Yen, Suwat Jirathearanat,
Jay Patchapol Sartkuvanich, Hari Palaniswamy, Giovani Spampinato and Hyunjoong
Cho) for their assistance and encouragement.
Finally, I would like to thank my parents, who live in Thailand. Without their
support, my studies would not have been possible. Their support made me pursue my
graduate studies and overcome many difficulties over the years.
v
VITA
October 18, 1973………………………….Born –Chachoengsao, Thailand
1995……………………………………….B.S. Mechanical Engineering, King Mongkut
Institute of Thonburi, Thailand
1995 -1997…………………………………Plant Engineer,
……………………………………………..Mercedes Thailand, Bangkok, Thailand
1997-2000………………………………….M.S. Mechanical Engineering,
The Ohio State University, Ohio, USA
2000-present……………………………… Graduate Research Associate,
ERC/NSM at The Ohio State University
PUBLICATIONS
Research Publication
1. Aue-u-lan, Y., Altan, T., (2002) “Process Simulation for Hydroforming Components from Sheet and Tube. How can we improve the accuracy of the predictions?”, Proceedings of Hydroforming for Car Body Components, The 3rd Chemnitz Car Body Colloquium, 2002, pp. 117-132 2. Aue-u-lan, Y., Ngaile, G., Altan, T., (2004), “Optimizing tube hydroforming using process simulation and experimental verification”, Journal of Materials Processing Technology, vol. 146, issue 1, pp. 137-143
vi 3. Aue-u-lan, Y., Guza, D., Marks, S., Shah, K., Muckatira, T., Altan, T., (2004) “Hydroforming lightweight aluminum and magnesium components from tube – Development and commercialization of a novel elevated temperature hydroforming system- Annual Phase II Report-, Report number: DE-FG02- 02ER86141-Phase II Interim report, submitted to Department of Energy 4. Choi, Y., Aue-u-lan, Y., Guza, D., Shah, T., Altan, T., 2003 “Hydroforming lightweight aluminum and magnesium components from tube –Development and commercialization of a novel elevated temperature hydroforming system- Semi-Annual Phase II Report-, Report number: DE-FG02-02ER86141-Phase II Interim report, submitted to Department of Energy 5. Koc, M., Aue-u-lan, Y., Altan, T., 2001 “On the characteristics of tubular materials for hydroforming design rules, Analysis and Experimentation”, International Journal of Manufacturing and Machine Tools vol. 41, no. 5, pp. 761-772 6. Altan, T., Palaniswamy, H., Aue-u-lan, Y., 2005 “Tube and sheet hydroforming –Advances in material modeling, tooling and process simulation”, Advanced Materials Research, vols. 6-8 (May 2005), pp. 1-12 7 Altan, T., Muammer, K., Aue-u-lan, Y., and Tibari, K., “ Formability and design issues in tube hydroforming”, International Conference on Hydroforming, Oct / 12 – 13/ 1999, Fellbach Stuttgart, Germany, pp.1217 – 1222 8. Altan, T., and Aue-u-lan, Y., “Tube & Sheet Hydroforming: What's new and important?" presented at the SME Hydroforming Conference, Sept. 2005, Chicago, IL. 9. Altan, T., Kaya, S., and Aue-u-lan, Y., 2007 “ Forming Al and Mg alloy sheet and tube at elevated temperatures”, Advanced Materials Research (12th International Conference on Sheet Metal), Palermo, April 1st -4th, 2007
FIELDS OF STUDY
Major Field: Industrial and Systems Engineering
Minor File: Operation Research and Machine Design
vii TABLE OF CONTENTS
Page
ABSTRACT...... ii
DEDICATION...... iv
ACKNOWLEDGEMENT ...... v
VITA...... vi
TABLE OF CONTENTS...... viii
LIST OF TABLES...... xii
LIST OF FIGURES ...... xiv
NOMENCLATURE ...... xxiv
CHAPTER 1 INTRODUCTION/ MOTIVATION ...... 1
1.1. Characterization of material properties for tubular materials...... 1 1.2. Forming of lightweight alloys at elevated temperatures...... 2
CHAPTER 2 RESEARCH OBJECTIVES...... 6
CHAPTER 3 STATE OF THE ART...... 9
3.1. Tube hydroforming (THF) process...... 9 3.1.1. Tube hydroforming process as a system...... 11 3.2. Tube manufacturing processes...... 12 3.2.1. Tube manufacturing by roll forming ...... 13 3.2.2. Extrusion process...... 15 3.3. Mechanical tests...... 17 3.3.1. Tensile test...... 17 3.3.2. Summary of the standard test for tubular materials...... 19 viii 3.4. Variation of material properties...... 21 3.5. Forming technology at elevated temperatures of lightweight alloys.... 26
CHAPTER 4 DEVELOPMENT OF AN ANALYTICAL MODEL TO DETERMINE FLOW STRESS OF TUBES ...... 31
4.1. Description of hydraulic tube bulge test ...... 34 4.2. Constitutive models ...... 37 4.3. Development of analytical model ...... 39 4.3.1. Membrane theory...... 39 4.3.2. Calculation of the curvature radius at the apex of the dome ...... 40 4.3.3. Relationship between incremental strains along hoop and longitudinal directions...... 40 4.3.4. Calculation of wall thickness at the apex of the tube ...... 42 4.4. Procedure to determine flow stress by using an analytical model...... 54 4.5. Flow stress determination of Stainless steel AISI 304...... 55 4.5.1. Dimensions and mechanical properties ...... 55 4.6. Experimental results and flow stress determination of Low Carbon Steel grade AISI 1008 ...... 59 4.7. Conclusions...... 63
CHAPTER 5 INVESTIGATION OF THE EFFECT OF MANUFACTURING PROCESS UPON TUBE QUALITY ...... 65
5.1. Experimental procedure...... 66 5.1.1. Material used in this study...... 66 5.1.2. Experimental matrix ...... 67 5.2. Experimental results...... 69 5.2.1. Experimental results of tube set no. 1...... 69 5.3. Discussions ...... 74 5.3.1. Effect of sheet properties used to manufacture the tubes ...... 74 5.3.2. Effect of the roll forming and welding processes...... 77 5.4. Summary and conclusions ...... 79
CHAPTER 6 DESIGN OF WARM TUBE HYDROFORMING SYSTEM...... 80
6.1. Design considerations ...... 80 6.2. Proposed design for warm tube hydroforming process ...... 84 6.3. Part selections ...... 87 6.4. Specification of warm tube hydroforming system...... 89 6.4.1. Determination of maximum flow rate and volume required to form the selected part ...... 89 6.4.2. Axial feed control ...... 93 6.4.3. Estimation of heating system unit ...... 94 6.4.4. Fluid selection...... 98
ix 6.5. Design of the heating channels ...... 101 6.6. Insulation...... 103 6.7. Thermal analysis of the forming die ...... 106 6.7.1. Geometric Modeling and Boundary Conditions of the Die...... 106 6.7.2. Thermal properties...... 108 6.7.3. Determination of heat transfer coefficient...... 110 6.7.4. Finite Element Modeling (FEM) ...... 112 6.8. Stress analysis ...... 117 6.8.1. Finite element model ...... 117 6.8.2. Simulation results ...... 118 6.9. Temperature measurement...... 121 6.9.1. Set 1: Die temperature measurement without any fluid in the tank 122 6.9.2. Set 2: Temperature measurement of the tube under submerged condition……...... 128 6.10. Design of heating system...... 135
CHAPTER 7 INVESTIGATION OF PROCESS PARAMETERS BY USING THE PROTOTYPE WARM TUBE HYDROFORMING SYSTEM...... 139
7.1. Experimental conditions ...... 139 7.2. Investigation of tube manufacturing process on a quality of incoming tubes…………… ...... 140 7.2.1. Magnesium alloy (AZ31B) tubes (seamed tube)...... 140 7.2.2. Aluminum alloy (AA6061) tubes (seamless tube) ...... 141 7.2.3. Experimental results ...... 141 7.3. Investigation of the effect of process parameters of AA6061-O ...... 145 7.3.1. Experimental procedures and conditions...... 145 7.4. Experimental results...... 148 7.4.1. Effect of the temperature on the formability ...... 148 7.4.2. Effect of the temperature on the pressure profiles...... 150 7.4.3. Forming of AA 6061-O Aluminum tubes at 230 °C with axial feed………………...... 151 7.5. Summary and discussions...... 156
CHAPTER 8 FLOW STRESS DETERMINATION OF AA6061-O AT ELEVATED TEMPERATURES ...... 160
8.1. Tensile test ...... 160 8.1.1. Test procedures and conditions ...... 160 8.1.2. Flow stress results...... 161 8.1.3. Constitutive models ...... 168
CHAPTER 9 FINITE ELEMENT MODELING FOR WARM TUBE HYDROFORMING PROCESS...... 175
x 9.1. Finite element model and boundary conditions...... 175 9.2. Simulation results...... 184 9.2.1. Results at the forming temperature of 230oC without axial feed ... 184 9.2.2. Results at the forming temperature of 250oC without axial feed ... 186 9.3. Results at the forming temperature of 230oC with axial feed...... 189 9.3.1. Simulation results ...... 189 9.3.2. Comparison of the simulation and experimental results ...... 193
CHAPTER 10 OVERALL SUMMARY AND CONTRIBUTIONS ...... 196
REFERENCES ...... 200
APPENDIX A ANALYTICAL MODELS TO DETERMINE FLOW STRESS BASED ON DEFORMATION THEORY ...... 213
APPENDIX B TEMPERATURE EQUIPMENT TO MEASURE THE DIE AND TUBE TEMPERATURE ...... 219
APPENDIX C HEATING MEDIA EVALUATION FOR WARM TUBE HYDROFORMING PROCESS...... 226
APPENDIX D PROCESS SEQUENCE FOR SUBMERGED DESIGN CONCEPT . 236
APPENDIX E SYSTEM FOR MEASURING the BULGE HEIGHTs IN THE FORMING DIE ...... 243
xi
LIST OF TABLES
Page
Table 3.1: List of United States Standards and Specifications applicable to tube hydroforming...... 20
Table 4.2: Material properties and geometry of Stainless Steel 304 specimens. The material properties are applicable to flat sheet metal, using the Holloman’s Law n (σ = Kε ) (source: Honda R&D, 1998)...... 56
Table 4.3: Material properties and geometry of low carbon steel grade 1008 specimens. The material properties are applicable to flat sheet metal, using the Holloman’s Law n (σ = Kε ) (source: LTV steel)...... 60
Table 5.1: Chemical composition of Low Carbon Steel Tube grade AISI 1010...... 66
Table 5.2: Dimensions of the tube used in this study ...... 67
Table 5.3: Experimental matrix ...... 69
n Table 5.4: Flow Stress of LCS 1010 tubing to Krupkowsky’s Law (σ = K(ε 0 + ε) ) obtained from an analytical technique obtained by hydraulic bulge test for tube set no. 1 ...... 71
Table 5.5: Summary of all the flow stress and results of the tube set numbers 2 to 6 obtained from the hydraulic bulge test ...... 73
Table 6.1: Process parameters of the warm hydroforming system...... 94
Table 6.2: Specifications of the heating fluids...... 99
Table 6.3: Material properties of FOAM GLASS insulation ...... 106
Table 6.4: Material properties of H-13 [DEFORM 3DTM database] ...... 108
xii Table 6.5: Thermal properties of Dynalene 600 at various temperatures...... 109
Table 6.6: Material property of H13 at room and 482˚F (250°C) temperature ...... 117
Table 6.7: Computed results compared with the yield values of the tool material...... 121
Table 6.8: Experimental conditions for temperature validations...... 121
Table 7.1: Dimensions and mechanical properties of aluminum alloy tube obtained by tensile test at different temperatures (AA6061-O) [Kaufman, 2002]...... 147
Table 7.2: Comparison of calculated and measured sealing forces used in the experiments ...... 147
Table 7.3: Experimental conditions ...... 148
Table 9.1: Tube geometry and mechanical properties of AA6061-O...... 178
Table B.1: Specifications of the applied thermocouples [www.omega.com]…………………………………………………………………….221
Table B.2: Material properties of high temperature cement “Omega-bond 700” [www.omega.com]……………………………………………………………………. 222
Table C. 1: Experimental results and observations…………………………………… 231
xiii
LIST OF FIGURES
Page
Figure 1.1: Comparison of mechanical properties [Novotny, 2001] ...... 5
Figure 1.2: Warm tube hydroforming as a system –a summary of parameters contributing to the development of warm tube hydroforming process...... 5
Figure 3.1: Tube Hydroforming process sequence; a) place a tube, b) seal and fill fluid, c) pressurize and feed material, d) take a tube out [Leitloff, 1997] ...... 10
Figure 3.2: Examples of the automotive parts manufactured by tube hydroforming process ...... 11
Figure 3.3: System in Tube Hydroforming process...... 12
Figure 3.4: Overview of processes to manufacturing tubes in THF process...... 13
Figure 3.5: Continuous roll forming process to manufacture a steel tube [Singh, 2002].14
Figure 3.6: W- flower sequence for manufacturing the tubes [Gehrish, 1993] ...... 14
Figure 3.7: Schematic illustrates the mandrel extrusion die [Singh, 2002]...... 16
Figure 3.8: (a) an extruded AA6061-T6 cross section for tubes, and (b) –(d) components of various dies for extruding intricate hollow shapes [Kalpakjian, 2001] ...... 17
Figure 3.9: Ring hoop tension test specimen and fixture [Wang, 2001] ...... 18
Figure 3.10: Specimens used for tensile testing of sheet metals. A) Sheet metal specimen, B) Tubular Specimen, C) Longitudinal sectioned specimen for tubular material testing [after ASTM E8-96 (modified)] ...... 19
Figure 3.11 Hardness distribution of roll formed tube [Hielscher, 2001] ...... 23
xiv Figure 3.12 Engineering strain (%) distribution after tube bursting tests [Hielscher, 2001] ...... 24
Figure 3.13: Schematic sketch for the regions from which the tubes are made from the same sheet strip...... 24
Figure 3.14: Influence of the locations (see Figure 3.13) of the sheet used to produce the tubes on the percent elongation of the tube (Test results were conducted by hydraulic bulge test) [Schuler, 2005] ...... 25
Figure 3.15: Tube cross-section showing different testing locations ...... 25
Figure 3.16: Temperature range to identify the forming conditions [Novotny, 2002]..... 27
Figure 3.17: WTHF design concept of University of Darmstadt [Dorr, 2004] ...... 29
Figure 3.18: Schematic of hot gas metal forming equipment [Jager, 2003]...... 29
Figure 3.19: A) a simplified illustration of the heating plant and pressure system, and B) a cross section of the tempered hydroforming-tool (the front view and top view) indicating the heating channels [Neugebauer, 2003] ...... 30
Figure 4.1: State of stress in Hydraulic Bulge Test ...... 32
Figure 4.2 Hydraulic bulge test tooling in the press ...... 35
Figure 4.3 Schematic of hydraulic bulge tooling...... 36
Figure 4.4: Geometry of the deformed tube and the nomenclature used in calculations..36
Figure 4.5: Geometry of the bulge and stress components acting at the apex of the dome. rz, radius of curvature in the axial direction; rθ, radius of curvature in the axial
direction, t, wall thickness, σ θ = hoop stress and σ Z = longitudinal stress...... 37
Figure 4.6: Schematic of geometrical relationships to determine the curvature radius.... 38
Figure 4.7: Relationship of strain along the hoop and longitudinal directions...... 44
Figure 4.8: Schematic of the infinitesimal element at the apex of the dome...... 45
Figure 4.9: Flow chart of thickness calculation...... 46
Figure 4.10: Comparisons of the wall thickness at the apex of the tube among experimental, calculated by incremental theory and deformation theory [Aue-u-lan, 1999]...... 48
xv Figure 4.11: Comparisons of the wall thickness at the apex of the tube among experimental, calculated by incremental theory and deformation theory [Aue-u-lan, 1999]...... 49
Figure 4.12: Comparisons of the wall thickness at the apex of the tube among experimental, calculated by incremental theory and deformation theory [Aue-u-lan, 1999]...... 50
Figure 4.13: Comparisons of the wall thickness at the apex of the tube among experimental, calculated by incremental theory and deformation theory [Aue-u-lan, 1999]...... 51
Figure 4.14: Comparisons of the wall thickness at the apex of the tube among experimental, calculated by incremental theory and deformation theory [Aue-u-lan, 1999]...... 52
Figure 4.15: Comparisons of the wall thickness at the apex of the tube among experimental, calculated by incremental theory and deformation theory [Aue-u-lan, 1999]...... 53
Figure 4.16: Flow chart of flow stress determination by using analytical model...... 55
Figure 4.17: Bulge height versus internal pressure of stainless steel grade AISI 304 obtained from the hydraulic bulge test [OD = 57.15mm (2.25in), initial wall thickness, t0, = 0.61 mm (0.024in), tube length, L0, = 203.2 mm (8.0in) and bulge width, 2w, = 50.8 mm (2.0 in)...... 57
Figure 4.18: Wall thickness at the apex of the dome versus bulge height of stainless steel grade AISI 304 obtained from the hydraulic bulge test [OD = 57.15mm (2.25in), initial wall thickness, t0, = 0.61 mm (0.024in), tube length, L0, = 203.2 mm (8.0in) and bulge width, 2w, = 50.8 mm (2.0 in) ...... 58
Figure 4.19: Comparison of the effective stress versus effective strain of stainless steel grade AISI 304 obtained from the hydraulic bulge test (from tube) and tensile test (of sheet) [OD = 57.15mm (2.25in), initial wall thickness, t0, = 0.61 mm (0.024in), tube length, L0, = 203.2 mm (8.0in) and bulge width, 2w, = 50.8 mm (2.0 in) ...... 59
Figure 4.20: Bulge height versus internal pressure of low carbon steel grade AISI 1008 obtained from the hydraulic bulge test [OD = 88.9mm (3.5in), initial wall thickness, t0, = 2.00mm (0.079in), tube length, L0, = 228.6mm (9.0in) and bulge width, 2w, = 76.2mm (3.0 in) ...... 61
Figure 4.21: Wall thickness at the apex of the dome versus bulge height of low carbon steel grade AISI 1008 obtained from the hydraulic bulge test [OD = 88.9mm (3.5in), initial wall thickness, t0, = 2.00mm (0.079in), tube length, L0, = 228.6mm (9.0in) and bulge width, 2w, = 76.2mm (3.0 in)...... 62 xvi Figure 4.22: Comparison of the effective stress versus effective strain of low carbon steel grade AISI 1008 obtained from the hydraulic bulge test (tube) and tensile test (sheet) [OD = 88.9mm (3.5in), initial wall thickness, t0, = 2.00mm (0.079in), tube length, L0, = 228.6mm (9.0in) and bulge width, 2w, = 76.2mm (3.0 in)...... 63
Figure 5.1: Tube cross-section showing different testing locations ...... 68
Figure 5.2: Pressure vs. bulge Height curves for LCS 1010 tubing; OD = 63.5 mm (2.5 in) and t0 = 2.00 mm (0.079 in) obtained from hydraulic bulge test for tube set no. 1 ...... 70
Figure 5.3: Maximum bulge height and percent thinning at the bursting pressure at different location around the circumference for LCS 1010 tubing; OD = 63.5 mm (2.5 in) and t0 = 2.00 mm (0.079 in) obtained from hydraulic bulge test for tube set no. 1 ...... 71
Figure 5.4: Flow Stresses for set no. 1 of LCS 1010 tubing at the different location around the circumferential directions; OD = 63.5 mm (2.50 in) and t0 = 2.0 mm (0.079 in)...... 72
Figure 5.5: Maximum bulge height at the bursting pressure measured at the different location around the circumferential direction of each tube set...... 75
Figure 5.6: Strain-hardening coefficient (n-value) in each location around the circumferential direction of each tube set ...... 76
Figure 5.7: Maximum percent thinning at different locations around the tube circumference ...... 76
Figure 5.8: Maximum bulge height and percent thinning at different locations around the circumference of Tube set#3 ...... 78
Figure 5.9: Maximum bulge height and percent thinning at different locations around the circumference of Tube set#6 ...... 78
Figure 6.1: Schematic of warm hydraulic bulge test [Patil, 2002] ...... 82
Figure 6.2: Asymmetric expansion due to non-uniform fluid temperature distribution during forming [Patil, 2002] ...... 83
Figure 6.3: Temperature measurement at the tube areas (see Figure 6.2) ...... 83
Figure 6.4: Warm hydroforming system...... 86
Figure 6.5: Section of the designed tool and the names of parts...... 86
xvii Figure 6.6: Schematic of the part selected for this study...... 87
Figure 6.7: Schematic to demonstrate the submerged design concept ...... 88
Figure 6.8: A picture of submerged design concept- Dies are emerged inside the hot liquid bath...... 89
Figure 6.9: Deformed shape of the tube according to the flow rate. The flow rate function is linear as shown in Figure 6.10 and the maximum flow rates are: (a) 1.4 in3/sec, (b) 2.2 in3/sec, (c) 2.8 in3/sec, and (d) 3.6 in3/sec. Only (d) does not make any wrinkle...... 91
Figure 6.10: Flow rate curve of the pressure fluid...... 92
Figure 6.11: Tube internal volume is changing as the tube deforms. This volume is obtained from the simulation...... 92
Figure 6.12: Axial feed speed of the punches...... 93
Figure 6.13: Cross section of the forming dies...... 95
Figure 6.14: Schematic of heating channel for heating the selected die geometry (dimensions in centimeters)...... 103
Figure 6.15: Schematics of the Warm THF-tooling with the installed insulation...... 104
Figure 6.16: Schematics of the insulated tank ...... 105
Figure 6.17: Lower die and quarter die (used for the thermal simulations) ...... 107
Figure 6.18: Boundary conditions used to determine temperature distributions at the die surface (h = convection coefficient, W/m2-K) ...... 107
Figure 6.19: Temperature distributions at the die surface ...... 114
Figure 6.20: Temperature distribution at Time = 25 min at the square section (section A- A, see Figure 6.19) of the die ...... 115
Figure 6.21: Temperature distribution at Time = 25 min at the circular section (section B- B, see Figure 6.19) of the die ...... 115
Figure 6.22: Temperature vs. time curve for point P1 (see Figure 6.19)...... 116
Figure 6.23: Temperature vs. time for point P2 (see Figure 6.19)...... 116
Figure 6.24: Boundary and loading conditions...... 118
xviii Figure 6.25: Stress concentration after applying 5000 psi at temperature of 250°C (A) stress distributions for the whole die and (B) stress distributions at the cross section C-C ...... 120
Figure 6.26: Thermocouples attached in the lower die surface ...... 122
Figure 6.27: Schematic of the thermocouple layout in the lower die ...... 123
Figure 6.28: Thermocouples attached in the upper die...... 123
Figure 6.29: Schematic of the thermocouple layout in the upper die ...... 124
Figure 6.30: Defined profiles in order to identify the thermocouples in the same cross section perpendicular to X direction...... 125
Figure 6.31: Temperature measurements with the error range in the lower die for the different profiles defined in the Figure 6.30...... 126
Figure 6.32: Temperature gradient between the upper and the lower die ...... 128
Figure 6.33: Thermocouples attached to the tube surface ...... 129
Figure 6.34: Locations of each thermocouple attached on the tube ...... 129
Figure 6.35: Thermocouples in the upper die ...... 130
Figure 6.36: Temperature measurements at the tube and upper die surfaces ...... 132
Figure 6.37: Temperature distributions around the circumference at different sections (See Figure 6.34) at the steady state condition of the tube (the measurement error = 3oC)...... 133
Figure 6.38: Schematic of the heating system for warm tube hydroforming process .... 137
Figure 6.39: A picture of the warm tube hydroforming system designed for this study 138
Figure 7.1: Fracture of Mg tube during forming...... 143
Figure 7.2: Fracture and excessive wrinkling when forming Mg tube with axial feed .. 143
Figure 7.3: Thickness distributions measured around the circumferential direction from AA6061 ...... 144
Figure 7.4: Picture of the formed aluminum alloy tube AA6061-O conducted at the temperature of 250oC (482oF) and the volumetric flow rate of 3.28x10-6 m3/s (0.2in3/s) ...... 144
xix Figure 7.5: Tube with clamping rings in the forming die...... 146
Figure 7.6: Picture of experimental results conducted at different forming temperatures ...... 149
Figure 7.7: Maximum percentage expansion of AA6061-O at various forming temperatures (the flow rate = 0.2in3/s) ...... 149
Figure 7.8: Comparison of pressure profiles obtained from the experiment with constant volumetric flow rate at different temperatures ...... 150
Figure 7.9: Loading path obtained by the experiments of forming part at the temperature of 230°C ...... 153
Figure 7.10: Comparison between the forming dies geometry and the measurement of 4 profiles (A, B, C, and D) ...... 153
Figure 7.11: Picture of the formed tube and cross sections...... 154
Figure 7.12: Wall thickness distribution at the rectangular cross section (section A-A in Figure 7.11) ...... 155
Figure 7.13: Measured thickness distribution around the circumferential direction of section B-B (see Figure 7.11]...... 155
Figure 7.14: Internal pressure versus time and bulge height versus time of AA6061-0 at the forming temperature of 250oC (482oF) and flow rate of 1.6x10-5 m3/s (0.98in3/s). The bulge height was measured until the tube touched the die surface...... 158
Figure 7.15: Internal pressure versus time and bulge height versus time of AA6061-0 at the forming temperature of 250oC (482oF), and flow rate of 3.28x10-6 m3/s (0.2in3/s). The bulge height was measured until the tube touched the die surface...... 159
Figure 8.1: Dimensions of tensile specimen [ASTM A513] ...... 161
Figure 8.2: Engineering stress-strain curves obtained from tensile test at 100oC for different strain rates...... 163
Figure 8.3: Engineering stress-strain curves obtained from tensile test at 150oC for different strain rates...... 164
Figure 8.4: Engineering stress-strain curves obtained from tensile test at 200oC for different strain rates...... 165
Figure 8.5: Engineering stress-strain curves obtained from tensile test at 250oC for different strain rates...... 166
xx Figure 8.6: Effect of strain rates and forming temperatures on the uniform elongation of AA6061-O ...... 167
Figure 8.7: Effect of strain rates and forming temperatures on the total elongation of AA6061-O ...... 167
Figure 8.8: Relationship between true stress and strain in the log-log scale at the temperature of 200oC...... 171
Figure 8.9: Relationship between true stress and strain in the log-log scale at the temperature of 250oC...... 171
Figure 8.10: Relationship between true stress and strain rate in the log-log scale at the temperature of 200oC...... 172
Figure 8.11: Relationship between true stress and strain rate in the log-log scale at the temperature of 250oC...... 172
Figure 8.12: Effect of forming temperature on the strength coefficient (K) ...... 173
Figure 8.13: Effect of forming temperature on strain hardening coefficient (n) ...... 173
Figure 8.14: Effect of forming temperature on strain rate hardening coefficient (m) ... 174
Figure 9.1: Simulation model used in this study...... 177
Figure 9.2: Boundary conditions for the simulation ...... 178
Figure 9.3: Flow stress curves at temperature of 230oC for different strain rates (Flow stress was fit up to the uniform elongation, and the dot line represents the extrapolated data) ...... 179
Figure 9.4: Flow stress curves at temperature of 250oC for different strain rates (Flow stress was fit up to the uniform elongation, and the dot line represents the extrapolated data) ...... 179
Figure 9.5: Measured internal pressure vs. time curve obtained at the forming temperature of 250°C with the volumetric flow rate of 1.6x10-5 (0.98in3/s)...... 180
Figure 9.6: Measured internal pressure vs. time curve obtained at the forming temperature of 230°C with the volumetric flow rate of 1.6x10-5 (0.98in3/s)...... 180
Figure 9.7: Loading path obtained by the experimental trials of forming part at the temperature of 230°C ...... 181
Figure 9.8: Measured tube profile along the longitudinal direction obtained at the forming temperature of 230°C ...... 181 xxi Figure 9.9: Measured tube profile along the longitudinal direction obtained at the forming temperature of 230°C ...... 182
Figure 9.10: Location of the profile extracted to be used for flow stress determination 182
Figure 9.11: Measured tube profile along the longitudinal direction obtained at the forming temperature of 250°C ...... 183
Figure 9.12: Comparison between the forming dies geometry and the measurement of 4 profiles (A, B, C, and D) ...... 183
Figure 9.13: Comparison of the deformation between the FEM and experimental results at temperature of 250°C ...... 185
Figure 9.14: Comparison of the displacement profile between the experimental and simulation results at temperature of 250°C ...... 185
Figure 9.15: Comparison of the thickness distribution between the experimental and simulation results at temperature of 250°C ...... 186
Figure 9.16: Comparison of the deformation between the FEM and experimental results at temperature of 230°C ...... 187
Figure 9.17: Comparison of the displacement profile between the experimental and simulation results at the temperature of 230°C ...... 188
Figure 9.18: Comparison of the thickness distribution between the experimental and simulation results at temperature of 230°C ...... 188
Figure 9.19: Contour plot of the effective strain distribution at the final stage of the simulation ...... 189
Figure 9.20: Wall thickness distribution along the circumferential direction at the rectangular cross section...... 190
Figure 9.21: Deformation behavior at different stages ...... 191
Figure 9.22: Simulation result demonstrating a small wrinkle at the formed tube...... 193
Figure 9.23: Comparison of the displacement profile between the simulation and the averaged experimental results ...... 194
Figure 9.24: Comparison of wall thickness distribution between the simulation and experimental results at the rectangular cross section ...... 195
xxii Figure A.1: Geometry of the deformed tube: nomenclature used in calculations…………………………………………………………………………….. 214 Figure A.1: State of stress on an element at the apex of the hydroformed tube………. 218
Figure B.1: Process of attachment for the thermocouple junction on the die surface…………………………………………………………………………………..222
Figure B.2: National Instrument’s “Channel Wizard” used to select the thermocouple signal……………………………………………………………………………………224
Figure B.3: LabView® environment to acquire measurement data……………………. 225
Figure C. 1: Schematic of experimental set-up to test the fluid………………………. 228
Figure C.2: Photograph of the experimental set-up…………………………………….229
Figure C.3: Photograph of the fluid and heating coil in the beaker…………………….230
Figure C.4: Temperature vs. Time plot for Calflo™ HTF…………………………….. 232
Figure C.5: Temperature vs. Time plot for Dynalene™ 600…………………………...233
Figure C. 6: Temperature vs. Time plot for Dow Corning 550………………………... 234
Figure E.1: Mechanism of measurement components used to measure the bulge height as a function of time……………………………………………………………. 243 Figure E.2: Picture to show the fluid release channels………………………………… 244
Figure E.3: Schematic to demonstrate the measurement system used to measure the bulge height in the forming die…………………………………………………………244
xxiii NOMENCLATURE
SYMBOL DESCRIPTION
εZ = longitudinal strain
εθ = circumferential strain
εt = thickness strain
li = incremental tube length
lo = original tube length
W = bulge width
l2 = distance below die surface where tube is locked in axial direction
ti = incremental tube thickness
to = original tube thickness
Re = die corner radius
rI = incremental tube radius
ro = original tube radius
σ = flow stress
εo = pre-strain
ε = effective strain
K = strength coefficient of the material n = strain hardening exponent
Pi = internal pressure
σ θ = hoop stress
σ Z = longitudinal stress
rθ = incremental radius of tube in circumferential direction
rZ = radius of arc in axial direction
h = bulge height
S = Engineering Stress
e = Engineering Strain
%t = Percent thinning
xxv CHAPTER 1
INTRODUCTION/ MOTIVATION
1.1. Characterization of material properties for tubular materials
Tube hydroforming (THF) process is used widely in various applications, i.e. to
manufacture automotive, appliance and aerospace parts. The increase in functional
requirements makes the parts be more complex. As a result, the process window to
manufacture such parts is getting smaller. Finite Element Method (FEM) is used as a tool to design and develop the process. Accuracy of the FEM results relies heavily on the input parameters such as material property and interface friction condition. The tensile test normally used to determine the mechanical properties (i.e. flow stress and
formability) is conducted under uniaxial state of stress. Therefore, the tensile properties
cannot emulate the real state of stress (biaxial state of stress) happening in THF process.
Hydraulic bulge test in which the tube is subjected to biaxial state of stress has been
suggested by many researchers [Woo, 1978], [Fuchizawa, 1993], [Sokowloski, 2000],
[Koc, 1999] and [Aue-u-lan, 2000]. However, in order to conduct tube bulge experiments,
special measurement devices, such as curvature radius measurement tool, linear
potentiometer and ultrasonic device, are required to measure the curvature radius along
the longitudinal direction and the wall thickness at the apex of the dome during forming.
1 [Koc, 1999] and [Aue-u-lan, 2000] have attempted to develop a mathematical model based on deformation theory (assuming a proportional strain path) to determine the wall thickness at the apex of the dome by using only the bulge height. However, this methodology is accurate only up to the small radius of the deformed tube because the assumption of the proportional strain path is still valid. When the radius of the deformed tube is bigger, the assumption of the proportional strain path is not valid. Therefore, in this study, the incremental theory (non-proportional strain path) is developed to determine the wall thickness at the apex of the dome. This development makes the hydraulic bulge test be easy to use in the industrial environment to determine the tube properties.
Currently, the quality of the incoming tubes plays a significant role to control the
quality of the product and the percent scrap rate in production. The conventional tests based on ASTM standards, such as conical, flattening, expansion, and even tensile test, cannot evaluate the quality of the tubes used in THF process. The methods to evaluate the quality of the material need to be able to emulate the real conditions happening in the process. The hydraulic bulge test is proposed in this study to evaluate the quality of the tubes used in THF process because the tube is subjected to that internal pressure that happens in the process.
1.2. Forming of lightweight alloys at elevated temperatures
Lightweight alloys, such as aluminum and magnesium alloys, have gained more interest in the automotive industry due to the high strength per weight ratio when compared with the regular steel normally used. Furthermore, they have much better 2 mechanical properties, i.e. dent resistance and shell stiffness, than steel as shells with the same area weight and have a higher wall thickness due to their lower density. [Kleiner,
1999]
The major problem in forming of the lightweight alloys is their low formability at room temperature due to the microstructure (hexagonal closed pack in magnesium alloys) and alloy elements (in aluminum alloys) that limit the number of slip planes at the room temperature [Droder, 1999] and [Shehta, 1978]. Many studies have revealed that the formability of these alloys increase significantly at elevated temperatures (200 to 300oC).
Warm forming technology, especially for sheet forming (stamping) and bulk forming processes, has recently been investigated intensively by [Takuda, 1998&1999&2003];
[Lee, 2002]; [Yukutake, 2002&2003], [Doege, 2001]; [Iwanaga, 2004]; [Behrens, 2004];
[Jager, 2004]; [El-magd, 2003]; [Ogawa, 2002]; [Siegert, 2003]; [Matsumoto, 2002];
[Yoshihar, 2003]; [Yashihara, 2004]; [El-Morsy, 2002]; [Sillekens, 2003], [Groche,
2002]; [Neugebauer, 2003]. However, the warm forming technology for forming tubular magnesium and aluminum alloy tubes has not been well established.
Figure 1.2 summarizes various components of the warm tube hydroforming process that needs to be considered. Factors affecting the development of warm tube hydroforming process can be summarized in 4 areas as follows:
1. Tubular materials and their properties: Quality of incoming material, flow
stress and formability, process conditions and quality of products (surface
finish and mechanical properties of formed parts)
3 2. Interface conditions: Friction affects the movement of the tube at the contact
area with the die which leads to the quality of surface finish and thickness
distributions of the formed product. Interface heat transfer affects the friction
conditions and performance of lubricants.
3. Equipments and tooling: Temperature gradients at the tube and die surfaces
cause local deformation which leads to the local necking of the tube during
forming. Therefore, the heating system needs to be designed carefully in order
to avoid the temperature gradients.
4. Process conditions: Process conditions, such as forming temperature, rates and
loading paths (internal pressure versus axial feed), affect the forming behavior
and formability of the tube. Therefore, the process conditions need to be
controlled to successfully form the part.
The main goal for this research is to develop a prototype warm tube hydroforming system that could form lightweight tubes at elevated temperatures. Thus, it is possible to investigate the effect of the tube manufacturing processes on the quality of incoming tubes and the process conditions on the formability and forming behavior of the tube.
The use of Finite Element Method (FEM) will also provide a better understanding of the mechanics of forming lightweight alloys.
4
Figure 1.1: Comparison of mechanical properties [Novotny, 2001]
Figure 1.2: Warm tube hydroforming as a system –a summary of parameters contributing to the development of warm tube hydroforming process
5 CHAPTER 2
RESEARCH OBJECTIVES
Finite element method (FEM) has been used as a tool to design tooling (i.e.
forming die) and determine process parameters (i.e. loading path (axial feed vs. internal
pressure) in tube hydroforming (THF) process of steels at room temperature. The
accuracy of the FEM results depends upon reliable input parameters such as mechanical
properties (flow stress and formability), interface friction parameters and boundary
conditions. One of the significant parameters is flow stress (relationship between
effective true strain and strain) of the tubular materials. Normally, the flow stress of the
tube is obtained from the uniaxial tensile test, where the state of stress does not emulate
to the real state of stress (biaxial state of stress) occurring in THF process. Hydraulic
bulge test was introduced to determine the flow stress of the tube in biaxial state of stress.
However, the methods used to determine the flow stress from the hydraulic bulge test are based on simplified assumptions (i.e. deformation theory –assumed a proportional loading path). In other words, the relationship between major (hoop direction) and minor
(longitudinal direction) strains is assumed to be linear. Experimental as well as FEM results have revealed that the deformation path is not linear due to the state of stress changes due to the geometrical change. In this study, a methodology based on the
6 incremental strain theory is proposed to improve the accuracy of flow stress
determination.
The effect of tube manufacturing processes (i.e. roll forming process) on the
quality of incoming tubes is one of the major factors that affect the scrap rates in
production. The ASTM standards (i.e. tensile, conical, expansion, micro-hardness and
flattening tests) cannot be used as tools to evaluate the quality of incoming tubes for tube
hydroforming processes. The hydraulic bulge test should be used to evaluate the quality
of the tube received from different suppliers.
Currently research in the warm tube hydroforming of the lightweight alloys is at
the early stage. In order to successfully develop this process, this study needs to focus on
1) the design and development of economical/ robust warm tube hydroforming system
with emphasis on a) incoming material shape and properties, b) the forming temperatures,
c) the tribological condition (friction and heat transfer) at the interface, d) the tool
temperature, e) tooling/ equipment design, and f) forming speed (or strain rate); 2) the
influence of the forming equipment on the final product shape and properties; and 3) the
economics of the process.
Thus, the main objectives of the research are to develop a) an analytical model to
determine material properties of the steel tubes at room temperature and b) a robust warm
tube hydroforming process for forming of the lightweight alloy tubes. The specific
research objectives are as follows:
7 A. Characterization of an incoming tubular material
• Development of method to determine flow stress under biaxial state of
stress
• Investigation of the effects of tube manufacturing processes
upon the quality of welded steel tubes
B. Design of the experimental warm tube hydroforming system
C. Investigation of the effect of process parameters (i.e. forming temperature, forming rate and tube properties) on forming behavior of lightweight alloy (aluminum
and magnesium alloys) tubes
D. Development of Finite Element Model to simulate and analyze the warm tube
hydroforming process
8
CHAPTER 3
STATE OF THE ART
3.1. Tube hydroforming (THF) process
The Tube Hydroforming (THF) process is a relatively new manufacturing
technology, which has been used in the past decade. THF offers potential alternatives in
the use of lightweight materials and hence can have a great impact in energy saving in
automotive industry. Furthermore, THF also offers potential in design of structures with
high stiffness [Morphy, 1998]. THF offers several advantages as compared to
conventional manufacturing via stamping and welding. These advantages include: (a)
part consolidation, for example stamped and welded sections to form a box section, can
be formed as one single piece from a tubular material using hydroforming, (b) weight
reduction through more efficient section design and tailoring of the wall thickness in
structural components, (c) improved structural strength and stiffness via optimized section geometry, (d) lower tooling costs due to fewer parts, (e) fewer secondary operations (less welding and punching of holes during hydroforming), and (f)tighter tolerances and reduced springback that facilitates assembly, and (g) reduced scrap since
9 trimming of excess material is far less in tube hydroforming than in stamping [Dohman,
1996].
Figure 3.1: Tube Hydroforming process sequence; a) place a tube, b) seal and fill fluid, c) pressurize and feed material, d) take a tube out [Leitloff, 1997]
A typical THF process sequence is shown in Figure 3.1. A tube is placed between two dies. The dies are closed and held under pressure while the tube is internally pressurized and axially compressed to force the material into the deformation zone. Thus, the tube material is forced to expand to acquire the shape of the die cavities. During this process, axial feed and increase in internal pressure are controlled simultaneously to improve the material shaping capabilities [Leitloff 1997]. The main objective is to achieve the configuration represented by the die without any defects (wrinkles or fractures) in the product. Figure 3.2 illustrates some of the automotive components produced by THF process.
10
Figure 3.2: Examples of the automotive parts manufactured by tube hydroforming process
3.1.1. Tube hydroforming process as a system
In order to successfully design and develop a THF process or operation, improvements in each area of the THF technology and their interactions should be considered. The main components and key issues of a complete THF system (see Figure
3.3) can be listed as follows: a) Quality of incoming tubes, b) Performing and bending design and production methods, c) Die and tool design guidelines, d) Die-workpiece interface issue (friction and lubrication), e) Deformation mechanics (metal flow) in different zone, f) Equipment, press and environment related issues, and g) Dimensions and properties of the hydroformed part. [Jirathearanat, 2004]
11
Figure 3.3: System in Tube Hydroforming process
3.2. Tube manufacturing processes
Figure 3.4 shows the processes used to manufacture tubes. Most of the steel tubes are manufactured by using roll forming and welding. The lightweight alloy tubes are normally manufactured by using extrusion processes. There are 2 types of the extrusion processes; extrusion with a porthole die and extrusion with a mandrel. Advantages and disadvantages for both processes are described in Section 3.23. Alternatively, some types of lightweight alloys such as AA5XXX series alloys can be manufactured by using continuous roll forming process. However, the quality of welding process is the major effect on the quality of the tube and the cost for manufacturing the tube is relatively high.
12 Tubular materials
Steel Lightweight alloys
Slab sheet Ingot
Continuous Rolling process Extrusion process (Thin gage sheet)
Roll forming process Extrusion with Extrusion with a porthole die a mandrel
Continuous roll forming Press break bending
Figure 3.4: Overview of processes to manufacturing tubes in THF process
3.2.1. Tube manufacturing by roll forming
The sheets used for manufacturing the tubes are made by conventional strip rolling operations. Roll forming process is one of the most common processes used to produce thin walled welded steel tubes. The roll forming process is normally divided into
2 types: a) continuous roll forming process and b) press brake bending process. The continuous roll forming process is used often to produce the tube because of higher production rate. [Rempe, 2000]
Figure 3.5 illustrates the overall process to manufacture the tubes by the continuous roll forming process. The sheet from the coil is transferred to the roll forming
13 passes (shape roll forming) to gradually form the sheet to a tube. The roll formed sheet is then welded and sized to produce accurate tube dimensions.
Figure 3.5: Continuous roll forming process to manufacture a steel tube [Singh, 2002]
Figure 3.6: W- flower sequence for manufacturing the tubes [Gehrish, 1993]
14 3.2.2. Extrusion process
Extrusion process is a common process to produce aluminum and magnesium alloy tubes. The extrusion process starts by heating up the billet to temperature depending on the types of materials. Then, the heated billet will be forced with a ram through the die with the required cross section shapes, as seen in Figure 3.7.
The design of the extrusion dies for manufacturing a hollow cross section (i.e.
circular cross section) is divided mainly into 2 designs as follows:
3.2.3-A: Mandrel extrusion die: This extrusion die design, as seen in Figure 3.7,
is suitable for single cavity sections. The billet needs to be pierced through the center
prior to the extrusion step. The main advantage of this extrusion die is that there are no
welding lines in the extruded cross section. Therefore the property of the tube is much
more homogenous. The main drawback of this method is the large thickness variation
because the mandrel can deflect or wander from side to side during the extrusion process.
[Singh, 2002] The variation in wall thickness of the tube also could cause of the variation
of the formability of the tube. [Shirayori, 2003] studies show that a) the material trends to
expand first at the area that has the lowest thickness, and b) the deviation of the wall
thickness is increased when the tube keeps expanded. As a result, non-uniform
deformation of the tube around the circumferential direction may occur. [Shirayori, 2003]
3.2.3-B Porthole extrusion die: In this extrusion die, the hollow cross sections, as seen in Figure 3.8, are extruded by welding-chamber methods and the use of various dies known as spider dies, porthole dies, and bridge dies. During the extrusion, the metal divides and flows around the supports for the internal mandrel into strands; these strands
15 are then re-welded under the high pressures existing in the welding chamber, before they exist through the die.
The quality of the extruded tubes by using the porthole die is mainly depended on the quality of welding lines, when the tubes are subjected to severe internal pressure or expansion in the practical use. [Kim, 2002]
Figure 3.7: Schematic illustrates the mandrel extrusion die [Singh, 2002]
16 (a)
Porthole die Spider die Bridge die
Figure 3.8: (a) an extruded AA6061-T6 cross section for tubes, and (b) –(d) components of various dies for extruding intricate hollow shapes [Kalpakjian, 2001]
3.3. Mechanical tests
The quality of the incoming tubular materials can be tested in different ways depending on ASTM standards. The purposes of mechanical tests are to a) evaluate the quality of welding line and tubes (i.e expansion, conical and flattening tests) and b) determine flow stress and formability of the materials (i.e. tensile and biaxial tests) used in Finite Element Modeling.
3.3.1. Tensile test
The material, i.e. tubular materials, during hydroforming is always subjected to the biaxial state of stresses. Therefore in order to determine the quality and mechanical properties of the tubes, the mechanical test needs to be able to emulate the real state of stresses that occurs in tube hydroforming process [Aue-u-lan, 1999]. ASTM A513 17 standard, which most of the tube suppliers follow, provides basic quality control measures for hydroforming tubes. This standard also provides some guidelines for mechanical testing of tubes by means of uniaxial tensile test, which is more suitable for structural tube allocations than metal forming applications. The tensile specimen is cut and tested only along the longitudinal direction of the tube, as seen in Figure 3.10.
Furthermore, [Wang, 2001] has developed the ring hoop tension test (see Figure 3.9) technique to determine the mechanical properties of the tube along the circumferential
direction. However, none of these tests is really reliable to determine the quality and
mechanical properties of the tubes, because they do not really reflect the real state of
stresses that exist in hydroforming.
Figure 3.9: Ring hoop tension test specimen and fixture [Wang, 2001]
18 A)
B)
C)
Figure 3.10: Specimens used for tensile testing of sheet metals. A) Sheet metal specimen, B) Tubular Specimen, C) Longitudinal sectioned specimen for tubular material testing [after ASTM E8-96 (modified)]
3.3.2. Summary of the standard test for tubular materials
There are very few standards for testing tubular materials that are applicable to tube hydroforming. Most of the material specifications are for sheet material, from which, the tubing is rolled or for tubular materials in general-purpose applications.
There are some standards, which may be applicable, including specifications for weld seam quality, eccentricity, etc. These specifications are for materials for use as structural or mechanical tubing. Table 3.1 lists some of these standards.
19
Standard Title Source B313/313M – A92 Specifications for Aluminum and Aluminum- Annual Book of Alloy Welded Tubes ASTM Standards A500 - 96 Cold-Formed Welded and Seamless Carbon Steel Annual Book of Structural Tubing in Rounds and Shapes ASTM Standards A501-96 Hot-Formed Welded and Seamless Carbon Steel Annual Book of Structural Tubing ASTM Standards A511 – 96 Seamless Stainless Steel Mechanical Tubing Annual Book of ASTM Standards A512 – 96 Cold-Drawn Butt-weld Carbon Steel Mechanical Annual Book of Tubing ASTM Standards A513 – 96 Electric - resistance – Welded Carbon and Alloy Annual Book of Steel Mechanical Tubing ASTM Standards A519 – 96 Seamless Carbon and Alloy Steel Mechanical Annual Book of Tubing ASTM Standards B547/B547M – 93 Standard Specifications for Aluminum and Annual Book of Aluminum-Alloy Formed and Arc-Welded Tube ASTM Standards A554 – 94 Welded Stainless Steel Mechanical Tubing Annual Book of ASTM Standards A618 – 96 Hot-Formed Welded and Seamless High- Annual Book of Strength Low-Alloy Structural Tubing ASTM Standards A778 – 90 (1995)a Welded, Unannealed Austenitic Stainless Steel Annual Book of Tubular Products ASTM Standards A787 – 96 Electric-Resistance-Welded Metallic-Coated Annual Book of Carbon Steel Mechanical Tubing ASTM Standards A953 – 96 Austenitic Chromium-Nickel-Silicon Alloy Steel Annual Book of Seamless and Welded Tubing ASTM Standards SAE J 356-91 Welded Flash Controlled Low Carbon Steel Tubing Normalized for Bending, Double Flaring, and Beading, Standard SAE AMS 5077E- Steel Tubing, Welded (SAE 1025) Normalized or 89 Stress Relieved
Table 3.1: List of United States Standards and Specifications applicable to tube hydroforming
20 3.4. Variation of material properties
The major cause of the formability variations in tubular materials produced by the roll forming process could be divided into 2 categories as follows:
A. Variation in material properties of the tubes from the same rolling
strip but different slit portions along width direction
As these sheets have large widths, many sets of tubes are made from the same sheet (For example, in the sketch shown in Figure 3.13, 3 sets of tubes were made from the sheet at different location along the width direction). Sheets were cut along the width
direction to get to the required dimension for making the tube with the required diameter.
This sheet is passed through many rolls to bend the flat sheet into a circular sheet. This
sheet is then welded to produce a tube (roll forming process).
In hot-rolled process the heat transfer rates of the sheet material at the outer
region (tube I and III) are different from that of the sheet material at the center (tube II) as
seen in Figure 3.13. This leads to differential cooling of the sheet material at different
regions. Hence the tubes, though made from the same sheet (wide sheet), may have
different material properties due to the different cooling rate at different location of the strip sheet. [Schuler, 2005] has studied this phenomenon extensively. The location of the slit from the big rolled sheet was monitored and recorded before the slit was sent to the roll forming process. Once the tubes were manufactured, the hydraulic bulge test was used to test the tubes. Figure 3.14 shows the variation in percentage elongation from the tubes produced from the slits at different locations on the rolled sheet. According to the
results, the tubes produced from the slits located at the middle of the big sheet have a
21 higher formability than those produced from the slits located at the middle of the big sheet. The difference is significant may cause failure (bursting) during hydroforming process.
B. Variation in material properties of the tubes at different locations
around the circumferential direction
Many researchers have investigated the variation in tube properties caused due to tube-making process. Boyles and Davies [Boyles, 1999] investigated the change in material properties from sheet steel coil through to finished hydroform for a range of body-in-white components. They studied the magnitude of changes in the work hardening of input material and whether it was significant for component design. They measured the variation in yield strength around the circumference of ERW tube. They concluded that the combined thermal and mechanical history imparted during conversion from coil to hydroform, coupled with the metallurgical response of different steel grades, can have significant influence on material properties which in turn impact performance and durability of hydroformed components. Hielscher [Hielscher, 2001] measured the hardness distribution in a roll-formed tube and conducted tube bursting tests. In this study he found that the hardness changes over the circumference of the tube, as seen in Figure
3.11. Figure 3.12 shows that the strain distributions after bursting test have similar non- uniform distributions.
Carleer [Carleer, 2001] studied the influence of roll-forming process on properties of tubes and concluded that the tube hydroforming process chain should be simulated
22 starting from tube making process. In his study, FE simulation on roll forming process was conducted and the variation of properties around the tube circumference was studied.
Figure 3.11 Hardness distribution of roll formed tube [Hielscher, 2001]
23
Figure 3.12 Engineering strain (%) distribution after tube bursting tests [Hielscher, 2001]
Middle
Edge Slit (1) End Central Slit Front
Edge Slits (2)
Figure 3.13: Schematic sketch for the regions from which the tubes are made from the same sheet strip
24
Figure 3.14: Influence of the locations (see Figure 3.13) of the sheet used to produce the tubes on the percent elongation of the tube (Test results were conducted by hydraulic bulge test) [Schuler, 2005]
Figure 3.15: Tube cross-section showing different testing locations
25 3.5. Forming technology at elevated temperatures of lightweight alloys
Some hard-to-form materials, such as Mg and Al alloys, were formed by a superplastic forming process (temperature >0.5Tm, where Tm =melting temperature). The
drawbacks of this process are; a) low forming rate (strain rate is in the range of 10-4 to 10-
2), b) low stiffness of the part after forming c) high energy to heat the part up to superplastic temperature ranges and d) precise and sophisticated temperature controllers to maintain the uniform temperature. In order to avoid the drawbacks, most researchers are interested in forming the materials at a lower temperature range (0.2 to 0.5Tm). This
process is called “Warm Forming Process”. Warm Forming is relatively new. Several
universities /institutes around the world are in the initial stage of design and development of warm forming systems. In this review the emphasis will be on a warm hydroforming system design. The warm hydroforming process employs a heated medium to form a sheet or tube instead of using the solid punch commonly used in a conventional stamping process.
26
Figure 3.16: Temperature range to identify the forming conditions [Novotny, 2002]
The biggest challenges of the development of this process are designing heating techniques to heat and maintain the sheet/ tube and dies at the designed temperature, and choosing the type of pressurizing medium used to form the material. [Neugebauer, 2003] has employed heated fluid (high temperature oil) to heat the die and sheet to the designed temperature (200 –300oC) for his warm sheet hydroforming process of an automotive door panel (see Figure 3.19B). The main advantages of this design are that only one source of heat energy is required to heat the fluid medium and that maintain a uniform temperature distribution at the sheet is possible due to the direct contact between the sheet and heated medium. As seen in Figure 3.19A, the heating source of the system comes from an electrical heater (60KW). Fluid, heated up by the electrical heater, heats the upper and lower dies as well as the pressuring liquid medium inside the heat exchanger. The drawback of this design is the high smoke (oxidation of the fluid) formation after forming due to the direct exposure of the fluid to the air. However, this 27 problem could be solved if a high oxidation resistant fluid is selected and a ventilation system is installed to eliminate the smoke.
Dorr, 2004 (see Figure 3.17) and Jager, 2003 (see Figure 3.18) have used electrical heaters (cartridge heaters) to heat the dies for their warm tube hydroforming designs and they have a separate system to heat the pressurizing medium (using heat pump). In the Dorr’s design concept, he heats the tube by using the heated die and fluid.
He employs a heat transfer fluid as the pressuring medium. One of his core design concepts was to heat the die at the deformation zone area and cool it at the guiding zone area (see Figure 3.17). With this design the axial feed at the guiding zone can be enhanced due to the low friction (friction is direct proportional to the temperature) as well as a low risk of wrinkle at the guiding zone. The drawback of this design is high temperature gradients at the deformation zone, especially in the transition zone, between the deformation and guiding zones. These gradients may cause non-uniform deformation.
Also even though an air gap was used as an insulator between the guiding and deformation zones, there is a significant energy loss at the deformation zone.
While [Dorr, 2004] used fluid as the pressuring medium, [Jager, 2003] used hot air. He used the heated die to heat the tube to the designed temperature and employed hot air to form the tube. The drawback of this design is the temperature gradients that exist in the tube because the tube has a direct contact to the air during forming. Also the amount of energy used to heat the air is wasted to the environment every time after forming.
28
Figure 3.17: WTHF design concept of University of Darmstadt [Dorr, 2004]
Retainer Insulation Tie rod Sealing punch hydraulic-cylinder Insert for Cooling plate expansion Band heater Cross head
le Hydraulic-- Insert cylinder „guided zone“
Pressure inlet Insulation Band heater
Bottom plate Cooling plate Main body
Figure 3.18: Schematic of hot gas metal forming equipment [Jager, 2003]
29
Figure 3.19: A) a simplified illustration of the heating plant and pressure system, and B) a cross section of the tempered hydroforming-tool (the front view and top view) indicating the heating channels [Neugebauer, 2003]
30
CHAPTER 4
DEVELOPMENT OF AN ANALYTICAL MODEL TO DETERMINE FLOW STRESS OF TUBES
The accuracy of FEM simulations of tube hydroforming (THF) is strongly
dependent on the parameters of the flow stress law used to describe the plasticity of tubular materials used. In the current industrial practice of tube hydroforming (THF) operations, very often the mechanical properties and the formability of tubes are derived from the tensile test data of the flat sheets used to manufacture the tubes. Alternatively, the material data are determined by running a tensile test directly on the tubes, rather than on the sheets. [Wang, 2001]
In both cases, these practices present some drawbacks, as also stated in previous
works (see as an example [Fuchizawa, 1993]). One disadvantage is that the maximum
effective strain value achievable with an ordinary tensile test before localized necking
occurs is remarkably lower than the effective strain values usually reached during the
hydroforming process. Furthermore, when using material data obtained by tensile tests of
sheets, they should at least be corrected to consider the straining due to the bending
process used to form the tubes (roll forming process).
31 Moreover, though the results of the tensile test can provide information about the stress-strain relationship, they can hardly be used to evaluate formability of tubes for hydroforming, since the tensile test induces a uniaxial state of stress, while the THF process is mainly biaxial. In other words, a test generating a biaxial tensile stress state in the sample (such as a hydraulic bulge test) would be closer to the real process conditions and this would insure a much more effective evaluation of formability [Jevons, 1942].
Figure 4.1 illustrates the state of stress occurring in the hydraulic bulge test.
σ θ = σ Z σ Z Stress state of Hydraulic bulge test
1 σ = σ θ 2 Z Stage 2
Uniaxial tensile test σ θ Stage 1
Figure 4.1: State of stress in Hydraulic Bulge Test
For the reasons stated above, several alternative testing procedures and tooling have been proposed so far, like the sheet bulge test (extensively described in the literature), the tube bulge test [Fuchizawa 1993] or more complex combined tests [Hora, 32 2000]. The hydraulic bulge test for tubes is gaining always more and more attention from the hydroforming industry. Bulge test equipment has been developed by several research institutes, hydroforming press manufacturers and tube suppliers. [Aue-u-lan, 1999]
The main problem of using the tube bulge test for determining the stress strain relationship is the measurement of the tube radius of curvature in the axial direction and the wall thickness at the apex of the dome (rz and t in Figure 4.5). This curvature is
required to calculate the stresses, based on stress balance equations. Therefore, systems
for tube bulge testing must be equipped with devices able to measure the bulge height,
the tube curvature and the wall thickness. [Fuchizawa, 1993] This requirement may result
in an increase of cost and make it difficult to use the bulge test in practice for quality
control. Furthermore, in some cases it can difficult to obtain a good precision in the
measurement of curvature [Rees, 1995] and wall thickness [Fuchisawa, 1993], thus
causing a loss of accuracy in the flow stress curve. This author [Aue-u-lan, 1999]
attempted to develop the mathematical models to approximate the curvature radius based
on the geometrical relationship between the bulge height and curvature radius (see in
Figure 4.6). The results after calculating the flow stress were acceptable. In contrast, this
author has developed the analytical model based on the assumption of the proportional
strain path to predict the wall thickness at the apex of the tube. The results were
reasonable only at the low hoop strain (low bulge height). When the deformation is
higher, the deformation path starts to deviate from the linear relationship. This causes of
the error in thickness predictions.
33 In this study, the new mathematical model based on the incremental strain theory was developed by this author to predict the wall thickness at the apex in order to improve
the accuracy in flow stress calculations.
4.1. Description of hydraulic tube bulge test
In this present study, hydraulic bulging unit was developed to biaxially deform
tubular samples since tube hydroforming also applies a biaxial loading to the tubes. This
tool set is suitable for use with different types of materials, different tube sizes, and
different pressure levels.
The basic system consists of a press, hard tooling set (Figure 4.2), hydraulic
pressurization system, and data acquisition equipment (for online measurement of
internal pressure and bulge height).
Figure 4.3 shows the schematic of hydraulic bulge test used to determine the
properties of tubes (effective stress versus strain and formability). Tube is placed between
the upper and lower dies. Dies are closed using hydraulic press. The tube is then
pressurized till it bursts while the ends of the tube are prevented from moving axially.
Internal tube pressure and the bulge height are recorded continuously during the
experiment, using pressure transducer (Sensotec) and linear potentiometer respectively.
From this data, the effective stress (σ ) and effective strain (ε ) can be calculated and
then plot as the flow stress curve. Details on how to calculate the flow stress by using
analytical model are described in the next section of this chapter.
34
Figure 4.2 Hydraulic bulge test tooling in the press
35
Figure 4.3 Schematic of hydraulic bulge tooling
Bulge width Final w thickness Bulge height t1 Initial h thickness
r 1 Final tube t0 r 0 Initia l tube radius radii
Figure 4.4: Geometry of the deformed tube and the nomenclature used in calculations.
36
Figure 4.5: Geometry of the bulge and stress components acting at the apex of the dome. rz, radius of curvature in the axial direction; rθ, radius of curvature in the
axial direction, t, wall thickness, σ θ = hoop stress and σ Z = longitudinal stress
4.2. Constitutive models
The flow stress curve can be represented by points or by equations. The most common equations used to represent the flow stress of tubular materials are:
n • Krupkowsky’s law (σ = K(ε 0 + ε) , where K is strength coefficient, n
is strain hardening coefficient, and ε0 is pre-strain). This equation
considers the effect of pre-strain due to tube manufacturing by roll
forming and welding.
37 n • Hollomon’s law (σ = Kε , where K is strength coefficient and n is
strain-hardening coefficient). This equation does not consider the effect
of pre-strain. Normally, it will be used from the tube manufactured by
extrusion process.
For this study, the flow stress equation used is Krupkowsky’s law since the amount of formability is consumed during the manufacturing process (i.e. roll forming process).
Figure 4.6: Schematic of geometrical relationships to determine the curvature radius
38 4.3. Development of analytical model
4.3.1. Membrane theory
The membrane theory can be used to calculate the flow stress of thin wall tube
(OD/t0 >>20). The major assumption of the membrane theory is that the bending and
thickness stresses are neglected. Equation 4.1 shows the relationship among stresses, tube
geometries and internal pressure. (see Figure 4.5)
P σ θ σ z = + Equation 4.1 t i rθ rz
where σ θ and σ Z are stresses along hoop and longitudinal directions , respectively.
P is the internal pressure applied at the tube wall and ti is wall thickness at the apex of the
tube. rθ and rZ are hoop and curvature radius, respectively.
The stress in the longitudinal direction can be calculated in terms of internal pressure, the bulge radius and thickness as:
Prθ σ Z = Equation 4.2 2ti
With Equations 4.1 and 4.2, the hoop stress can be calculated as:
Prθ σ z rθ σ θ = − Equation 4.3 ti rz
With Equations 4.2 and 4.3, the effective stress at the apex of the dome can be calculated based on Von Mises Yield criterion as follows:
2 2 σ = σ θ − σ θ σ Z + σ Z Equation 4.4
39 The effective strain can be calculated based on Von Mises Yield criterion as follows:
n ⎛ 2 ⎞ ⎜ 2 2 2 ⎟ ε = ∑ ⎜ ()dε + dε + dε ⎟ Equation 4.5 i=1 ⎝ 3 ⎠
where dε θ , dε Z and dε t = increment hoop, longitudinal and thickness strain, respectively.
4.3.2. Calculation of the curvature radius at the apex of the dome
The curvature radius at the apex of the dome as seen in Figure 4.4 can be
calculated by assuming that the cross-section of the bulged tube can be approximated
to the two circular arcs. Then, from geometry, the followings are obtained.
rZ sinφc + (Re + to / 2) sinφc = w
rz (1 − cosφc ) + (Re + to / 2)(1 − cosφc ) = h + to / 2 − t / 2
2 2 2 (rZ + Re + t o / 2) = w + (rZ + Re + t / 2 − h)
2 2 w + (h + to / 2 − t / 2) rZ = − (Re + to / 2) Equation 4.6 2(h + to / 2 − t / 2)
4.3.3. Relationship between incremental strains along hoop and longitudinal
directions
dε Define A = Z , during the bulging process; let the tube element with width s dε θ
expands from the current location rθ to the location of (rθ + drθ ) with width (s + ds) .
40 Let’s assume that the expansion of the tube element is due to the work done by the internal pressure p .
Work increment done by the internal pressure to the tube element:
dW = pdV ,
2 2 ⎡ 2 ⎛ ds ⎞⎤ where dV = π (rθ + drθ ) (s + ds) − πrθ s − 2⎢πrθ ⎜ ⎟⎥ = 2πrθ sdrθ is a volume of ⎣ ⎝ 2 ⎠⎦
the pressurized internal media.
The volume of the tube element:
vs = 2πrθ ts
Therefore, the work done by the pressure to the unit volume of the tube element
is:
dV 2πrθ sdrθ p p dw = p = p = drθ = rθ dε θ , vs 2πrθ ts t t
drθ where dε θ = is a strain increment in the circumferential direction. rς
The work required to deform the tube element in a stress status of (σ Z ,σ θ ) is:
dw = σ ij dε ij = σ Z dε Z + σ θ dε θ = (Aσ Z + σ θ )dε θ ,
where A = dε Z / dε θ = ε&Z / ε&θ is a ratio between the strain-increments or the
strain-rates. By comparing the above equations, we obtain the following relationship.
p dw = (Aσ + σ )dε = r dε Z θ θ t θ θ
Aσ + σ p Z θ = rθ t
41 If we remind the equilibrium equation of the tube element with two radii of
curvature, (rZ ,rθ ) ,
σ σ p Z + θ = rZ rθ t
the following relationship between A = dε Z / dε θ and the radii of curvature can
be obtained.
1 A dε r = or A = Z = θ Equation 4.7 rZ rθ dε θ rZ
The above equation is true when rZ = ∞ and rZ = R .
4.3.4. Calculation of wall thickness at the apex of the tube
The mathematical model used to calculate the wall thickness at the apex of the dome was developed by [Aue-u-lan 1999] based on the proportional strain path
dε ε ( Z ≅ Z ). However, based on the FEM results as seen in Figure 4.7, the strain path dε θ ε θ will be linear only at the small bulge height. The strain path will be deviated from the linear at the higher bulge height. Therefore, the methodology to calculate the wall thickness at the apex can be improved by using the incremental strain approximation.
Recall the relationship between the incremental strains along the hoop and
dε r longitudinal direction ( A = Z = θ ) and the volume consistency dε θ rZ
42 (0dε θ + dε Z + dε t = ). Therefore, the thickness at the bulge height can be calculated as
follows.
dε t = −(1 + A)dε θ
Assuming the incremental strain is the infinitesimal strain. Thus,
∆ε t ≅ −(1 + A)∆ε θ
dt drθ Since ∆ε t = and ∆ε t = t rθ
Therefore,
dt dr = −(1 + A) θ t rθ
or
(1+ A) ⎛ r ( j−1) ⎞ t ( j) = t ( j−1) exp(∆ε ) = t ( j−1) exp[−(1 + A)∆ε ] = t ( j−1) ⎜ θ ⎟ Equation 4.8 θ θ ⎜ ( j) ⎟ ⎝ rθ ⎠
( j) ( j−1) ( j) ( j−1) where ∆ε t = ln(t / t ) and ∆ε θ = ln(rθ / rθ ) are used.
The wall thickness of the tube could not be calculated directly from the equation.
As seen in Equation 4.8 the “A” term used to calculate the wall thickness is a function of
the thickness as well. Therefore, the wall thickness needs to be calculated numerically.
Figure 4.9 illustrated the procedure to calculate the wall thickness of the tube. First, the
wall thickness is guessed (in this case the initial wall thickness (t0) could be used as an
initial guess). The “ rZ ” and “A” could be calculated by using Equations 4.6 and 4.7,
respectively. Then, the wall thickness could be calculated by using Equation 4.8. The
calculated and guessed wall thickness is compared. If both values were not the same, then
43 the next guessed wall thickness would be selected until the values are the same or at least closed to 1xe-4. Then, program will be stopped.
0.08 0.07 n i 0.06 a r
st 0.05 l
na 0.04 udi t 0.03 0.02 Longi 0.01 0 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 Hoop strain
Figure 4.7: Relationship of strain along the hoop and longitudinal directions
44
Figure 4.8: Schematic of the infinitesimal element at the apex of the dome
45
Figure 4.9: Flow chart of thickness calculation
Figures 4.10 to 15 show some of the examples of wall thickness predictions for various steel materials (Low carbon steel and Stainless steel). The observation from this
calculation can be summarized as follows:
A. Both techniques developed by using deformation and incremental
theories could predict wall thickness of the tubes very well within the
reasonable error (less than 10%).
46 B. The accuracy in the thickness predictions developed by the
deformation theory [Aue-u-lan, 1999] was low when the bulge height
(h) was higher. This is due to the fact that as seen in Figure 4.7 the
strain path (ratio between hoop and longitudinal strains) is no longer
proportional. As a result, by assuming the ratio of both strains was
proportional could result more error in the wall thickness calculation.
The incremental theory could improve the thickness predictions at the
higher bulge height.
47 Low carbon steel grade AISI 1008 (Galvanized steel)
Dimension of the initial tube and die geometry (Figure 4.6) OD L0 t0 2w 2.25 in (57.15 mm) 8 in (203.2 mm) 0.079 in (2.00 mm) 1 in (25.4 mm)
Figure 4.10: Comparisons of the wall thickness at the apex of the tube among experimental, calculated by incremental theory and deformation theory [Aue-u- lan, 1999]
48 Material: Low Carbon Steel grade AISI 1008
Dimension of the initial tube and die geometry (Figure 4.6) OD L0 t0 w 2.25 in (57.15 mm) 8 in (203.2 mm) 0.079 in (2.00 mm) 1 in (25.4 mm)
Figure 4.11: Comparisons of the wall thickness at the apex of the tube among experimental, calculated by incremental theory and deformation theory [Aue-u- lan, 1999]
49 Material: Low Carbon Steel grade 1008
Dimension of the initial tube and die geometry (Figure 4.6) OD L0 t0 w 3.50 in (88.90 mm) 9.0 in (228.6 mm) 0.079 in (2.00 mm) 1.5 in (38.1 mm)
Figure 4.12: Comparisons of the wall thickness at the apex of the tube among experimental, calculated by incremental theory and deformation theory [Aue-u- lan, 1999]
50 Stainless Steel grade AISI 304
Dimension of the initial tube and die geometry (Figure 4.6) OD L0 t0 w 2.25 in (57.15 mm) 8 in (203.2 mm) 0.024 in (0.61 mm) 1 in (25.4 mm)
Figure 4.13: Comparisons of the wall thickness at the apex of the tube among experimental, calculated by incremental theory and deformation theory [Aue-u- lan, 1999]
51 Stainless Steel grade AISI 304
Dimension of the initial tube and die geometry (Figure 4.6) OD L0 t0 w 2.25 in (57.15 mm) 8 in (203.2 mm) 0.049 in (1.25 mm) 1 in (25.4 mm)
Figure 4.14: Comparisons of the wall thickness at the apex of the tube among experimental, calculated by incremental theory and deformation theory [Aue-u- lan, 1999]
52 Stainless Steel grade AISI 304
Dimension of the initial tube and die geometry (Figure 4.6) OD L0 t0 w 2.25 in (57.15 mm) 8 in (203.2 mm) 0.065 in (1.65 mm) 1 in (25.4 mm)
Figure 4.15: Comparisons of the wall thickness at the apex of the tube among experimental, calculated by incremental theory and deformation theory [Aue-u- lan, 1999]
53 4.4. Procedure to determine flow stress by using an analytical model
The procedure to determine flow stress is listed as follows:
Step 1: During the hydraulic bulge test experiment, the internal pressure and bulge height will be recorded continuously until the tube bursts.
Step 2: The bulge height and tube geometry (i.e. tube radius and initial wall thickness) will be used to calculate the curvature radius and wall thickness at the apex of the dome.
Step 3: The bulge height, internal pressure, wall thickness and curvature radius will be used to calculate the effective stress and strain.
Step 4: The effective stress and strain will be fit to the Krupkowsky’s law
n (σ = K(ε 0 + ε ) ) by using the least square method.
54 Continuously measured data
Pressure (P) ) ) Bulge height (h) Membrane theory ) a a a P P P 600 M M M 500 ( ( ( s s s s s Effective stress s 400 e e e r r r 300 e st e st σθ ,σ Z ⇒ σ e st 200 0.25 v v v i i i σ = 600 (ε +0.1) ct ct ct 100 e e e f f f f f Analytical model f 0 E E E (Geometrical relationship) Effective strain 0 0.1 0.2 0.3 0.4 Effective strain εθ ,ε t ⇒ ε Curvature radius (rz)
Wall thickness
at the apex (ti)
Figure 4.16: Flow chart of flow stress determination by using analytical model
4.5. Flow stress determination of Stainless steel AISI 304
4.5.1. Dimensions and mechanical properties
The Stainless Steel 304 tubes, used in these tests, were roll-formed and laser welded, as commonly used for the tube hydroforming applications. This material is normally used for the exhaust components in the automotive parts.
The material properties obtained from uniaxial tensile test of flat sheet (supplied by the tube manufacturer) and the dimensions of this tube are shown in Error! Reference source not found..
55
Parameter Value Outside Diameter (D0) 57.15 mm (2.25 in) Wall Thickness (t0) 0.61 mm (0.024 in) Tube length (L0) 203.2mm (8.0 in) Bulge width (2w) 50.8mm (2 in) Strain hardening coefficient, n 0.48 Strength coefficient, K 1000MPa (210.4ksi) Yield Stress (σy) 260MPa (37.7ksi)
Pre-strain (ε 0 ) 0.06
Table 4.1: Material properties and geometry of Stainless Steel 304 specimens. The material properties are applicable to flat sheet metal, using the Holloman’s n Law (σ = Kε ) (source: Honda R&D, 1998)
Experimental results and flow stress determination with AISI 304
Figure 4.17 shows the bulge height versus internal pressure obtained from the hydraulic bulge test. The experiments were conducted 3 times. The measurement location was at the 180 degree from the welding line in order to avoid the effect of the welding on the forming behavior. Based on the incremental theory explained in Section 4.3.4, Figure
4.18 illustrates the wall thickness calculations at the apex of the dome. Then, the internal pressure, bulge height and wall thickness were used to calculate the effective stress and strain as seen in Figure 4.19.
56
Figure 4.17: Bulge height versus internal pressure of stainless steel grade AISI 304 obtained from the hydraulic bulge test [OD = 57.15mm (2.25in), initial wall thickness, t0, = 0.61 mm (0.024in), tube length, L0, = 203.2 mm (8.0in) and bulge width, 2w, = 50.8 mm (2.0 in)
57
Figure 4.18: Wall thickness at the apex of the dome versus bulge height of stainless steel grade AISI 304 obtained from the hydraulic bulge test [OD = 57.15mm (2.25in), initial wall thickness, t0, = 0.61 mm (0.024in), tube length, L0, = 203.2 mm (8.0in) and bulge width, 2w, = 50.8 mm (2.0 in)
58
Figure 4.19: Comparison of the effective stress versus effective strain of stainless steel grade AISI 304 obtained from the hydraulic bulge test (from tube) and tensile test (of sheet) [OD = 57.15mm (2.25in), initial wall thickness, t0, = 0.61 mm (0.024in), tube length, L0, = 203.2 mm (8.0in) and bulge width, 2w, = 50.8 mm (2.0 in)
4.6. Experimental results and flow stress determination of Low Carbon
Steel grade AISI 1008
Low carbon steel tubes grade 1008 used in these tests were roll-formed and solid state (High frequency) welded, as commonly used for the tube hydroforming applications.
This material is normally used making an engine cradle or side frame for a vehicle.
The material properties obtained from uniaxial tensile test of flat sheet (supplied by the tube manufacturer) and the dimensions of this tube are shown in Table 4.2. 59 Parameter Value Outside Diameter (D0) 88.9 mm (3.50 in) Wall Thickness (t0) 2.00 mm (0.079 in) Tube length (L0) 228.6mm (9.0 in) Bulge width (2w) 76.2mm (3 in) Strain hardening coefficient, n 0.23 Strength coefficient, K 548 MPa (210.4ksi) Yield Stress (σy) 190 MPa (27.6 ksi)
Table 4.2: Material properties and geometry of low carbon steel grade 1008 specimens. The material properties are applicable to flat sheet metal, using the n Holloman’s Law (σ = Kε ) (source: LTV steel)
Figure 4.20 shows the bulge height versus internal pressure obtained from the hydraulic bulge test. The experiments were conducted 3 times. The measurement location was at the 180 degree from the welding line in order to avoid the effect of the welding on the forming behavior. Based on the incremental theory explained in Section 4.3.4, Figure
4.21 illustrates the wall thickness calculations at the apex of the dome. Then, the internal pressure, bulge height and wall thickness were used to calculate the effective stress and strain as seen in Figure 4.22.
60
Figure 4.20: Bulge height versus internal pressure of low carbon steel grade AISI 1008 obtained from the hydraulic bulge test [OD = 88.9mm (3.5in), initial wall thickness, t0, = 2.00mm (0.079in), tube length, L0, = 228.6mm (9.0in) and bulge width, 2w, = 76.2mm (3.0 in)
61
Figure 4.21: Wall thickness at the apex of the dome versus bulge height of low carbon steel grade AISI 1008 obtained from the hydraulic bulge test [OD = 88.9mm (3.5in), initial wall thickness, t0, = 2.00mm (0.079in), tube length, L0, = 228.6mm (9.0in) and bulge width, 2w, = 76.2mm (3.0 in)
62
Figure 4.22: Comparison of the effective stress versus effective strain of low carbon steel grade AISI 1008 obtained from the hydraulic bulge test (tube) and tensile test (sheet) [OD = 88.9mm (3.5in), initial wall thickness, t0, = 2.00mm (0.079in), tube length, L0, = 228.6mm (9.0in) and bulge width, 2w, = 76.2mm (3.0 in)
4.7. Conclusions
A. The hydraulic bulge test was selected in this study to determine the
flow stress of the tubular materials.
B. The analytical model based on the incremental theory was developed to
predict the wall thickness at the apex of the dome.
C. The wall thickness calculated by using deformation theory was
compared with that calculated by using the incremental theory. At the
63 small bulge height (less than 12mm), both theories could predict very
well. However, when the bulge height is higher than 12mm, the
predicted wall thickness from both theories were different. The wall
thickness predicted by the incremental theory is agreeable to the
measured wall thicknesses.
D. The analytical model was developed to determine the flow stress of the
tube.
E. The flow stress obtained from the hydraulic bulge test is higher than
that obtained from the tensile test because the flow stress for the tensile
test came from the sheet prior to the tube manufacturing process. The
tube manufacturing process is consumed the formability of the sheet.
64
CHAPTER 5
INVESTIGATION OF THE EFFECT OF MANUFACTURING PROCESS UPON TUBE QUALITY
The main objective in this study is to apply the hydraulic bulge test and analytical model developed in previous chapter to investigate the effect of tube manufacturing processes. This study is emphasized only on the roll formed tubes, since the majority of the steel tubes or even some of aluminum alloy tubes, such as AA5XXX series aluminum alloys, used in THF process is produced by the roll forming process. As mentioned in
Chapter 3 on the cause of the variation of the formability and flow stress of the tubes comes from a) variation of the sheet used to manufacture the tube and b) the roll forming
and welding processes. Unfortunately, in this study the location of the slit (a small sheet
used to produce the tube, such as Tube I in Table 5.4) from the big sheet was not properly
indicated. The big sheet itself was cut along the width and lengthwise to become small
pieces. Then those small pieces were welded together to become a big coil. This coil was
used to manufacturing the tubes.
Later, the different criteria such as a) maximum bulge height at the bursting
pressure, b) strain hardening coefficient (n) and c) maximum percent thinning were tested
65 to determine which of the criteria could be used as a formability index to justify the quality of the tubes.
5.1. Experimental procedure
5.1.1. Material used in this study
The material used in this study is low carbon steel AISI grade 1010. The general chemical composition of this material is shown in Table 5.1. This material is normally used to manufacture an engine cradle and side frame of an automobile. The process to manufacture the tube is a continuous roll forming process. The type of the welding process is high frequency welding process.
Table 5.2 shows the dimensions and the mechanical properties of the tubes.
Content (%) Material C Fe Mn P S AISI 1010 0.08 – 0.13 99.18 – 99.62 0.3 – 0.6 Max 0.4 Max 0.05
Table 5.1: Chemical composition of Low Carbon Steel Tube grade AISI 1010
66 Parameters Value Outside diameter, OD (mm / in) 63.5/2.5 Tube wall thickness, t0 (mm / in) 0.079/2.0 Bulge width, w (mm / in) 76.2/3.0 Hardness, Brinell 95 Ultimate Tensile strength (MPa /ksi) 325/47.10 Yield Strength (MPa /ksi) 180 /26.1 Elongation @break (percent) 28% Modulus of Elasticity (GPa /Msi) 200 / 29.0
Table 5.2: Dimensions of the tube used in this study
5.1.2. Experimental matrix
Table 5.3 shows the experimental matrix for this study. The 6 sets of the tubes
were received from the tube suppliers of the well known car maker. Each set of the tubes
came from different coils of the sheet used to manufacture the tubes. Therefore, it could
be assumed that the properties of the sheet received may be different because the results
from the car maker on the percent scrap rates were different. In order words, the quality
of tubes produced from different coils was not the same. However, this information is
considered to be propriety. Therefore, it is not reported in this document.
The hydraulic bulge test was used to test these tubes. The locations of the measured bulge height (h) and wall thickness (t) were at 3 different locations around the circumference: welding line and 90 and 180 degree from the welding line, as seen in
Figure 5.1.
67
Figure 5.1: Tube cross-section showing different testing locations
For representation, the numbering system is Tube X.Y, where X denotes the tube
SET (numbers 1 to 6 denote SET 1 to SET 6) and Y denotes the testing location (1 denotes the bulge height measuring location as the welding line, 2 denotes the measuring location is at 90o from welding line and 3 denotes the measuring location is at 180o from welding line).
For example, a tube numbered Tube 4.3 denotes that the tube of SET 4 and the bulge height is measured at 180o from the welding line.
Each experiment was replicated 3 times in order to ensure the consistency of the
measurement results. The results obtained from the experiments were bulge height versus
internal pressure and flow stresses calculated by using the analytical model described in
Chapter 4.
68 Tube set Tube Bulge height measurement No. of tubes number location tested Tube 1.1 at the welding line 3 1 Tube 1.2 900 from the welding line 3 Tube 1.3 1800 from the welding line 3 Tube 2.1 at the welding line 3 2 Tube 2.2 900 from the welding line 3 Tube 2.3 1800 from the welding line 3 Tube 3.1 at the welding line 3 3 Tube 3.2 900 from the welding line 3 Tube 3.3 1800 from the welding line 3 Tube 4.1 at the welding line 3 4 Tube 4.2 900 from the welding line 3 Tube 4.3 1800 from the welding line 3 Tube 5.1 at the welding line 3 5 Tube 5.2 900 from the welding line 3 Tube 5.3 1800 from the welding line 3 Tube 6.1 at the welding line 3 6 Tube 6.2 900 from the welding line 3 Tube 6.3 1800 from the welding line 3
Table 5.3: Experimental matrix
5.2. Experimental results
5.2.1. Experimental results of tube set no. 1 Figure 5.2 represents the measured bulge height vs. pressure of each tube.
According to this figure, the variation of bulge height versus internal pressure at different location around the circumferential directions was very small, less than 1 percent.
However, when comparing the maximum bulge at the bursting pressure at the different locations, it shows that at 90 degree of the welding line, the bulge height is the highest
(see Figure 5.3). At the welding line, the maximum bulge height is the lowest. This means that the welding line was stronger than the base material. 69 The flow stress curves obtained from all the specimens for tube set #1 are shown in Figure 5.4. The parameters are listed in Table 5.4.
For the flow stress results obtained from different sets of the tubes are summarized and tabulated as seen in Table 5.5.
Figure 5.2: Pressure vs. bulge Height curves for LCS 1010 tubing; OD = 63.5 mm (2.5 in) and t0 = 2.00 mm (0.079 in) obtained from hydraulic bulge test for tube set no. 1
70
Figure 5.3: Maximum bulge height and percent thinning at the bursting pressure at different location around the circumference for LCS 1010 tubing; OD = 63.5 mm (2.5 in) and t0 = 2.00 mm (0.079 in) obtained from hydraulic bulge test for tube set no. 1
Value Parameters Tube 1.1 Tube 1.2 Tube 1.3 Strength Coefficient –K- (MPa /ksi) 643.8/93.4 652.7/94.7 634.8/92 Strain Coefficient –n- 0.220 0.226 0.207 ε 0.01 0.01 0.01 Pre-strain - 0 - Max effective strain 0.363 0.400 0.376 Maximum bulge height (mm/in) 10.41/0.41 11.42/0.45 10.78/0.42 Maximum thinning (%) 28.75 31.4 29.65
Table 5.4: Flow Stress of LCS 1010 tubing to Krupkowsky’s Law n (σ = K(ε 0 + ε) ) obtained from an analytical technique obtained by hydraulic bulge test for tube set no. 1
71
Figure 5.4: Flow Stresses for set no. 1 of LCS 1010 tubing at the different location around the circumferential directions; OD = 63.5 mm (2.50 in) and t0 = 2.0 mm (0.079 in).
72 Value Parameters Tube 2.1 Tube 2.2 Tube 2.3 Strength Coefficient –K- (MPa/ksi) 615.5/89.26 660.1/95.73 644.6/93.49 Strain Coefficient –n- 0.210 0.248 0.239
Pre-strain -ε 0 - 0.01 0.01 0.01 Maximum bulge height (mm/in) 10.23/0.40 11.68/0.46 11.74/0.46 Maximum percent thinning (%) 28.25 32.1 32.25 Value Parameters Tube 3.1 Tube 3.2 Tube 3.3 Strength Coefficient –K- (MPa/ksi) 733.3/106.35 667.7/96.83 692.8/100.48 Strain Coefficient –n- 0.311 0.250 0.277
Pre-strain -ε 0 - 0.01 0.01 0.01 Maximum bulge height (mm/in) 11.30/0.44 11.97/0.47 11.62/0.46 Maximum percent thinning (%) 31.10 32.85 31.95 Value Parameters Tube 4.1 Tube 4.2 Tube 4.3 Strength Coefficient –K- (MPa/ksi) 621.5/90.14 634.9/92.08 632.6/91.75 Strain Coefficient –n- 0.196 0.204 0.206
Pre-strain -ε 0 - 0.01 0.01 0.01 Maximum bulge height (mm/in) 9.43/0.37 10.13/0.40 10.06/0.40 Maximum percent thinning (%) 26.20 28.05 27.85 Value Parameters Tube 5.1 Tube 5.2 Tube 5.3 Strength Coefficient –K- (MPa/ksi) 656.2/95.17 643.7/93.36 649.7/94.22 Strain Coefficient –n- 0.241 0.241 0.248
Pre-strain -ε 0 - 0.01 0.01 0.01 Maximum bulge height (mm/in) 10.68/0.42 10.91/0.43 11.71/0.46 Maximum percent thinning (%) 29.20 30.05 32.15 Value Parameters Tube 6.1 Tube 6.2 Tube 6.3 Strength Coefficient –K- (MPa/ksi) 600.6/87.10 634.3/92.00 659.4/95.63 Strain Coefficient –n- 0.196 0.243 0.261
Pre-strain -ε 0 - 0.01 0.01 0.01 Maximum bulge height (mm/in) 10.03/0.40 11.30/0.45 11.93/0.47 Maximum percent thinning (%) 27.75 31.10 32.75
Table 5.5: Summary of all the flow stress and results of the tube set numbers 2 to 6 obtained from the hydraulic bulge test
73 5.3. Discussions
5.3.1. Effect of sheet properties used to manufacture the tubes
The variation of the formability of the tube may come from the variations in the mechanical properties of the sheet used to manufacture the tube or the roll forming and welding processes. In order to investigate the effect of the variations in sheet properties,
six sets of the tubes were tested by using the hydraulic bulge test. To avoid the effect of
the roll forming and welding processes, the results at 180 degree from the welding line
are discussed.
Figure 5.5 shows the variation of the maximum bulge height at the bursting
pressure. According to the results, Tube set# 6 has the highest bulge height among the
tube sets. Tube set#4 has the lowest maximum bulge height, while Tube sets #2 and #5
have the same maximum bulge height.
Figure 5.6 illustrates the strain hardening coefficient (n). As known, the strain
hardening coefficient (n) is usually used to identify the formability of the materials. The
higher the strain hardening, the better the formability is going to be. Tube set #3 and 4
has the highest the strain hardening coefficient, while Tube set#4 has the lowest the strain
hardening coefficient.
Maximum percent thinning has been practically used as a fracture criterion in the
metal forming processes, because it is easy to use and measure directly from the part.
Also the thickness is used to determine the quality of the products. As seen Figure 5.7,
overall the maximum thinning varies in the range of 25 to 31%. Tube set#4 has the
74 lowest maximum thinning, while tube set#2, 3 and 6 has the almost the same maximum thinning of 32%.
In conclusion, the maximum bulge height at the bursting pressure seems to be able to distingue the formability variations of the tubes among other criteria. It can be seen that Tube set # 3 and 6 are the best quality tubes.
Figure 5.5: Maximum bulge height at the bursting pressure measured at the different location around the circumferential direction of each tube set.
75
Figure 5.6: Strain-hardening coefficient (n-value) in each location around the circumferential direction of each tube set
Figure 5.7: Maximum percent thinning at different locations around the tube circumference
76 5.3.2. Effect of the roll forming and welding processes
In order to study the effect of the roll forming and welding processes the formability variations (bulge height and percent thinning) of Tube set #3 and 6 would be compared. According to Figure 5.8, the maximum bulge heights of Tube set#3 are evenly distributed around the circumference, while for Tube set#6 the maximum bulge heights are varied as seen in Figure 5.9. This indicates the strong influence of the roll forming and welding processes. At the welding line, the material seems to be stronger than the base material. Therefore, during deformation the material at the welding line may pull the material at the neighborhood that would cause of the local deformation near the welding line. This observation could be confirmed by the thinning distributions shown in Figure
5.9. At the welding the maximum thinning is much lower than that at the rest of the areas.
Even though the tube set #6 is considered to be the best quality among all the tubes, tube set#3 is still considered to be better because of the uniform deformation around the circumference.
77
Figure 5.8: Maximum bulge height and percent thinning at different locations around the circumference of Tube set#3
Figure 5.9: Maximum bulge height and percent thinning at different locations around the circumference of Tube set#6
78 5.4. Summary and conclusions
• 6 sets of the tubes under this study came from the same supplier but different coils
show a significant difference in the formability as well as flow stress..
• 6 sets of tubes were tested to identify that which set has the highest formability.
According to the all the criteria used to measure the formability of the material, it
concludes that the tube set no. 3 has a highest formability among those sets of the
tube. However, when considering the effect of the roll forming and welding
processes Tube set#3 is considered to be the best quality.
• Among the criteria discussed above, the maximum bulge height seems to provide
the distinguishable results from the different sets of the tubes. Therefore, the
maximum bulge height may be the good indicator for the evaluation of the quality
of the tubes.
• Hydraulic bulge test can be used to identify the difference in the tube properties
and as a tool for evaluating the quality of incoming tubes.
79 CHAPTER 6
DESIGN OF WARM TUBE HYDROFORMING SYSTEM
In this chapter, the new design so called “Submerged Design Concept” of warm
tube hydroforming is explained. The design criteria and some of the potential problems in
the design are described. The medium used to heat and pressurize the tube and die is
liquid. Extensive tests were performed on the liquids in order to ensure the safety of the system. The heat transfer analysis would be used to select the power of the heating system. The forming die was designed based on the selected geometry and the heating channel used to heat the forming die would be determined. 3D Finite Element Method
(DEFORM 3DTM) was used to approximate the temperature distributions at the die
surface in order to ensure that the selected heating system as well as the heating channel
design could be enough to provide the heat energy to uniformly heat the die surface.
Later, the temperature distributions would be confirmed experimentally.
6.1. Design considerations
Preliminary studies have demonstrated that the formability of magnesium and
aluminum alloy tubing is greatly dependent on the forming temperature. [Patil, 2000]
Therefore, it is crucial that a chosen hydroforming tool at elevated temperatures should
provide a uniform and steady temperature distribution throughout the die and the tube 80 surfaces. For conducting the preliminary warm tube hydroforming studies, the hydraulic tube bulge test tooling as mentioned in Chapter 4 was modified by adding the heating elements, in order to heat the tube and fluid medium to form the tube at elevated temperatures. The revised test set up allowed simple hydroforming experiments of magnesium and aluminum alloy tubes. Figure 6.1 illustrates the modified warm bulge test design without axial feeds. In other words, the tube material is restricted at both ends and
only the material at the deformation area (center of the tube) is formed.
`The tube is heated indirectly through the applied heat transfer fluid, CALFLO
HTF, which fills the tube completely. Two cartridge heaters (1000W each) are placed
inside the axial punches, transferring the heat to the fluid by the means of natural
convection. The dies are not equipped with a heating device. After the tube has reached
the required temperature the fluid is pressurized and the part is formed. In order to
achieve a constant process temperature one thermocouple has been attached on the tube
surface and a controller unit (WATLOW 935A) maintains a constant temperature. Using
this heating design it is theoretically possible to heat the tube up to 300°C.
81 Support Plate
Cartridge Hydraulic Punch Cylinder Cylinder Rod Tube Heater
Clamping Die Bottom Die Base Plate Holder
Figure 6.1: Schematic of warm hydraulic bulge test [Patil, 2002]
During the first tests at elevated temperatures, in order to maintain the uniform temperature distributions of the tube was difficult. As a result, the tube expanded non- uniformly, with the main expansion occurring on the top face (see Figure 6.2).
To properly investigate this problem, thermocouples were attached on the tube surface at four different locations as shown in Figure 6.2. Figure 6.3 shows the temperature measurement. According to the results, the temperature on the top face
(location A) is significantly higher than that of the other locations [Patil, 2002] over the entire heating period.
One could explain that this phenomenon is due to the effect of natural convection between the internal heater element and the fluid in the upper part of the tube being significantly stronger than that in the lower section, since cool layers of fluid stay on the
82 bottom and prevent natural convection in the lower area of the tube. This effect contributes to the observed temperature difference of approximately 40°C.
A C
D B
Figure 6.2: Asymmetric expansion due to non-uniform fluid temperature distribution during forming [Patil, 2002]
300
250 C) C)
° 200
° 200 e ( e ( r r u u 150 at at er er p p m m 100 Temp. at A e e T T Temp. at B Temp. at C 50 Temp. at D
0 0 5 10 15 20 25 Time (min)
Figure 6.3: Temperature measurement at the tube areas (see Figure 6.2)
83 The fact that areas C and D exhibit almost equal temperatures, although they are on opposite areas of the tube can be explained by a strong dissipation of heat from the tube surface to the die, due to conduction. Even though the natural convection heat transfer to the top region is much stronger, the conduction between the tube and die overbalances this superior heat supply. In the bulging area however, the tube surface is surrounded by air, which reduces the heat flux to the environment, and contributes to the temperature difference between the surfaces. This negative effect could be addressed by a steady heat source that surrounds the tube and heats it additionally from the outside.
6.2. Proposed design for warm tube hydroforming process
Figure 6.4 shows a schematic overview of the warm hydroforming system. The
THF tooling was designed for installation in a 160-ton Minster hydraulic press.
A schematic section view of the die tooling is shown in Figure 6.5. The basis for this proprietary hydroforming process, which differs from conventional hydroforming systems, is the use of a "submerged" approach whereby the formed part is totally submerged in a heated heat transfer fluid before and during the hydroforming process.
The upper and lower dies are heated and submerged in the heating fluid tank and the axial punches feed the tube ends into the expansion zone during part forming. The docking rods that are attached to the punches penetrate the fluid tank walls and are sealed with o- ring type seals to allow axial movement while preventing heating fluid leakage from the tank. The forming dies and the heat transfer fluid contained within the tank, illustrated in
Figure 6.5, are heated to the optimum part forming temperature. The heat transfer fluid preheats the tube to forming temperature, acts as the hydroforming pressurizing medium, 84 and lubricates the dies. This concept suggests a novel method to eliminate part blank pre- heat and pressurizing fluid pre-fill time, making it competitive with conventional room temperature hydroforming.
Figure 6.7 demonstrates the concept of the submerged concept. The heated dies are located inside the tank. In the open stage, only the lower die is emerged in the heat transfer fluid, as seen in Figure 6.8. Upon the closure of the upper die, a displacement body that floats on the fluid bath is forced into the fluid that causes of the level of the heating medium to increase which at the same time pre-fills and heat the tube. Details of the operation sequence are described in Appendix D.
85 160 Ton Minster Axial Feed Press Pr. Gage Cylinder
TOOLING Motor Pump Flow Rate Control
Tank Temp. Axial control Feed Pr. Intensifier Control
Control Signal High Pressure Line Medium Pressure Line Thermocouples
Figure 6.4: Warm hydroforming system
Dynamic seal Dynamic seal
Figure 6.5: Section of the designed tool and the names of parts.
86 6.3. Part selections
Figure 6.6 shows a schematic of a selected part for this project. This part geometry was selected because it demonstrates the various combinations such as rectangular and circular cross sections and transition zones that may represent the complex part geometry normally occurring in the industrial parts produced by the hydroforming process.
Section A-A Section B-B Section C-C Section D-D
B C A D
A D B C
Figure 6.6: Schematic of the part selected for this study
87 Upper die
Displacement body
Fluid tank Tube Lower die
Submerged die
Fluid bath
Submerged tube
Figure 6.7: Schematic to demonstrate the submerged design concept
88
Figure 6.8: A picture of submerged design concept- Dies are emerged inside the hot liquid bath
6.4. Specification of warm tube hydroforming system
6.4.1. Determination of maximum flow rate and volume required to form the
selected part
In order to compute the flow rate, a preliminary computer simulation was used.
The flow rate function was considered as linear function as seen in Figure 6.10 and several maximum flow rates were examined to find the condition which deforms the tube without wrinkling and within the strain rate limit of 0.5 sec-1. [Droder, 1999]
89 As seen in Figure 6.9, the deformed tubes (a), (b), and (c) have wrinkles when the maximum flow rate is smaller than 3.6 in3/sec (0.92 GPM). Therefore, the maximum flow rate was selected as 3.6 in3/sec (0.92 GPM) and the flow rate function shown in
Figure 6.10 satisfied the requirements. The internal volume change of the tube using the
selected flow rate is shown in Figure 6.11. The maximum flow rate for designing the tool
is selected as 2.0 GPM in order to give an enough range to explore variable flow rate in the optimizing process. The total volume change of the tube is 17 in3 (0.07 gallon). The flow rate curve was determined for a 30-sec cycle time.
90 (A) (B) (C) (D)
Figure 6.9: Deformed shape of the tube according to the flow rate. The flow rate function is linear as shown in Figure 6.10 and the maximum flow rates are: (a) 1.4 in3/sec, (b) 2.2 in3/sec, (c) 2.8 in3/sec, and (d) 3.6 in3/sec. Only (d) does not make any wrinkle.
91 1
0.8 ) M P
G 0.6 ( e t a
r 0.4 ow Fl 0.2
0 0 5 10 15 20 25 30 Time (sec)
Figure 6.10: Flow rate curve of the pressure fluid
80
70
60 ) n i
- 50 u c
( 40 e m
lu 30 o V 20
10
0 0 5 10 15 20 25 30
Time (sec)
Figure 6.11: Tube internal volume is changing as the tube deforms. This volume is obtained from the simulation.
92 Axial feed speed
0.25 0.2 sec) / B
n 0.15 i A 0.1 eed ( p
S 0.05 0 0 102030
Time (sec)
Figure 6.12: Axial feed speed of the punches.
6.4.2. Axial feed control
The selected part has complicated cross sections that are not symmetric along the longitudinal direction. Therefore, the amounts of needed axial feed for both ends differ.
The optimal axial feed has been investigated through the preliminary computer simulation, and is shown in Figure 6.12.
Additional information obtained from the part selection task provided the nominal process parameters of the warm hydroforming process needed to form the selected part.
Table 6.1 shows the resulting specifications necessary to form the selected part, which was used as a guide to specify tooling requirements for the warm hydroforming prototype system.
93 Description Value Unit Max needed clamping force 60 Tons Axial feed velocity 1.30 in/sec Max axial displacement 2.63 in Max pressure 4000 psi Flow rate 0 – 1.0 GPM Temperature 482/250 ˚F/˚C Internal volume change 17 cu-in
Table 6.1: Process parameters of the warm hydroforming system
6.4.3. Estimation of heating system unit
Figure 6.13 illustrates the cross section of the die and the layer of insulators. The capacity of heating system unit used in this process were estimated from Equations:
Qstore Ptotal = + qLoss Equation 6.1 t required
where
o Qstore = the amount of heat energy required to heat the forming die to 250 C
o Trequired = the total time required to heat the forming die up to 250 C
qLoss = the amount of heat loss due to heat conduction, convection and radiation
Calculation of total heat stored in the die
Qtotal = mc P (Tdesigned − Troom )
Where
3 Vdie = Die block – die cavity = 6inx6inx19in – 125.1= 558.9 in
m = die mass = 0.282lb/in3 x 558.9 in3 = 157.61 lbs
o Cp (H-13 tool steel) = 0.11 BTU/lb- F 94 o Tdesigned = 482 F
o Troom = 77 C
Qtotal 7021.53 qstore = = =11.70BTU =12.34KW t required 10 min x60sec
Press bed
Cooling plate Ceramic insulator Upper die Mica insulator qh qh Tank
Air Air Lower die Mica insulator Ceramic insulator Cooling plate
Press bed
Figure 6.13: Cross section of the forming dies
Calculation of heat loss to environment
Heat loss due to conduction
Since the die was installed on the insulator, the heat loss due to the conduction can be neglected.
95 Heat loss due to convection
At the outer die surface only natural convection occurs between the die surface and air or fluid around the die. The die is surrounded by the heated liquid during the experiment. However, in order to maximize the amount of heat loss, it is assumed that the
o die is surrounded by the air (Tair = 50 C). The air layer at the die surface heat up faster
than the ambient layers and begin to rise, due to density differences, with the maximum
fluid velocity close to the die wall. This naturally induced air/fluid flow develops to a
laminar or turbulent flow condition, which is not expressed in terms of the Re number,
but with the Rayleigh number. [Incorpera, 1981]
The convection coefficient “h” for all die surfaces can be approximated by using
Equation 6.2:
k h = Nu ⋅ = 7.91 w/m2.K (Equation 6.2) L
where:
L - Die height = 0.1524 m
A widely accepted Nusselt correlation that describes external natural convection in the laminar zone is [Incropera, 2002]:
1/ 6 2 h ⋅ L ⎪⎧ 0.387Ra L ⎪⎫ Nu = = ⎨0.825 + 9 /16 8 / 27 ⎬ = 35.7 (Equation 6.3) k ⎩⎪ [1 + (0.492 / Pr) ] ⎭⎪
where Ra is the Rayleigh number. The coefficient C and the exponent n depend on the Rayleigh number range that determines, whether the flow is laminar or turbulent.
The Rayleigh number can be calculated with [Incropera, 1981]:
96 g ⋅ β ⋅ ()T − T ⋅ L3 Ra = S ∞ =16.46x106 (laminar zone) (Equation 6.4) ν ⋅α where: g - Gravitational force (9.8m/s2)
β - Expansion coefficient of air (1/Tf = 0.0024 1/K)
ν - Kinematic viscosity of air/fluid (26.4x10-6m2/s) [Incropera, 1981]
L - Length of the die surface (die height) = 0.1524m
o TS - Die surface temperature (250 C or 523K)
o T∞ - Air temperature/Fluid temperature (50 C)
α - Diffusivity of air (38.3x10-6 m2/s) [Incropera, 1981]
Tf -[Tf = (TS +T∞ )/2] = 423 K
with:
v α = (Equation 6.5) Pr where:
ν Pr - Prandtl number = = 0.69 α
ν - Kinematic viscosity of air
• qh = hA(TS − T∞ ) = 7.91x0.193x(250 − 50) = 306.19W
Heat loss due to radiation
• 4 -8 4 qr = ε nσTs A= 0.4x5.67x10 x (523) x 0.1935 = 328 W
Therefore the total power for the heating system is
97 Ptotal = 12340 + 306.19 + 328 = 12975W
With the safety factor of 3, the total power of the heating unit is 38.9KW
According to the catalog of the heating unit available, the heating unit with the
power of 48KW was selected in this study.
6.4.4. Fluid selection
Since this novel submerged hydroforming concept relies critically on the
employment of a heat transfer fluid, early efforts were directed during the preliminary design phase towards identifying the optimum heat transfer fluid into which to submerge the hydroformed parts. Many heat transfer fluid were obtained for testing. The ideal fluid will have a high flash point and will also able to withstand the 7.5 ksi maximum expected system pressure. Thermal stability tests were conducted with selected candidate heat transfer fluids to determine which one works best this application. Table 6.2 illustrates some of the candidate fluids and some of their physical properties.
Heat transfer fluids are limited in their operation as determined by flash point, fire point, and auto-ignition temperatures that are now defined. The flash point, as defined by the National Fire Protection Agency (NFPA) is defined as the lowest temperature at which a flammable liquid gives off sufficient vapor to form an ignitable mixture with air near its surface. Ignition of these vapors can occur in the presence of an ignition source, but ignition is not self-sustaining. The fire point is defined as that temperature at which a flame becomes self-sustained so as to continue burning the liquid. The auto-ignition temperature describes the minimum temperature to which a substance must be heated,
98 without the application of an ignition source, which will cause the substance to ignite.
Consequently for safety reasons, the desirable characteristics of a heat transfer fluid are those with high flash, fire and auto-ignition temperatures.
Dow Royal Company Petro Canada Paratherm SASOL Chemical Dynalene Purple Corp Hy- Marlo- Calflo™ Paratherm Dow Dynalene Product Therm™ therm™ HTF ™ NF Therm 550 ™ 600 707 SH Flash (°F) 439 345 460 421 322 600 Fire (°F) 462 385 520 475 342 626 Auto ignition 687 690 N/A 842 716 N/A (°F) Viscosity cSt @ 104 35.6 20 32 16.5 N/A 72 cSt @ 212 6 3.5 5.9 3.1 N/A 29 cSt @ 600 0.73 0.55 N/A 0.43 0.58 5 Thermal K 0.078 - 0.052- 0.08 - 0.07 N/A 0.07 - 0.05 0.057 (BTU/hr F ft) 0.065 0.061 2.5-1.6 E6 Bulk Modulus 1.91 E6 psi N/A N/A N/A N/A psi
Table 6.2: Specifications of the heating fluids
A major limitation in the use of a commercially available heat transfer fluid is
their property of rapidly oxidizing at elevated temperatures when exposed to oxygen.
Evidence of oxidation is evolution of smoke at increased temperatures, an increase in
fluid viscosity, and color-change. It was found that most of these fluids were designed to
work in closed-loop systems (i.e., systems not exposed to the atmosphere), and tend to
degrade when heated in the presence of air. Since the hydroforming approach pursued
99 here employs an open tank filled with a heat transfer fluid exposed to the air, methods of minimizing this oxidation problem had to be identified. Vendor suggestions to prevent fluid oxidation in an open bath apparatus were the use of an inert gas (nitrogen) blanket over the exposed fluid surfaces. Although it is possible to enshroud the dies and tank within an inert gas chamber, the system would require additional methods of moving parts into and out of the chamber, somewhat complicating the part-flow through the process. Use of fluids capable of operating at elevated temperatures while exposed to the air simplifies the design of the system by eliminating the need for the inert gas chamber described above. A continued search ultimately revealed that silicone-based fluids resist oxidation better at elevated temperatures than paraffinic or aromatic hydrocarbon-based fluids. As a contingency, the inert gas chamber option was considered a possible alternative if the evaluation of the silicone fluids proved unsatisfactory. One such product called Dynalene™ 600 (Dynalene Heat Transfer Fluids, Whitehall, PA) is advertised to be capable of being specifically used in open bath apparatus up to temperatures of 300°C, which adequately exceeds the expected nominal processing temperatures. Several other heat transfer fluid samples were requested from the vendors for in-lab testing and evaluation. Three candidate fluids were ultimately selected and evaluated at elevated temperatures. These were: a) Dynalene™ 600, b) Calflo™ HTF and c) Dow Corning™ 550. Appendix C describes details of the testing method and
results. Dynalene™ 600 was ultimately selected as the fluid used.
100 6.5. Design of the heating channels
In order to reach a stable temperature in a reasonably short time, an efficient heat transfer from the fluid to the die must be provided. This can be achieved with a high heat transfer coefficient in the heating channel, which is mainly dependent on the flow condition. The flow condition is expressed in terms of the Reynolds number and could be laminar (Re < 2300), transient (2300< Re < 4000) or turbulent (Re > 4000) [Wagner,
2004]. Since the heat transfer coefficient (convection coefficient) is significantly higher in a turbulent flow condition, a high Reynolds Number above 4000 is desired. The
Reynolds Number (Re-Number) is calculated as follows:
V ⋅δ Re = (Equation 6.6) v
V - Fluid velocity
δ - Channel diameter
v - Kinematic viscosity
Based on the Equation 6.6, a high fluid velocity, a considerable channel diameter
and a low fluid viscosity result in a high Re-Number. The maximum volumetric flow rate
of the heating medium is limited by the flow rate of the selected heating pump (20 GPM).
In order to achieve a decent fluid velocity per die, a single channel design was chosen,
which then results in a flow rate of 10 GPM per die. With the given flow rate, a heating
channel diameter was designed. Designing a large channel diameter per die would be
beneficial for an efficient temperature distribution, but the fluid velocity decreases
reciprocally proportional. A compromise between a reasonable channel diameter and 101 fluid velocity was found in a diameter of 12.7 mm (1/2 inch). The fluid velocity is then calculated to be 5.12 m/s per die. As mentioned before, the fluid that will be used in this system is “Dynalene 600”, which provides a high resistance to viscosity breakdown at elevated temperatures. With the given fluid velocity, channel diameter and fluid viscosity
(ν = 8.5·10-6m2/s @ 250°C), the Re-number in the heating channel was calculated to be
7650. This is in the range of a turbulent flow.
Figure 6.14 shows the pattern of the heating channel for the lower die. The channel runs parallel and close to the cavity surface to achieve a uniform surface temperature along the longitudinal axis and combines a total length of 1700 mm (67 in) per die.
102
Figure 6.14: Schematic of heating channel for heating the selected die geometry (dimensions in centimeters)
6.6. Insulation
In order to maintain a consistent fluid bath temperature and to minimize heat
losses from the die to the press beds two kinds of insulation were designed.
A heat flux reduction from the lower die to the tank and from the upper die to the
upper press bed was achieved by inserting a thin layer (0.036 inches) of MICA insulation sheet between the dies and die shoes. Figure 6.15 illustrates the insulation of the tooling.
MICA sheet insulation combines a very low thermal conductivity, a small thickness and high compression strength, which makes it suitable for the die shoe insulation. To further reduce excessive heat transfer from the tooling to the press beds, ceramic insulation plates and water-cooled aluminum plates are inserted between the tank and lower press bed and between the upper die and upper press bed. 103 Catch basin
FOAM Glass Insulation Insulation Fluid tank Lower die
MICA Insulation Lower die shoe Sheet (0.036 in) Tank Wall CERAMIC Insulation Catch Water Cooling Basin Plate
Press Bed
Figure 6.15: Schematics of the Warm THF-tooling with the installed insulation
In order to reduce heat dissipation from the hot tank to the environment “FOAM
GLASS” insulation around the outer tank walls was installed. Figure 6.16 illustrates the insulated tank. Due to the hot conditions on the tank wall, it was difficult to attach the insulation with adhesives only and a mechanical fixture was designed in order to assure a tight fit between insulation and tank wall. The chosen design of metal straps distributes
104 the pressure uniformly across the porous insulation and outperforms localized clamping fixtures.
Side View Front View
Fixture for Insulation connecting plates Metal the straps straps 10. 25 0 i
Docking n Insulation rod plates passage 37 inches 12.725 in
Figure 6.16: Schematics of the insulated tank
“FOAM GLASS” insulation is an all glass closed-cell structure and was selected based on its high temperature resistance, low conductivity and high chemical resistance in the case of contact with the heating fluid. The material properties of “FOAM GLASS” are summarized in Table 6.3. With a low conduction coefficient of 0.085 W/m·K and an insulation thickness of roughly 1 inch at the tank side and 1.5 inches at the tank front, it is theoretically possible to reduce the tank surface temperature from 250˚C to 65˚C and
45˚C, respectively.
105 Description Value Max. operating temperature [ºF/ºC] 900/482 Thermal conductivity [W/m⋅K ] 0.085 @ 250ºC Density [kg/m3] 128 @ 24ºC Specific heat [kJ/kg⋅K] 0.84 @24ºC
Table 6.3: Material properties of FOAM GLASS insulation
6.7. Thermal analysis of the forming die
To ensure that the process heater requirements selected were sufficient to heat the
die to the designed temperature of 482°F (250˚C), 3D thermal FEM analyses were
conducted using DEFORM 3DTM.
6.7.1. Geometric Modeling and Boundary Conditions of the Die
The initial geometries of the die were modeled in the CAD Software Solid Edge
V15. A surface mesh was created using the build-in mesh generator of DEFORM 3D with an element size between 4 and 12 mm. In particular, the following mesh sizes were assigned: a) the die cavity mesh element size was 3 mm, b) the outer die surfaces were assigned an element size of 8 mm, and c) the heating channel mesh element size was 4 mm. This resulted in a total element number of 72334. In order to reduce the
computation time for the simulation, only ¼ of the entire die was modeled, because the
die is symmetric along the longitudinal axis (Figure 6.17). Figure 6.18 illustrates the
applied symmetry conditions. The die was assumed to be rigid.
106
Figure 6.17: Lower die and quarter die (used for the thermal simulations)
Figure 6.18: Boundary conditions used to determine temperature distributions at the die surface (h = convection coefficient, W/m2-K)
107 6.7.2. Thermal properties
The material data of the tool steel H-13 was obtained from the DEFORM 3DTM material database, which provides the data over a wide temperature range. Table 6.4 lists the thermal material data of H-13 at 250˚C.
Description Value for H-13 @ 250°C Thermal conductivity [W/m⋅k] 24.55 Heat capacity [N/mm2⋅K] -
Table 6.4: Material properties of H-13 [DEFORM 3DTM database]
Dynalene provides the material data for “Dynalene 600” at selected temperatures
(Table 6.5). Material properties for other temperatures than the listed temperatures had to
be interpolated or can be reviewed at [www.Dynalene.com].
108 Description Value at different temperatures Specific Heat c p 1.423 @100ºC 1.7055 @250ºC 1.799 @300 ºC [kJ/kg⋅K] Density ρ [kg/m3] 968 @25ºC 760.88 @250ºC - Thermal conductivity k 0.1557@25ºC 0.115 @250ºC - [W/m⋅k] Kinematic viscosity ν 100·10-6 @ 25ºC 8.5 x10-6@250ºC 5.5 x10-6@ 300ºC [m2/s] Volumetric expansion 1.77·10-3 coefficient β [K-1]
Table 6.5: Thermal properties of Dynalene 600 at various temperatures
Commercially available software, DEFORM 3DTM, based on a non-isothermal analysis was used for this study. This software is capable of solving thermo-mechanical problems in warm forming. In other words, this software can consider heat transfer and
deformation calculations simultaneously. For this study, only the heat transfer subroutine
was used to determine the temperature distribution of the dies and tube surfaces. The
temperature pattern will be transferred to the forming process later. The thermal analysis
of the forming die was done in the steady stage condition only, because during the
transient stage (heating up stage) the temperature of the fluid or air around changes and
this information was not known at the time. As a result, the heat convection coefficients,
which are strongly affected by the temperature change, could not be approximated.
Therefore, the heat convection coefficients were calculated based on assumed final
tooling temperatures and were considered to be constant for the entire heat transfer
calculations. From a thermodynamic point of view the tooling represents then a closed
system that will balance itself based on the energy input (assigned heat transfer 109 coefficients in the heating channel) and output (assigned heat transfer coefficients at outer die surfaces). It is noted that the assumptions for the final tooling temperature must be verified experimentally.
6.7.3. Determination of heat transfer coefficient
The bulk fluid temperature for the convection boundary condition was assumed to be 482°F (250˚C), and the convection coefficient was calculated using the material
properties of the silicone-based oil Dynalene™ 600. Using 20 GPM as the flow rate
based on the expected output from the fluid heater system and the illustrated flow pattern
with 1/2-inch diameter flow channels, the flow was found to have fully developed
turbulent flow (Re > 2300) assuming 2 channels with a 10 GPM flow rate (1 for the top
die and 1 for the bottom die) each.
With the theory of convection heat transfer it is possible to reduce the number of
parameters, thus allowing to formulate heat transfer theories generally and using
empirical correlations to solve them. Wilhelm Nusselt, who made significant
contributions to this theory, developed a dimensionless heat transfer coefficient, termed
“Nusselt Number“ [Incropera, 1981]:
δ Nu = h ⋅ (Equation 6.7) k
h - Convection coefficient
δ - Tube diameter
k - Thermal conductivity
110 The Nusselt number represents the enhancement of heat transfer through a fluid layer, as a result of convection relative to conduction across the same fluid layer and the effectiveness of the convection heat transfer increases with a larger Nusselt Number.
With Nu = 1 the heat transfer is pure conduction. By rearranging the formula, the heat convection coefficient in the die heating channel can be calculated with:
Nu ⋅ k h = (Equation 6.8) δ
Since the thermal conductivity of the heating fluid and the channel diameter are
known, only the Nusselt Number for the developing flow region and the developed flow
region has to be calculated. The literature offers a number of empirical correlations.
For turbulent pipe flow “Gnielinsky” [Holman, 1972] recommends:
0.6 < Pr < 2000 2300 < Re < 106 0 < d/L < 1
⎡ 2 ⎤ 0.5 ⋅ f ⋅ Pr⋅ (Re−1000) ⎛ d ⎞ Nu = ⎢1+ 3 ⎜ ⎟ ⎥ (Equation 6.9) 3 2 ⎢ ⎝ L ⎠ ⎥ 1+12.7 ⋅ 0.5 ⋅ f ⋅ ( Pr −1) ⎣ ⎦
1 f = 1.58 ⋅ ln(Re) − 3.28
where:
f - Friction factor
Re - Reynolds number (see Equation 3-1)
Pr - Prandtl number
d - Channel diameter
L - Channel length
111 The Prandtl Number is calculated by:
µ ⋅ c p Pr = (Equation 6.10) k
where:
µ - Dynamic viscosity of the medium
cp - Specific heat of the medium
k - Thermal conductivity of the medium
The fluid properties are to be evaluated at the bulk fluid temperature (250°C).
The Prandtl Number is the ratio of the velocity boundary layer to the thermal boundary layer and is a measurement of the relative effectiveness of the momentum and energy transport by diffusion in the velocity and thermal boundary layers, respectively.
The convection coefficient was calculated to be 1.93 BTU/hr-in2-°F (1580
W/m²K) at the heating channel, which was used in the simulation. The die was also assumed perfectly insulated at the bottom since a mica insulation sheet was used there to insulate the die. Air at a temperature of 77°F (25˚C) was assumed for the sidewall boundary conditions.
6.7.4. Finite Element Modeling (FEM)
The simulated temperature distribution of the die after 25 minutes was shown in
Figure 6.19. Figure 6.20 and Figure 6.21 show the temperature distribution at square and circular cross sections.
112 Sample locations, P1 and P2 as seen in Figure 6.19, represented at the nodes located at the farthest away from the heating channel, but still on the surface of the die that will be in contact with the hydroformed part. These nodes happen to also be the nodes with the overall minimum temperature of each cross section. The temperature plotted against time curves have been created for those locations, and are shown in Figure
6.22 and Figure 6.23 for points P1 and P2, respectively.
As seen from the curves, the temperature levels off for both nodes of interest to
466°F (241˚C). Both nodes reach a steady temperature in about 12 minutes. Therefore, it was concluded that the heating channels would provide sufficient heat to bring the temperature of the dies up in a timely fashion.
The points P1 and P2 represent the furthest location from the heating channel on the surface, which is in contact with the tubing.
113
Figure 6.19: Temperature distributions at the die surface
114
Figure 6.20: Temperature distribution at Time = 25 min at the square section (section A-A, see Figure 6.19) of the die
Figure 6.21: Temperature distribution at Time = 25 min at the circular section (section B-B, see Figure 6.19) of the die
115 300 500 250 ] ] F C 400 ° ° 200 e [ e [ r r u u 300 t 150 at a er er p p 100 200 m m e e T T 50 100
0 0 0 5 10 15 20 25 Time [min]
Figure 6.22: Temperature vs. time curve for point P1 (see Figure 6.19)
300
500 250 ] F] C 400 ° 200 [° e e [ r ur u 300 t a
at 150 r e er p p
200 m m 100 e Te T 50 100
0 0 0 5 10 15 20 25 Time [min]
Figure 6.23: Temperature vs. time for point P2 (see Figure 6.19)
116 6.8. Stress analysis
Not only should the die set be heated in a relatively short time using hot fluid circulating through heating channels, but also the die must withstand the stress that will be generated when the tubular part is internally pressurized to the maximum anticipated forming pressure of up to 5000 psi. To meet these stress requirements, a separate Finite
Element (FE) simulation analysis was performed.
AISI H13 tool steel hardened to 52~54 HRC was the material used for manufacturing the die. The material properties of this material at 482˚F (250˚C) and room temperature are shown in Table 6.6.
Description Values Young modulus 210 GPa / 30500 ksi Poisson ratio 0.3 Ultimate tensile strength @ room temp 1937.5 Mpa / 281 ksi Yield strength @ room temp 1723.8 Mpa / 250 ksi Hardness, Rockwell C 52~54 Ultimate tensile strength @ 482˚F(250°C) 1820 Mpa / 264 ksi Yield strength @ 482˚F(250°C) 1579 Mpa / 229 ksi
Table 6.6: Material property of H13 at room and 482˚F (250°C) temperature
6.8.1. Finite element model
Using the FEM software package 3D-Simulation, DEFORM 3DTM, a simulation was conducted for the structural analysis of the die. All structural simulations were based on isothermal conditions, assuming at temperature of 250°C (482oF). The boundary and loading conditions are shown in Figure 6.24, considering that the die holders are attached 117 to the die on the top and bottom. There is no support on the sidewall since the die is open to the sides. The size and pattern of the heating channels were selected from the thermal analysis, which is described in the previous section.
Symmetry
Internal Pressure
Symmetry
Fixed
Figure 6.24: Boundary and loading conditions
6.8.2. Simulation results
The simulations showed that the weakest section of the die is the square section as shown in Figure 6.25a (stress plot for the whole die) and Figure 6.25b (stress plot at the square cross section). The results also showed that the stress is concentrated at the bottom of the square section and around the heating channels. The maximum Von Mises stress of this section is approximately 10% of the yield strength of the hardened H13. Therefore,
118 the designed die structure was considered to be safe under the expected operating pressure and temperature. The analysis results are shown along with the material yield strength in Table 6.7.
119
Figure 6.25: Stress concentration after applying 5000 psi at temperature of 250°C (A) stress distributions for the whole die and (B) stress distributions at the cross section C-C
120 Stress Deflection Model Strain Remarks (MPa) (mm) Material Yield 1579 0.007519 AISI H13 @ 250˚C Strength 3D Value is approximately 161 0.00111 0.0215 Simulation 10%
Table 6.7: Computed results compared with the yield values of the tool material
6.9. Temperature measurement
After assembling the warm tube hydroforming tooling and establishing the operating sequence, the heating system was tested in order to determine the temperature distribution at the die and tube surfaces. Subsequently, preliminary parts were formed to validate the approach.
The different experiment conditions and the status are explained in the Table 6.8.
Experiment Description Die temperature measurement without fluid Set 1 in the tank. Temperature of the tube surface under Set 2 submerged condition
Table 6.8: Experimental conditions for temperature validations
The selection of the thermocouples and data acquisition for the experiments is explained in explained in Appendix B.
121 6.9.1. Set 1: Die temperature measurement without any fluid in the tank
Experimental set up
For this experiment 24 thermocouples (T/C) were attached at the dies: 21 T/Cs at the lower die and 3 T/Cs at the upper die to check the symmetry of temperature pattern between the lower and upper dies. The distribution of the thermocouples in the lower die is shown in Figure 6.26 and Figure 6.27.
Figure 6.26: Thermocouples attached in the lower die surface
122
Figure 6.27: Schematic of the thermocouple layout in the lower die
The pattern of the thermocouples in the upper die is shown in Figure 6.28 and
Figure 6.29.
Figure 6.28: Thermocouples attached in the upper die
123
Figure 6.29: Schematic of the thermocouple layout in the upper die
Experimental condition for Set 1
1) Upper and lower dies were closed without fluid in the tank
2) Dynalene™ 600 was circulated through the heating channel of both dies
with the flow rate of 10 GMP (Gallon per minute) each.
3) In order to avoid thermal shock, the temperature of the fluid was increased
stepwise (50, 100, 150, 200, 250 and 260ºC) setting at the heating system
unit.
4) After the fluid temperature reached steady state at 260ºC at the heating
system, temperature measurements were taken.
124 Experimental results
The temperature of the dies was measured in the steady temperature condition for a fluid setting temperature of 260ºC at the heating system.
Figure 6.30 and Figure 6.31 show the temperature distribution along the lower die surface.
Figure 6.30: Defined profiles in order to identify the thermocouples in the same cross section perpendicular to X direction
125
Figure 6.31: Temperature measurements with the error range in the lower die for the different profiles defined in the Figure 6.30
The maximum temperature gradient at the die surface was about 8±3.4ºC
(Therefore the error range between thermocouples must be taken in account. In other words, the overall error |±3.4ºC| was obtained by considering the error of the first thermocouple |±1.7ºC|, plus the error of the second thermocouple in the comparison
|±1.7ºC|). However the maximum temperature gradient along the same cross section is
3±3.4ºC. For example, Z3-3, Z3-2 and Z3-1 were located in the same cross section and the maximum gradient among these three thermocouples is 1±3.4ºC (which in this case means that the temperature distribution in this section is quite uniform). The minimum temperature areas were near the surfaces A & B as shown in the Figure 6.30. Therefore,
126 this cross section should not be considered in the measurement because during the forming experiment the tube will not be in contact with this area. The maximum temperature gradient between two consecutive cross sections without taking in account the first and the last section (which are nearest to the side surface of the die) was about
2.5±3.4ºC.
According to the experimental results the variation of the temperature is within the error limit. Therefore, it was concluded that an acceptable uniform temperature distribution at the lower die was achieved.
Figure 6.32 shows the comparison of the temperature measured from upper and lower dies at selected locations. The purpose of this comparison is to ensure the symmetry of the temperature for both dies. According to the results, it was concluded that the symmetry of the temperature between upper and lower dies was achieved within the variation of 3±3.4ºC.
127 Z2-1 / Z6-2
Z2-2 / Z6-3
Figure 6.32: Temperature gradient between the upper and the lower die
6.9.2. Set 2: Temperature measurement of the tube under submerged condition
Experimental Set up
Magnesium alloy tube with dimension of 2.061in OD x 0.095in wall thickness x
16in length was cut and used for this experiment. Eight thermocouples were attached to this tube around the circumferential and longitudinal directions of the tube, as seen in
Figure 6.33. The exact locations of each thermocouple were shown in Figure 6.34. In order to have a comprehensive picture about the temperature variations, there are four locations along the longitudinal direction; i.e. two locations in the guiding zone and two locations in the deformation zone, and in each location along the longitudinal direction
128 four thermocouples were attached around the circumferential direction with 90 degrees apart (0, 90, 180, and 360 degrees). Each measurement was replicated three times.
Figure 6.33: Thermocouples attached to the tube surface
Figure 6.34: Locations of each thermocouple attached on the tube
129 Experimental Conditions
During the temperature measurement at the tube, an additional temperature measurement was done at the upper die surface as well. The layout of the thermocouples at the upper die was shown in Figure 6.35. This temperature measurement would be used as a reference to investigate the effect of the temperature in the die during inserting the tube inside the die.
Figure 6.35: Thermocouples in the upper die
The procedure to run the experiments is explained, as follows: a) Upper and lower dies are closed, and the system was heated until steady stage
temperature of 260ºC at the MOKON heating system, b) After reached to the steady temperature the upper die was opened and the tank was
filled with hot fluid flowing directly from the MOKON heating unit, c) Once the tank was completely filled, the upper die was closed.
130 d) After the temperature at the upper die surface reached the steady state temperature, the
upper die was opened in order to insert the tube into the die cavity, and then the upper
die was closed. e) During the experiment the temperature at the upper die and tube surfaces was
measured until the temperature reached the steady state.
The experiments were repeated 3 times from (d) to (e) in order to ensure the consistency of the temperature measurement.
131 Experimental Results
Figure 6.36 illustrates the temperature measurement vs. time curve for all the thermocouples located at the tube and upper die surfaces. Furthermore this figure also demonstrates the experimental sequence, which was started from heating up the die, filling the tank, inserting the tube and finally waiting until the tube reaches the steady state temperature. As mentioned before the measurement error for the thermocouple is +/-
3°C.
Figure 6.36: Temperature measurements at the tube and upper die surfaces
132
Section 1 Section 2 Section 3 Section 4
250 245 240
C) 235 o
e ( 230 r u
at 225 er
p 220 m
e 215 T 210 205 200 0 degree 90 degree 180 degree 270 degree Location around the circumference
Figure 6.37: Temperature distributions around the circumference at different sections (See Figure 6.34) at the steady state condition of the tube (the measurement error = 3oC)
133 Figure 6.36 illustrates the overall picture of the temperature at the upper die and tube surfaces and the operating sequence for the test (as mentioned in the testing procedure), and Figure 6.37 shows the temperature distributions of the tube surface around the circumference at different sections at the steady state condition (t = 4000sec).
According to both figures, the important observation of the results is as follows:
• Filling the tank: After filling tank with heated fluid of 250°C directly
from the MOKON heating unit, it took almost 250 sec (4.2 min) for the
upper die surface reaching the steady state temperature of 250°C.
• Heating up the tube: After the upper die reached the steady state
temperature, the upper die was opened. The tube was inserted inside the
die, and then the die was closed. The time spent for the tube reached the
steady state temperature was 625 sec (10.42 min)
• Overall temperature gradient at the tube: The temperature gradient of
the tube surface was approximately 10°C after reaching the steady state
temperature. The maximum temperature of the tube was 240°C and the
minimum temperature of the tube was 230°C. Therefore, during the
experiments the tube would be heated by circulating the heated fluid
through the tube in order to ensure the uniform temperature distributions
at the tube surface.
134 6.10. Design of heating system
The heating system design that will be used to heat the dies and fluid bath is schematically illustrated in Figure 6.38. The hydraulic system is comprised of two sub- circuits; high pressure (blue and red lines) and moderate pressure (orange line). The high-pressure circuit begins at the hydraulic power supply (HPS) (1), where the pressure is intensified (2) to the forming pressures required to form the part. The high-pressure fluid is routed through a water cooler (3), a preheater (4), filter (5), and through a docking rod inside the formed tube. Although the intensifier will be fabricated using high temperature seals, the water cooler is used to isolate the intensifier from excessive temperatures, which will ultimately degrade seal performance. The preheater (4) is used to bring the pressurizing fluid to proper temperature so as to not chill the part during forming.
The moderate pressure circuit begins at the heat pump (6). Heat transfer fluid is circulated through the preheater (4) and both upper and lower dies to heat them to designed temperature. A set of normally closed high- pressure valves (Hipco valves) (8 and 9) is used to isolate the high pressure (~5,000 - 7,500 psi) and moderate pressure
(100 psi) circuits from each other. During operation, the Hipco valves (8 and 9) are closed and a tube blank is positioned within the closed dies and the docking rods positioned to seal the tube ends. Because the tube will be filled and preheated upon insertion into the tank, only the differential forming volume of fluid is needed to form the part. The intensifier (2) is stroked and the part is formed. The docking rods are withdrawn and the intensifier (2) retracted to recover some of the pressurizing fluid from
135 the tank in preparation for the next part to be formed. The Hipco valves (8 and 9) are added as a precaution to allow an optional flow path of heated transfer fluid through the part in the event excessive thermal gradients are experienced within tube surface.
136 Hot system (Heating system)
Cold system
Figure 6.38: Schematic of the heating system for warm tube hydroforming process
137
Figure 6.39: A picture of the warm tube hydroforming system designed for this study
138 CHAPTER 7
INVESTIGATION OF PROCESS PARAMETERS BY USING THE PROTOTYPE WARM TUBE HYDROFORMING SYSTEM
The purpose of this study was to a) investigate the effect of manufacturing processes on the quality of incoming tubes and b) process parameters (i.e. forming temperature and forming speed) on the formability of the tubes by using the prototype warm tube hydroforming system.
7.1. Experimental conditions
The experiments were divided into 2 main groups as follows:
Group A: Investigation of the quality of tube manufacturing processes: In this group the tubes manufactured by using 2 types of the extrusion processes (extrusion process with a porthole die (seam tube) and with a mandrel (seamless tubes)) were investigated. The tube manufactured by the extrusion process with the porthole die was
AZ31B-F, and the tube manufactured by the extrusion process with the mandrel was
AA6061-O.
Group B: Investigation of process parameters: The experimental results from
Group A would be used to select the type of the tube to be further investigated in details on the effect of process parameters on the formability of the tubes. In this study, the
139 effect of forming temperatures and forming rates on the formability was investigated without applying the axial feed.
7.2. Investigation of tube manufacturing process on a quality of incoming
tubes
Materials used in this study were commercially available magnesium alloy
AZ31B (seam tube) and aluminum alloy AA6061 (seamless tube). Outside diameter of these tubes is 2in (50.8 mm), and wall thickness is 0.095in (2.5mm). The tubes were cut into the lengths of 14.4 inch that can fit into the existing die.
Both aluminum and magnesium alloy tubes were formed at the temperature of
250oC (482oF) and the pressure was generated by applying a constant volumetric flow rate of 3.28x10-6 m3/s (0.2in3/s) without axial feed. The maximum percentage of
Pf − Po elongation ( %elongation = x100% , where Pf = perimeter of formed tube and Po Po
= perimeter of the initial tube) would be used as a criterion to evaluate the formability of the tubes.
7.2.1. Magnesium alloy (AZ31B) tubes (seamed tube)
Magnesium alloy (AZ 31B) tubes were manufactured by an extrusion process with the porthole. During this process, the metal divides and flows around the mandrel supports and re-welded together before they exited through the die. The quality of tubes manufactured by this method was mainly dependent on the quality of welding lines that could be the weakest areas when subjected to internal pressures. [Schuster, 2005]
140 7.2.2. Aluminum alloy (AA6061) tubes (seamless tube)
A seamless mandrel extrusion process was used to manufacture the aluminum alloy tubing used for this investigation. The process was suitable for manufacturing single cavity sections whereby the raw material billet was pierced through the center prior to the extrusion step.
Originally, the aluminum alloy tubes were received at the –T6 conditions (heat- treated condition). In this condition, the material was too strong. The available intensifier for the prototype of the hydroforming system was not be able to generate the internal pressure enough to form the tube in –T6 condition even at the high temperature
(250oC/482oF). Therefore, the aluminum alloys were fully annealed (O-conditions) under
422.2°C (760 °F) for 3 hours and were cooled at the rate of 10 °C (50 oF) per hour till the temperature reaches 283.3 °C (510 °F). After that the tube was taken out of the oven and was allowed to cool in the air.
7.2.3. Experimental results
Figure 7.1 illustrated the experimental results conducted by using magnesium alloy tubes without axial feeding. The experimental result shows no significant improvement in formability. The magnesium alloy tubes were broken exactly at the
“welding” line (the location where the material was passed through the web of the porthole die and joined, as explained in Chapter 3). The location of the welding lines was identified visually. To determine if this effect could be mitigated, forming was then
141 attempted with the assist of the axial feed, which exacerbated the problem and resulted with tubes bursting with additional severe wrinkling, as shown in Figure 7.2.
In another effort to mitigate this problem, magnesium tubes were first annealed before forming was attempted. Annealing under conditions of 300°C for 4 hrs in an oven was done to improve the formability of the tubes. Unfortunately, the bursting behavior at the welding line remained and the wrinkling behavior worsened.
The seamless aluminum alloy tubes were manufactured by extrusion with mandrel process. During the process, the mandrel may elastically deflect resulting in non-uniform wall thickness. Therefore, the wall thickness distributions around the circumference of the initial tube were measured by using a micrometer. Figure 7.3 showed the thickness variations. According to the measurement results, the variation in thickness around the circumferential direction was considerably small (~1% from the manufacturer specified wall thickness (0.095 in)). Figure 7.4 illustrated the formed tubes without axial feeding.
The maximum percent expansion was more than 80% without cracking. The maximum expansion of the tube was limited by the dimensions of the forming die.
142
Figure 7.1: Fracture of Mg tube during forming
Figure 7.2: Fracture and excessive wrinkling when forming Mg tube with axial feed
143
Figure 7.3: Thickness distributions measured around the circumferential direction from AA6061
Figure 7.4: Picture of the formed aluminum alloy tube AA6061-O conducted at the temperature of 250oC (482oF) and the volumetric flow rate of 3.28x10-6 m3/s (0.2in3/s)
According to the experimental results, aluminum alloy tube AA6061-O was selected to further study the effect of process parameters (i.e. forming temperature and flow rate) on the formability.
144 7.3. Investigation of the effect of process parameters of AA6061-O
7.3.1. Experimental procedures and conditions
In order to prevent axial movement of the tube ends and to prevent the transmission of compressive force from the axial punch to the deformation area during sealing, clamping rings were installed to provide additional compression and axial restraint within the die, as seen in Figure 7.5. After the tube was placed inside the lower die, the upper die was closed. The axial punches were then moved to seal at both ends of the tube. The amount of sealing force was calculated based on the level of yielding stress of the tube at different temperatures. The mechanical properties of aluminum alloy
AA6061-O are tabulated in Table 7.1. The equation used to calculate the sealing force is shown in Equation 7.1.The amount of sealing force for each temperature conditions was calculated as seen in Table 7.1: Dimensions and mechanical properties of aluminum alloy tube obtained by tensile test at different temperatures (AA6061-O) [Kaufman, 2002]
.
T Fseal = 2πrtσ y Equation 7.1
where Fseal = sealing force
r = outside radius of the tube
t = initial wall thickness
T σ y = Yield stress of the tube at the forming temperature
145 To demonstrate temperature and forming rate dependent formability of the aluminum alloys, the experiments were conducted at four different temperatures (room temperature, 200°C, 230°C, and 250°C) and 2 volumetric flow rates (3.28x10-6 (0.2in3/s) and 1.6x10-5 (0.98in3/s) . The maximum and minimum volumetric flow rates were determined based on the limitation of the movement of the intensifier and valve used in the hydroforming system. Table 7.3 showed the experimental conditions.
Figure 7.5: Tube with clamping rings in the forming die
146 Descriptions Values Outside diameter (mm/ in) 50.8 / 2.00 Wall thickness (mm /in) 2.413/ 0.095 Tube length (mm/in) 366 /14.4 Yield strength (MPa / ksi) @ T= 25°C 55 /8.0 Ultimate Tensile strength (MPa / ksi) @ T= 25°C 125 /18.0 Percent elongation (%) @ T= 25°C 30 Yield strength (MPa / ksi) @ T= 205°C 55/8.0 Ultimate Tensile strength (MPa / ksi) @ T= 205°C 75/11 Percent elongation (%) @ T= 205°C 60 Yield strength (MPa / ksi) @ T= 230°C 45 /6.5 Ultimate Tensile strength (MPa / ksi) @ T= 230°C 59 /8.5 Percent elongation (%) @ T= 230°C 75 Yield strength (MPa / ksi) @ T= 260°C 38 /5.5 Ultimate Tensile strength (MPa / ksi) @ T= 260°C 48 /7.0 Percent elongation (%) @ T= 260°C 80
Table 7.1: Dimensions and mechanical properties of aluminum alloy tube obtained by tensile test at different temperatures (AA6061-O) [Kaufman, 2002]
Forming Yield stress Calculated sealing force Measured sealing force temperature (ksi / MPa) (lb / kN) (lb/ kN) (oF / °C) 75/25 8.0/55 4777/21.2 5600/24.91 400/205 8.0/55 4777/21.2 5600/24.91 450 /230 6.5 /45 3900 /17.40 4800 / 22.0 500 /260 5.5 /38 3283 /14.63 4500 / 20.0
Table 7.2: Comparison of calculated and measured sealing forces used in the experiments
147 AA 6061 Forming Volumetric Flow Aluminum alloy Temperature (°C) rate (in3/sec) Tube Exp #1 Room Temp 0.98 Exp #2 200 0.20 Exp #3 200 0.10 Exp #4 230 0.20 Exp #5 230 0.98 Exp #6 250 0.98 Exp #7 250 0.20
Table 7.3: Experimental conditions
7.4. Experimental results
7.4.1. Effect of the temperature on the formability
Figure 7.6 and Figure 7.7 illustrate the experimental results with different forming temperatures. According to the results, at a temperature of 250°C without axial feed, significant improvement in formability (Maximum Percentage Expansion of 80%) can be achieved. When forming was attempted at a lower temperature of 230°C, the tube burst with lower percentage expansion. It may be possible that by feeding the material (axial feed), the bursting may be prevented and the part can be successfully formed at a lower temperature.
From the results, it could be seen that at the temperature of 250°C without axial feed, the significant improvement in formability (maximum percentage expansion of
80%) could be achieved.
148 Room temp 200oC 230oC 250oC
Figure 7.6: Picture of experimental results conducted at different forming temperatures
Figure 7.7: Maximum percentage expansion of AA6061-O at various forming temperatures (the flow rate = 0.2in3/s)
149
Figure 7.8: Comparison of pressure profiles obtained from the experiment with constant volumetric flow rate at different temperatures
7.4.2. Effect of the temperature on the pressure profiles
Figure 7.8 illustrated the effect of forming temperatures on the pressure profiles.
As mentioned before, the method to generate the pressure was by applying a constant volumetric flow rate. According to the results, except at the room temperature the pressure has dropped during forming, and then when the tube touched the die, the pressure was increased.
150 7.4.3. Forming of AA 6061-O Aluminum tubes at 230 °C with axial feed
In this experiment, only at the temperature of 230oC would be studied the effect of the axial feed, because from the pure expansion experiment at temperature of 230oC the material could not be fully formed to the die geometry. Therefore, this temperature could be used to investigate the improvement of the formability by applying the axial feed.
Experimental preparation and procedure
Tubes were cut to a length of 16 inches and de-burred at the ends, after which were lubricated by using Tri-Flow TF23015 with Teflon high temperature (800°F) lubricant. The lubricant was applied on the tube by uniformly brushing along the longitudinal direction.
The strategy used to determine the feasible loading path (axial feed and internal pressure or volumetric flow rate) within the limited time domain available for this experiment come from the observation from the pure expansion experiments. The forming procedure was divided into 2 steps:
Step A: At the first step, the pressure control technique was used to pressurize the
tube because it allowed having the reliable pressure profile that could be used to
approximate the amount of the axial feed. If the pressure generated by the
controlled flow rate, it was difficult to obtain the real pressure curve because the
amount of pressure is depended on the amount of fluid supplied to the system and
amount of axial feed. In this stage the amount of internal pressure used was
151 determined from the pressure generated in the pure expansion experiments. The
amount of axial feed was approximated by the experimental trials.
Step B: Once the tube surface was touched with the die surface, the axial feed
would be stopped. Then, the method used to apply the pressure would switch
from pressure-control to flow rate-control. The constant flow rate would be used
to generate the pressure until the tube filled at the die corner radius. This stage
was called the calibration stage.
Figure 7.9 showed the loading path that was used successfully to form a part. The experiments were replicated twice times.
Figure 7.11 showed a picture of formed parts. The profile of the tube was measured along the longitudinal direction, as seen in Figure 7.10, and thickness distributions along the circumferential direction of rectangular and circular (section A-A and section B-B in Figure 7.13, respectively). Cross sections are shown in Figure 7.12 and Figure 7.13, respectively. The thickness variation was fairly uniform for both cross sections. The minimum wall thickness of 0.062 or 34.4% wall thinning (% wall thinning
t f − ti = *100% , where tf =final wall thickness and ti = initial wall thickness) occurred at ti the rectangular cross section.
152
Figure 7.9: Loading path obtained by the experiments of forming part at the temperature of 230°C
Figure 7.10: Comparison between the forming dies geometry and the measurement of 4 profiles (A, B, C, and D)
153
Figure 7.11: Picture of the formed tube and cross sections
154
Figure 7.12: Wall thickness distribution at the rectangular cross section (section A-A in Figure 7.11)
Figure 7.13: Measured thickness distribution around the circumferential direction of section B-B (see Figure 7.11]
155 7.5. Summary and discussions
A. The effect of the tube manufacturing process was studied. The 2 types
of the tubes (aluminum alloy AA6061 and magnesium alloy AZ31B)
were manufactured by using different type of extrusion processes (with
mandrel and porthole die). According to the results, the tubes
manufactured by the extrusion process with the mandrel would be
recommended in the warm tube hydroforming process, because they do
not have manufacturing defects that would cause of the early fracture
of the material.
B. The effect of the forming temperature on the formability of the tubes
was investigated. When the forming temperature increases the
formability increase. The maximum percent expansion at temperature
of 250oC is more than 80%. Due to the limitation of the geometry of
the die used in this experiment, the real maximum elongation of the
tube at 250oC could not be determined. Therefore, in the future the size
of the forming die should be increased in order to determine the
maximum formability limit of the AA6061-O at the forming
temperature of 250oC.
C. Since the internal pressure was generated by applying the constant
volumetric flow rate, the pressure output from the experiments would
reflect the real strength of the materials. In this case, the pressure
profile at the temperature above 200oC tended to increase, decrease and
156 increase again, as seen in Figure 7.8. In order to see the effect of deformation evolutions, the special measurement system to measure the bulge height versus time was designed. Details of the measurement mechanism was shown in Appendix E. Figure 7.14 and Figure 7.15 show the measurement of the bulge height evolutions versus time and an internal pressure versus time before the tube was touched the die surface at the temperature of 250oC with different volumetric flow rates (1.6x10-5 m3/s (0.98in3/s) and 3.28x10-6 m3/s (0.2in3/s)), respectively. According to both results, although the internal pressure was dropped, the bulge height was still increasing. During forming the strength of tube was reduced. The reasons for this behavior could be due to the strain softening occurring at the high temperature. According to the tensile test data presented in Chapter 8, when the strain was higher the stress was reduced significantly especially at the forming temperature of 250oC. The pressure could be approximately as
Equation 7.2 (in case of plane strain conditions):
2 t P = σ f Equation 7.2 3 rθ
where P = internal pressure, t = wall thickness, rθ = radius of the
formed tube and σ f = instantaneous yield stress
157 According to Equation 7.2, when the stress was dropped or maintain
constant while the deformation was progressed, the level of pressure
would be dropped.
Figure 7.14: Internal pressure versus time and bulge height versus time of AA6061-0 at the forming temperature of 250oC (482oF) and flow rate of 1.6x10-5 m3/s (0.98in3/s). The bulge height was measured until the tube touched the die surface.
158
Figure 7.15: Internal pressure versus time and bulge height versus time of AA6061-0 at the forming temperature of 250oC (482oF), and flow rate of 3.28x10-6 m3/s (0.2in3/s). The bulge height was measured until the tube touched the die surface.
159
CHAPTER 8
FLOW STRESS DETERMINATION OF AA6061-O AT ELEVATED TEMPERATURES
The main purpose of this chapter is to determine the flow stress of aluminum alloy tube at the elevated temperatures. The hydraulic bulge test at the elevated temperature was originally proposed to determine the flow stress. However, it could not been done due to the limit recourses available. Therefore, the uniaxial tensile test would be selected to determine the flow stress in this study. Later, the uniaxial tensile data would be modified to accurately predict the deformation behavior of the pure expansion experiments described in Chapter 7.
8.1. Tensile test
8.1.1. Test procedures and conditions
The material used in this study is aluminum alloy tube (AA6061-O). Magnesium alloys (AZ31B-F) would not be tested in this study due to the manufacturing defects.
The dimensions and chemical compositions of the AA6061-O are shown in
Chapter 7. The tensile specimens were cut by using Electro-discharged machine (EDM)
160 along the longitudinal direction based on ASTM A513 standards. Figure 8.1 illustrates the dimensions of the tensile specimens.
The tensile tests were conducted at 4 different temperatures (100, 150, 200 and
250 oC) and 3 different strain rates (0.001, 0.01 and 0.1 /s). The level of the strain rate is approximated from the process. The tests were carried out in the furnace.
Figure 8.1: Dimensions of tensile specimen [ASTM A513]
8.1.2. Flow stress results
Figures 8.2 to 8.5 represent the engineering stress-strain curves at different temperatures and strain rates. According to the results, at the temperatures of 100 and
150oC the effect of the temperature on the stress is not well pronounced, while at the 161 temperature of 200 to 250oC, the effect of forming temperature is significant. These results were confirmed by the experiments results conducted in Chapter 7, the formability up to bursting point of the tube was increased distinctively at the forming temperature which was higher than 200oC.
One of the interesting observations in the tensile results is that the level of uniform elongation is lower than that at the higher temperature as seen in Figure 8.6. In other words, at the temperature of 100oC the uniform elongation is 35%, while at the temperature of 250oC the uniform elongation is 18% at the strain rate of 0.001/s. In contrast, when comparing the total elongation the higher the forming temperature the better the maximum elongation would be. However, the maximum elongation could not be used to justify the formability of the material, because when the stress reaches at the maximum, the necking behavior starts to influence in the accuracy of the flow stress data.
162
Figure 8.2: Engineering stress-strain curves obtained from tensile test at 100oC for different strain rates
163
Figure 8.3: Engineering stress-strain curves obtained from tensile test at 150oC for different strain rates
164
Figure 8.4: Engineering stress-strain curves obtained from tensile test at 200oC for different strain rates
165
Figure 8.5: Engineering stress-strain curves obtained from tensile test at 250oC for different strain rates
166 T = 250C T = 200C T = 150C T = 100C
40 )
% 35 30 on ( i t 25 20 onga l
e 15 m r
o 10 f i
n 5 U 0 0.001 0.01 0.1 Strain rate
Figure 8.6: Effect of strain rates and forming temperatures on the uniform elongation of AA6061-O
T = 250C T = 200C T = 150C T = 100C
100
) 90
% 80 70 on ( i 60 50 40
elongat 30 al 20
Tot 10 0 0.001 0.01 0.1 Strain rate
Figure 8.7: Effect of strain rates and forming temperatures on the total elongation of AA6061-O
167 8.1.3. Constitutive models
The constitutive models are used to represent the material flow. Normally, they represent the relationship between true instantaneous stress and true strain. However, in case of forming at elevated temperatures, the effect of strain rate or rate of deformation needs to be considered into the models. These models are broadly divided into 2 types:
Phenomenological and Physically based models. Phenomenological based model has gained more interest in the numerical modeling (FEM) because the form of the model is simple and easy to use. The well known model is the Power’s law or the Nadai’s model
n n (σ = Kε where σ is true instantaneous stress, ε = true strain and “n” and “K” are material constants). The physically based model is based on the compositions and microstructures of the materials. The material parameters based on the physical model are difficult to measure. Therefore, this model has not gained interest from the researchers.
[Boogaard, 2002]
For this study, the phenomenological model would be used to represent the constitutive model of aluminum alloys at the elevated temperature. However, the Power’s law needed to be modified to represent the effect of strain rate or deformation rate, since the deformation rate has a significant effect on the material behavior at the elevated temperatures. Equation 8.1 represents the constitutive model used in this study.
168 • m(T ) n(T ) σ = K(T )ε ε Equation 8.1
• where σ = true stress, ε = true strain, ε = true strain rate and “K (T)”, “n (T)” and “m (T)” are material constants
The material constants (K, n and m) in the equation are functions of forming temperatures.
In order to ensure the accuracy of the use of Equation 8.1, the relationships between true stress and true strain and true stress and strain rate need to be examined.
[Takuda, 2005] It is noted that the stress and strain data were done only before the necking behavior starts, because the data after the necking is considered in accurate.
Figures 8.8 and 8.9 show the plot of the true stress and true strain in the log-log scale at the temperatures of 200 and 250oC, respectively. According to the results, the relationship between true stress and strain is linear. Therefore, the dependence of the stress on the strain can be expressed by using the strain hardening (n).
Figures 8.10 and 8.11 illustrate the relationships in the log-log scale between true stress and strain rate at the temperature of 200 and 250oC, respectively. The relationships are linear and can be represented by using the strain rate hardening (m). Also the slope of the graph at different true strains for each forming temperature is almost the same except at the strain of 0.02 for the temperature of 250oC. Therefore, it can be assumed that the
“m” is independent to the true strain.
Figures 8.12 to 8.14 illustrate the plot of K, n and m as a function of forming temperatures. By neglecting at the temperature of 100oC the Equations 8.2, 8.3 and 8.4
169 are represents the relationships of K, n and m as a function of forming temperatures, respectively.
K = 434.59exp(−0.0031T ) (MPa) Equation 8.2
n = 1.5111exp(−0.0072T ) Equation 8.3
m = 0.0014T − 0.2323 Equation 8.4
It is noted that these equations can be represented at the temperature and strain range of 150-250oC and 0.001-0.1/s, respectively.
170
Figure 8.8: Relationship between true stress and strain in the log-log scale at the temperature of 200oC
Figure 8.9: Relationship between true stress and strain in the log-log scale at the temperature of 250oC
171
Figure 8.10: Relationship between true stress and strain rate in the log-log scale at the temperature of 200oC
Figure 8.11: Relationship between true stress and strain rate in the log-log scale at the temperature of 250oC
172
Figure 8.12: Effect of forming temperature on the strength coefficient (K)
Figure 8.13: Effect of forming temperature on strain hardening coefficient (n)
173
Figure 8.14: Effect of forming temperature on strain rate hardening coefficient (m)
174 CHAPTER 9
FINITE ELEMENT MODELING FOR WARM TUBE HYDROFORMING PROCESS
9.1. Finite element model and boundary conditions
The commercial software package DEFORM3DTM was used for finite element method (FEM). Since the process was assumed to be axisymmetric, only a quarter of the tube and forming die was used in order to reduce the computation time. The initial geometry of the tube and the die were modeled and meshed with solid elements in
IDEASTM v.10.0. Due to the part’s symmetry as seen in Figure 9.1, only a quarter of the part was modeled. The total number of elements used to model the tube was 8000 elements.
Figure 9.2 shows the boundary condition of the simulation. The tube was assumed to be fully plastic. Since the temperature distributions, as seen in Chapter 6, was reasonably constant, the isothermal condition (no change of temperature during forming) could be assumed for the simulations in order to reduce the amount of computation time.
The interface friction coefficient was approximately to be 0.1 [Boogaard, 2002]. The simulation input data was tabulated in Table 9.1. The flow stress data obtained from tensile test at temperature of 230 and 250oC with different strain rates also are shown in
Figure 9.3 and Figure 9.4, respectively. Figure 9.13, Figure 9.6 and Figure 9.7 represent 175 the pressure versus time curve used in the simulations for the forming temperature of
250oC and 230oC without axial feed and 230oC with axial feed, respectively. Figures 9.8 to 9.12 represent the profiles of the deformed tubes at different forming conditions.
176 Z X Y
Figure 9.1: Simulation model used in this study
177
CL Symmetry Z plane
X Forming die Z
Y Tube Symmetry plane
Figure 9.2: Boundary conditions for the simulation
Parameter Values Outside diameter 2.0 in / 50.8mm Initial wall thickness 0.095 in / 2.413T mm Tube length 16.0 in/ 406.4 mm Modulus of Elasticity 57 GPa /8.3X103 ksi Forming temperature 230°C / 446°C Poisson’s ratio 0.35 Yield strength 45 MPa /6.5 ksi Ultimate tensile strength 59 MPa /8.5 ksi Time step 0.05sec Interface friction coefficient 0.1
Table 9.1: Tube geometry and mechanical properties of AA6061-O
178
Figure 9.3: Flow stress curves at temperature of 230oC for different strain rates (Flow stress was fit up to the uniform elongation, and the dot line represents the extrapolated data)
Figure 9.4: Flow stress curves at temperature of 250oC for different strain rates (Flow stress was fit up to the uniform elongation, and the dot line represents the extrapolated data)
179 1800 1600 ) i s 1400 p ( e
r 1200
su 1000 es r 800 p
al 600 n r e
t 400 n I 200 0 0 20 40 60 80 100 120 140 160 180 Time (sec)
Figure 9.5: Measured internal pressure vs. time curve obtained at the forming temperature of 250°C with the volumetric flow rate of 1.6x10-5 (0.98in3/s)
Figure 9.6: Measured internal pressure vs. time curve obtained at the forming temperature of 230°C with the volumetric flow rate of 1.6x10-5 (0.98in3/s)
180
Figure 9.7: Loading path obtained by the experimental trials of forming part at the temperature of 230°C
Profile 3
Profile 2
Profile 1
Figure 9.8: Measured tube profile along the longitudinal direction obtained at the forming temperature of 230°C
181 Profile 1 Profile 2 Profile 3
1.8 1.6 1.4 n) i
( 1.2 nt e
m 1 ce a
l 0.8 p s i 0.6 D 0.4 0.2 0 03691215 Curvilinear length (in)
Figure 9.9: Measured tube profile along the longitudinal direction obtained at the forming temperature of 230°C
Figure 9.10: Location of the profile extracted to be used for flow stress determination
182 1.8 1.6 1.4 n) i
( 1.2 nt e 1 Profile A Profile B m
ce 0.8 a l p s
i 0.6 D 0.4 0.2 0 0246810121416 Curvilinear length (in)
Figure 9.11: Measured tube profile along the longitudinal direction obtained at the forming temperature of 250°C
Figure 9.12: Comparison between the forming dies geometry and the measurement of 4 profiles (A, B, C, and D)
183 9.2. Simulation results
9.2.1. Results at the forming temperature of 230oC without axial feed
Figure 9.13 shows the comparison between the FEM results and the experimental results. The maximum effective strain is approximated 0.591 in the simulation. Figure
9.14 shows a comparison between the experimental profile (averaged displacement from two profiles as seen in Figure 9.11) and simulation displacement profile with the error of
2%.
Figure 9.15 shows a comparison of the thickness distributions between the simulation and experimental results of the section A-A (see Figure 9.13). The average error of the thickness distributions at section A-A was 8%. The thickness distributions are really uniform in the simulation, while they are fluctuated in the experimental results.
This deviation was due to fact that the interface friction coefficient was assumed to be constant in the simulation, but in the real conditions the friction coefficient is a function of temperature and contact pressure. During the calibration stage, the pressure required to form the tube to the designed corner radius was high (1600psi). This may cause of then increase of the friction coefficient.
184
Figure 9.13: Comparison of the deformation between the FEM and experimental results at temperature of 250°C
Figure 9.14: Comparison of the displacement profile between the experimental and simulation results at temperature of 250°C
185
Figure 9.15: Comparison of the thickness distribution between the experimental and simulation results at temperature of 250°C
9.2.2. Results at the forming temperature of 250oC without axial feed
Figure 9.16 shows the comparison between the experimental and simulation results. The maximum effective strain is 0.591, which is the same as the results obtained for the tube formed at the temperature of 230oC. This could be explained by that the tube was formed to the same geometry and also the effect of the friction coefficient at the die surface is not significant. The material only was experienced only pure stretching.
Therefore, the maximum effective strain should be the same. The results are confirmed by comparing the wall thickness distributions between the simulation and experiments as seen in Figure 9.18. The wall thickness distributions in simulation are fairly uniform, which is similar in the case of forming at 230oC. However, the wall thickness
186 distributions in the experiments are fluctuated. Therefore, the friction effect on the material needs to be studied intensively in the future.
Figure 9.17 shows a comparison between the experimental profile (averaged displacement from three profiles as seen in Figure 9.11) and simulation displacement profile. According to the figure, the average error between profiles was 1.2%.
Figure 9.16: Comparison of the deformation between the FEM and experimental results at temperature of 230°C
187
Figure 9.17: Comparison of the displacement profile between the experimental and simulation results at the temperature of 230°C
Figure 9.18: Comparison of the thickness distribution between the experimental and simulation results at temperature of 230°C
188 9.3. Results at the forming temperature of 230oC with axial feed
9.3.1. Simulation results
The loading path (axial vs. time and internal pressure vs. time) as seen in was applied into the FEM. At the last stage of the simulation Figure 9.19 shows the strain contour plot of the simulation results. The maximum effective strain is 0.475.
A rectangular cross section (Section A-A in Figure 9.20) was cut, and the thickness distribution in this cross section was extracted as shown the plot in Figure 9.20.
The minimum wall thickness of 0.067 in occurred at the corner of this area.
Figure 9.19: Contour plot of the effective strain distribution at the final stage of the simulation
189 0.075
0.07 ) ) ) n n n i i i ( ( (
s s s 0.065 1 2 s s s A 3 e e e n n n 4 k k k c c c i i i 0.06 5 h h h t t t ll ll ll a a a A W W W 0.055 Measurement location
0.05 0123456 Measurement location
Figure 9.20: Wall thickness distribution along the circumferential direction at the rectangular cross section
190
Stage 0 (t=0 sec)
Stage 1 (t=30 sec)
Stage 2 (t=45 sec)
Stage 3 (t=148 sec)
Figure 9.21: Deformation behavior at different stages
191 As mentioned before, the simulation was used as a tool to illustrate the forming behavior at each step of the forming. Figure 9.21 shows the forming behavior at each stage. The forming stages were divided into three stages as follows:
Stage 1: The tube was pressurized until it yielded. During this stage the axial feed at both ends were needed in order to prevent leakage.
Stage 2: Internal pressure was maintained be constant while the amount of axial feed was increased to supply the material to the deformation area. It was observed that at the end of this stage the tube at the rectangular side started to wrinkle.
Stage 3: After the amount of axial feed was sufficient to form the part, the internal pressure was increased by supplying a constant volumetric flow rate of 0.20 in3/s until the part was completely formed. During this stage the application of axial feed still needed in order to prevent leaking.
At the final stage of the simulation, a small wrinkle was produced in the model as seen in Figure 9.22. This wrinkle could come from the computational error in FEM.
192 Wrinkle
Figure 9.22: Simulation result demonstrating a small wrinkle at the formed tube
9.3.2. Comparison of the simulation and experimental results
After the simulation was completed, the profile of the formed tube obtained from the simulation was compared with that obtained from the experiments, as seen in Figure
9.23. The simulation results matched with the experimental results very well. The error was approximately 1%.
Figure 9.24 showed the comparison of the wall thickness distribution between the simulation and experimental results. The maximum error was 8%.
193 2 Simulation result Experimental result 1.8 1.6 n) n) 1.4 (i (i 1.2 nt nt e e m m 1 Profile direction ce ce Profile A a a l l 0.8 p p s s 0.6 Di Di ProfileD ProfileB 0.4
0.2 ProfileC Profile direction 0 0246810121416 Curvilinear length (in)
Figure 9.23: Comparison of the displacement profile between the simulation and the averaged experimental results
194 Experimental result Simulation result 0.09 0.08
) 0.07 4 3 2 0.06 5 1 ss (in e
n 0.05 6 20
ick 0.04 7 19 Measurement location 8 18
ll th 0.03 a 9 17
W 0.02 10 0.01 11 16 12 13 14 15 0 0 2 4 6 8 10121416182022 Measurement location
Figure 9.24: Comparison of wall thickness distribution between the simulation and experimental results at the rectangular cross section
195 CHAPTER 10
OVERALL SUMMARY AND CONTRIBUTIONS
The overall theme of this research could be divided into 2 major areas; a) mechanical property (flow stress) determination and investigation of the effect of manufacturing process on the formability variation of the steel tube at room temperature and b) design and analysis of warm tube hydroforming process for the lightweight alloys.
The hydraulic bulge test was selected in this study to determine the material properties of tubular materials. In addition to using the deformation theory, an analytical model based on the incremental strain theory (assumed non-linear strain paths) was developed and used to determine the wall thickness at the apex of the dome and the curvature radius. The thickness predictions were compared with those measured from the experiments and calculated with both incremental strain and deformation theories [Aue- u-lan, 1999]. The predictions agreed well with measured values at low bulge heights (less than 12mm). When the bulge height was higher than 12mm, the calculations based on the deformation theory did not give accurate results, while the calculations based on the incremental theory were within acceptable accuracy.
196 In the roll-formed tube, the variation of the formability came from the mechanical property variations of the sheet prior to the roll forming process and the roll forming and welding processes. In this study, 6 sets of the tubes produced from strip at different locations from the rolled sheet were studied. Unfortunately, the locations of the strip were not identified from the rolled sheet. The strips were passed through the same roll passes to make the tubes. The experiments were conducted by using the hydraulic bulge test.
The bulge height was measured at 3 different locations around the circumference.
According to the experimental results, the maximum bulge height at the bursting pressure could be used as a criterion to indicate the formability of different tube sets.
A prototype elevated temperature THF machine was designed, built, and tested using a fluid pressurizing and heating medium. Silicone-based fluids resist oxidation better at elevated temperatures than paraffinic or aromatic hydrocarbon-based fluids and are recommended for use in this application. The external fluid heating system used performed more reliably and required less maintenance than the electric heating methods using electric cartridges.
Both AZ-31B magnesium and 6061-O aluminum were tested in the prototype machine and showed improved formability at elevated temperatures. Aluminum 6061-O tube could be completely formed at elevated temperature showing expansion percentages approaching 80-percent at 250°C. Magnesium AZ-31B tubes could not be fully formed, due to what is believed to be metallurgical and structural defects in the tubes introduced using an extrusion process with a porthole die. These tubes that were not of seamless design, were acquired due to commercial availability. To eliminate this defect, it is
197 recommended that seamless tubing be selected for the elevated temperature THF process.
The submerged method of tube blank heating provided essentially uniform tube temperature distributions. It is recommended that a tube blank reservoir feature be incorporated into the prototype design to further demonstrate the advantage of the submerged method in speeding the fabrication process in a production mode of operation.
Flow stress data of AA 6061-O tubes were determined using tensile tests at different temperatures (150, 200 and 250oC) and strain rates (0.001, 0.01 and 0.1 /s). The
• m n Power’s law ( σ = Kε ε ) was used to fit the flow stress data up to the uniform elongation because after the uniform elongation necking starts resulting in the inaccuracy in flow stress data. FEM model was used to model the hydroforming process with loading path (internal pressure vs. time and axial feed vs. time) at the temperature of
230oC. Agreement was found between the simulations and experimental results. It is recommended that the FEM model used in this study be used as a basis for future optimization of the process conditions.
The following research contributions resulted from this dissertation work:
• Development of an analytical model, based on the incremental theory, to predict the
wall thickness at the apex of the dome in the bulge test
• Development of an analytical model to calculate the flow stress of a tube at room
temperature, by using the tube bulge test
• Demonstration of the effect of tube manufacturing processes on the formability
variations in the tube. For this purpose, tubes from the different batches were used
198 and deformation around the tube circumference was studied, using the hydraulic
bulge test
• Design and development of the prototype of the warm tube hydroforming system
• Demonstration of the use of the advanced Finite Element Method (FEM) to design
and analyze the prototype warm tube hydroforming process
• A fundamental understanding of the effect of the process parameters (i.e. forming
temperatures and rates) on the formability of the lightweight alloy tubes at elevated
temperatures
• Demonstration of the use of Finite Element Method (FEM) to simulate and optimize
the warm tube hydroformig process by predicting the “best” pressure versus time and
axial feed versus time curves.
199
REFERENCES
[Altan, 1983] Altan, T., Oh, S.I., and Gegel, L.H., 1983 "Metalforming - Fundamentals and Applications", ASM International.
[Altan, 1999] Altan, T., 1999 "Formability and Design Issues in Tube Hydroforming" proceedings of the International Conference on Hydroforming Stuttgart, Germany, October 12/13, p. 105.
[Altan, 2001] Jiratheranat, S., Strano, M., Altan, T., 2001 “Adaptive FEM Process Simulation for Hydroforming Tubes”, Proceedings of the International Conference on Tube Hydroforming, Fellbach (near Stuttgart), Germany, Nov 6 –7, pp. 363-384
[ASTM, 513-00] ASTM, Annual Book of Standards, Vol. 01.01, 2002, ASTM International, PA, USA A513-00: Specification for Electric- Resistance-Welded Carbon and Alloy Steel Mechanical Tubing
[Aue-u-lan, 2001] Aue-u-lan Y, Palaniswamy, H., Ngaile, G., and Altan T, 2001 "Determination of biaxial formability and flow stress for AZ31B magnesium alloy sheet and tube", ERC Report No; THF/S/ERC/NSM-01-R-07, 2001.
[Avedesian 1999] Avedesian, M. and Buker, H., 1999 "Magnesium and magnesium alloys", ASM International.
[Aue-u-lan, 2002] Aue-u-lan, Y., Patil, S., Altan, T., 2002 “Determination of biaxial formability and Flow Stress of Tubes for Low Carbon Steel Tubes”, ERC report THF/ERC/NSM-02-R-14
[Aue-u-lan, 1999] Aue-u-lan, Y., Koc, M., and Altan, T., 1999 “Prediction of Tubular Material Properties for Aluminum Alloy 6260-T4,” THF/ERC/NSM-99-R-09
200 [Bobbert, 1997] Bobbert S., Bischer M., Ahmetoglu M., Altan T., 1997, “Tool and Process Design for Tube Hydroforming: A State-of-the-art Review and Applications of Computer Simulations”, ERC report number: THF/ERC/NSM- 97-R-02.
[Behrens, 2005] Benhrens, B.A. and Kruz, G., 2005 “Heated hydro-mechanical deep drawing of magnesium sheet metal”, Production Engineering, vol. XII/1, pp. 21-26
[Behrens, 2004] Behrens, B.A., Kruz, G. and Hubner, S., 2004 “Heated hydro- mechanical deep drawing of magnesium sheet metal”, Proceedings of International Conference “New Developments in Sheet Metal Forming”, in Fellbach (near Stuttgart)- Germany, May 11th and 12th, 2004, pp. 379-397
[Bolt, 2001a] Bolt, P. J., Lamboo, N. A. P. M., and Rozier, P. J. C. M., 2001 "Feasibility of warm drawing of aluminum products", Journal of materials processing and technology, Vol 115, pp 118-121.
[Bolt, 2001b] Bolt, P. J., Werkhoven, R. J. and van de Boogaard, A. H., 2001 "Effect of elevated temperatures on the drawability of aluminum sheet components", Proceedings of ESAFORM, pp 769-772.
[Boogaard, 2001a] van de Boogaard, A. H., Bolt, P. J. and Werkhoven, R. J., 2001 "Aluminum sheet forming at elevated temperatures", Simulation of Materials Processing: Theory, Methods and Applications, pp. 819-824.
[Boogaard, 2001b] van de Boogaard, A. H., Bolt, P. J. and Werkhoven, R. J., 2001 "A material model for aluminum sheet forming at elevated temperatures", Proceedings of ESAFORM, pp 309-312.
[Braunig, 2003] Braunig, S., During, M., Hartmann,H. and Viehweger, B., 2003 “Magnesium sheets for industrial application”, Magnesium – Proceedings of the 6th International Conference Magnesium Alloys and Their Applications, pp. 955-961
[Carleer, 2000] Carleer, B., Kevie, G., Winter, L., and Veldhuizen, B., “Analysis of the Effect of Material Properties on the Hydroforming Process of Tubes”, Journal of Materials Processing Technology, Vol.104, 2000, pp/158-166 [Carrino, 1999] Carrino, L. and Giuliano, G., 1999 “Finite element modelling and the experimental verification of superplastic forming”, Advanced Performance Materials, vol. 6, pp. 149-159
201 [Choi, 2004] Choi, Y., Aue-u-lan, Y., Shah, K., Guza, D. and Altan, T, 2004, “Hydroforming lightweight aluminum components from tube- Development and commercialization of a novel elevated temperature hydroforming system - Semi-annual report-”, ERC/NSM report number: THF/ERC/NSM-04-R-39
[Chenot, 1996] Chenot, J.L., Massoni, E. and Fourment, L., 1996"Inverse problems in finite element simulation of metal forming processes", Engineering Computations, Vol. 13, No. 2/3/4, pp. 196-225.
[Carrino, 1997] Carrino, L. and Giuliano, G., 1997 “Modelling of superplastic blow forming”, International Journal of Mechanical Science, vol. 39, no. 2, pp. 193-199
[Chenot, 1998] Chenot, J.L. and Bay, F., 1998 “An overview of numerical modelling techniques”, Journal of Materials Processing Technology, vol. 80-81, pp. 8-15
[Doege, 2001] Doege E. and Dröder K., 2001, “Sheet metal forming of magnesium wrought alloys – formability and process technology”, Journal of materials processing technology, 115
[Doerr, 2003] Doerr, J., 2003, “Innenhochdruckumformen mit Werkzeug-bzw- Medienerwärmung”, TU-Darmstadt-Institut fuer Umformtechnik und Werkzeugmaschinen
[Dohman, 1992] Dohmann, F. and Böhm, A., 1992, “The Significance of Process Simulation in Liquid Bulge Forming” (translated from German), Bänder Bleche Röhre v. 1, pp. 26-34.
[Dohman, 1996] Dohmann, F. and Hartle, C., 1996, “Developments and Applications of Internal High Pressure Forming “(in German), Blech Rohre Profile, v. 43 pp. 30-35.
[Dohmann, 1997] Dohmann, F. and Hartl, Ch., 1997, “Tube hydroforming-research and practical application”, Journal of Materials Processing Technology, Vol.71, Issue-1, pp. 174-186
[DOE, 2003] Aue-u-lan, Y., Guza, D. and Altan, T., 2003 DOE Phase II Semi- annual Report”, 2004, “Elevated temperature tube hydroforming of lightweight materials”, ERC/NSM report number: THF/ERC/NSM-03-R-49a
202 [Dohmann, 1993] Dohmann, F., Bohm, A., Dudziak, K.U., 1993 “The shaping of hollow shaft-shaped workpieces by liquid bulge forming”, 1993 Advanced Technology of Plasticity –Proceeding of the Fourth International Conference on Technology of Plasticity, pp. 447- 452 [Doege, 1997] Doege, E. and Droder, K., 1997 "Processing of magnesium sheet metals by deep drawing and stretch forming", MATERIAUX & TECHNIQUES, Vol. 7-8, pp 19- 23.
[Doege, 1999] Doege, E. and Janssen, St., 1999 "Forming of magnesium alloys - a solution for light weight construction", Society of automotive engineers.
[Doege, 2000] Doege, E., Elend, L.E. and Meiners, F., 2000" Comparative study of massive and sheet light weight components formed of different light weight alloys for automobile applications", Proceedings of ISATA, pp 87- 94.
[Doege, 2001] Doege, E. and Kurz, G., 2001"Development of a formulation to describe the work softening behavior of magnesium sheets for heated deep drawing process" Annals of the CIRP 50/1. pp. 177- 180
[Doege, 2001] Doege, E., Sebastian, W., Droder, K. and Kurtz, G., 2001 "Increased formability of Mg-sheets using temperature controlled deep drawing tools", Innovations in processing and manufacturing of sheet materials, pp 53-60.
[Dunwoody, 2004] Dunwood, B.J., 2004 “Series production of aluminum body panels using a new press concept”, Proceedings of International Conference “New Developments in Sheet Metal Forming”, in Fellbach (near Stuttgart)- Germany, May 11th and 12th, 2004, pp. 353-360
[Emley 1966] Emley, E.F., 1966"Principles of Magnesium Technology", Pergamon Press.
[El-Morsy, 2002] El-Morsy, A., Manabe, K., Kang, D.M. and Hwang, J.K., 2002 “FE analysis on temperature and deformation of magnesium alloy sheet in warm deep-drawing process”, NUMISHEET 2002, October 21-25, 2002, Jeju Island, Korea, pp. 171-176
203 [EFB, 2003] Europäische Forschungsgesellschaft für Blechverarbeitung e.V., (European Research Association for Sheet Metal Working), 2003 “Temperierte Umformung von Blechen aus Magnesiumknetlegierungen mittels Hochdruck-blechumformung (HBU)”, EFB-Forschungsbericht Nr. 209, Sept 2003.
[Fridlyander, 2002] Fridlyander, I.N., Sister, V.G., Grushko, O.E., Berstenev, V.V. and Sheveleva, L.A., 2002“Aluminum alloy”, Metal Science and Heat Treatment, vol. 44, Nos 9-10, pp. 365-370
[Frielink, 1999] Frielink, R., 1999 “New Solutions for lightweight bodies”, NedCar, Automotive Body International Quarterly.
[Friedrich, 2001] Friedrich, H. and Schumann, S., 2001 “Research for a “new age of magnesium” in the automotive industry”, Journal of Materials Processing Technology, vol. 117, pp. 276-281
[Fuchizawa, 1987] Fuchizawa, S., 1987, “Influence of plastic anisotropy on deformation of thin-walled tubes in bulge forming,” Advanced Technology of Plasticity, v. II, pp 727-732.
[Fuchizawa, 1984] Fuchizawa, S., 1984, “Influence of strain-hardening exponent on the deformation of thin-walled tube of finite length subjected to hydrostatic internal pressure,” Advanced Technology of Plasticity, v. I, pp 297-302.
[Fuchizawa, 1993] Fuchizawa, S., Narazaki, M., 1993, “Bulge test for determining stress-strain characteristics of thin tubes,” Advanced Technology of Plasticity, pp 488-493.
[Gardner, 2001] Gardner, B.R., 2001 “The business case for the use of hot metal gas forming”, MetalForming, issue October 2001, pp. 36-37
[Ghosh, 1994] Ghosh, A.K., 1994 “A new physical model for superplastic flow”, Materials Science Forum, vol. 170-172, pp. 39-46
[Gronostajski, Gronostajski, Z., 2000 “The constitutive equations for FEM 2000] analysis”, Journal of Materials Processing Technology, vol. 106, pp. 40-44
[Gavrus, 1996] Gavrus, A., Massoni, E. and Chenot, J.L., 1996"An inverse analysis using a finite element model for identification of rheological process parameters", Journal of materials processing and technology, Vol 60, pp 447-454.
204 [Groche, 2002] Groche, P., Huber, R., Dorr, J. and Schmoekel, D., 2002 “Hydromechanical deep-drawing of aluminum-alloys at elevated temperatures”, Annals of the CIRP, vol. 51/1/2002, pp. 215-218
[Groche, 2005] Groche, P. and Dorr, J., 2005 “Tube hydroforming at locally elevated temperatures”, Production Engineering, vol. XII/1, pp. 63-66
[Groche, 2005] Groche, P., Peter, A. and Schafer, R., 2005 “Modeling of tube hydroforming processes with adaptive friction coefficients”, Production Engineering, vol. XII/1, pp. 81-84
[Geiger, 2005] Geiger, M., Heyd, G., Merklein, M. and Hubnatter, W., 2005 “Novel concept of experimental setup for characterization of plastic yielding of sheet metal at elevated temperatures”, Advanced Materials Research, vol. 6-8, pp. 657-664
[Geiger, 2005] Geiger, M., Merklein, M. and Hecht, J., 2005 “Warm hydroforming of magnesium sheets –results and microstructural changes”, ICTP 2005
[Geiger, 2002] Geiger, M. and Merklein, M., 2002 “Adaptive design of aluminum sheets for deep drawing processes”, Production Engineering, vol. IX/1, pp. 59-62
[Hambli, 2001] Hambli, R. and Potiron, A., 2001 “Comparison between 2D and 3D numerical modeling of superplastic forming processes”, Computer Methods Applied Mechanic Engineering, vol. 190, pp. 4871-4880
[Hosford, 1993] Hosford, W. and Caddell, R., 1993, “Metal Forming: Mechanics and Metallurgy” Chapter 2, 2nd ed. Prentice-Hall, Inc.
[Hora, 1999] Hora P., Skrikerud M. and Longchang T., 1999, “Possibilities and Limitations of FEM-Simulation in Hydroforming Operations”. Hydroforming of Tubes, Extrusions and Sheet Metals vol.1, ed. By K. Siegert.
[Holman, 1972] Holman, J.P., 1972, “Heat Transfer”, 3rd Edition, McGraw Hill
[Incropera, 1981] Incropera F.P., DeWitt D.P., 1981, “Introduction to Heat Transfer”, 4th Edition, Wiley [Jager, 2004] Jager, S., 2004 “Forming of magnesium sheet metal”, Proceedings of International Conference “New Developments in
205 Sheet Metal Forming”, in Fellbach (near Stuttgart)- Germany, May 11th and 12th, 2004, pp. 399-414
[Kim, 2002] Kim, K.J., Lee, C.H., Yang, D.Y., (2002) “Investigation into the improvement of welding strength in three dimensional extrusion of tubes using porthole dies”, Journal of Material Process Technology, vol. 130-131, pp. 426-431
[Kalpakjian , 2001] Kalpakjian, S., Schmid, S.R., (2001) “Manufacturing engineering and technology”, 4th edition Prentice-Hall, Inc., Upper Saddle River, New Jersey, 07458, pp. 320-329
[Kaese, 2003] Kaese, V., Greve, L., Juttner, S., Goede, M. and Schumann, S., 2003 “Approaches to use magnesium as structural material in car body”, Magnesium –Proceedings of the 6th International Conference Magnesium Alloys and Their Applications, pp. 949- 945
[Kleiner, 1999] Kleiner, M., Hellinger, V., Homberg, W. and Klimmek, Ch, 1999 “Trends in sheet metal hydroforming”, Hydroforming of tubes, Extrusions, and Sheet Metals, Edited by K. Siegert, vol. 1, pp. 249-260
[Koç, 1998] Koç M., Sokolowski T., Gerke K., Ahmetoglu M. and Altan T., 1998,“Evaluation of Tube Formability and Material Characteristics in Tube Hydroforming”, ERC/NSM report number: THF-98-R-25
[Keigler, 2004] Keigler, M., Bauer, H., Harrison, D.K. and Silva, A.K.M.D., 2004 “FE-simulation of the thermal hydroforming process”, LS- DYNA Anwenderforum, Bamberg 2004,pp. 1-8
[Kim, 1996] Kim, Y.H., Hong, S.S., Lee, J.S. and Wagoner, R.H., 1996 “Analysis of superplastic forming processes using a finite- element method”, Journal of Materials Processing Technology, vol. 62, pp. 90-99
[Kim, 2002] Kim, K.J., Lee, C.H. amd Yang, D.Y., 2002 “Investigation into the improvement of welding strength in three-dimensional extrusion of tubes using porthole dies”, Journal of Materials Processing Technology, vol. 130-131, pp. 426-431
[Keigler, 2005] Keigler, M., Bauer, H., Harrison, D. and Silva, A.K.M.D., 2005 “Enhancing the formability of aluminum components via temperature controlled hydroforming”, Journal of Materials 206 Processing Technology, vol. 167, pp. 363-370
[Koç, 1999] Koç, M., Aue-u-lan, Y. and Altan, T. 1999"On the Characteristics of Tubular Materials for Hydroforming Design Rules, Analysis and Experimentation" submitted to the International Journal of Manufacturing & Machine Tools, December 1999.
[Lass, 2003] Lass, J.F., Metz, A., Wappelhorst, M., Hombergsmeier, E., Lanzerath, H., Baumgart, P. and Cordini, P., 2003 “InMak- Innovation magnesium combined structures for car bodies”, Magnesium –Proceedings of the 6th International Conference Magnesium Alloys and Their Applications, pp. 1026-1033 [Lee, 2001] Lee, K.S. and Huh, H., “Numerical Simulation of the Superplastic Moving Die Forming Process with a Modified Membrane Finite Element Method”, Journal of Materials Processing Technology, 2001, v. 113, pp. 754-760 [Liew, 2003] Liew, K.M., Tan, H., and Tan, M.J., 2003 “Finite element modeling of superplastic sheet metal forming for cavity sensitive materials”, Transaction of the ASME, vol. 125, pp. 256-259
[Laue, et. al., 1976] Laue, K., Stenger, H., (1976) “Extrusion –Process, Machinery, Tooling” American Society for Metals
[Li, 2000] Li, D., Ripson, P. and Ghosh, A. K., "Warm forming of aluminum alloys", DoE- LEMS Project Quarterly Reports, 2000.
[Li, 2004] Li, H.Y., Wang, X.S., Yuan, S.J., Miao, Q.B., Wang, Z.R., 2004 “Typical stress states of tube hydroforming and their distribution on the yield ellipse”, Journal of Materials Processing Technology, vol. 151, pp. 345-349
[Li, 2004] Li, D. and Ghosh, A.K., 2004 “Biaxial warm forming behavior of aluminum sheet alloys”, Journal of Materials Processing Technology, vol. 145, pp. 281-293
[Lorenz, 2005] Lorenz, D. and Roll, K., 2005 “Modelling and analysis of integrated hotforming and quenching processes”, Advanced Materials Research, vol. 6-8, pp. 787-794
[Matsumoto, 2002] Matsumoto, R. and Osakada, K., 2002 “Lubrication and friction of magnesium alloys in warm forging”, Annals of the CIRP, vol. 51/1/2002, pp. 223-226
207 [Malinowski, 1994] Malinowski, Z., Lenard, J.G. and Davies, M.E., 1994 “A study of the heat-transfer coefficient as a function of temperature and pressure”, Journal of Materials Processing Technology, vol. 41, pp. 125-142
[Moll, 2003] Moll, F., Mekkaoui, M., Schumann, S. and Friedrich, H., 2003 “Application of Mg sheets in car body structures”, Magnesium – Proceedings of the 6th International Conference Magnesium Alloys and Their Applications, pp. 936-942
[Marciniak, 1987] Marciniak, Z. and Konieczny, A., “The Flow Stress of Metals in Warm Forming”, Advanced Technology of Plasticity, 1987, v.1, pp. 17-22
[Manabe, 1983] Manabe, K. and Nishimura, H., 1983, “Influence of material properties in forming of tubes”, Bänder Bleche Rohre, 9/1983 pp 266-269.
[Moran, 2003] Moran M., 2003, “Introduction to thermal systems engineering : thermodynamics, fluid mechanics, and heat transfer”, Wiley
[Müller, 1997] Müller, W. and Pöhlandt, K., 1997, “Direction Related Behavior of Sheet Metal in the Cross-Tensile Test” (in German), Bänder Blech Rohre Profile, 3/1997 pp43-45.
[Naka, 1999] Naka, T. and Yoshida, F., 1999 “Deep drawing of type 5083 aluminum-magnesium alloy sheet under various conditions of temperature and forming speed”, Journal of Materials Processing Technology, vol. 80-90, pp. 19-23
[Naka, 2001] Naka, T., Torikai, G., Hino, R. and Yoshida, F., 2001"The effects of temperature and forming speed on the forming limit diagram for type 5083 aluminum-magnesium alloy sheet", Journal of materials processing and technology, Vol. 113, pp 648-653.
[Naka, 2003] Naka, T., Nakayama, Y., Uemori, T., Hino, R. and Yoshida, F., 2003 “Effect of temperature on yield locus for 5083 aluminum alloy sheet”, Journal of Materials Processing Technology, vol. 140, pp. 494-499
[Novotny, 2001] Novotny, S., 2001 "Hydroforming of sheet metal pairs from Al and Mg alloys using a warm pressure fluid", Lehrstuhl fur Fertigungstechnologie, Annual Report, pp 53.
208 [Neugebauer, 2005] Neugebauer, R., Sterzing, A., Kurka, P. amd Seifert, M., 2005 “The potential and application limits of temperature-supported hydroforming of magnesium alloys”, ICTP 2005
[Neugebauer, 2003] Neugebauer, R. and Seifert, M., 2003 “Results of tempered hydroforming of magnesium sheets with hydroforming fluids”, Magnesium –Proceedings of the 6th International Conference Magnesium Alloys and Their Applications, pp. 306-312
[Novotny, 2002] Novotny, S., Geiger, M. and Feldmann, K., 2002, “Innerhochdruck-Umformen von Blechen aus Aluminum- und Magnesium-legierungen bei erhoehter Temperatur”, Bericht aus dem Lehrstuhl fuer Fertigungstechnologie, Universitaet Erlangen, PhD dissertation (in German)
[Patil, 2002] Patil, S., Shah, K., Aue-u-lan, Y., Pisoni, C., Kaya, S., Ngaile, G., Altan, T. and Guza, D., 2002, “Elevated temperature tube hydroforming of light weight materials”, ERC/NSM report number: THF/ERC/NSM-03-R-49
[Pai et. al., 1992] Pai, D., Livatyali, H., Tszeng, O., Kinzel, G., and Altan, T., “Roll Forming Process State of the Technology and Analysis”, ERC/NSM Internal Report 92-53-S
[Patil, 2002] Patil, S., Aue-u-lan, Y., and Altan, T., “Variations in the Material Properties of Tubes Used for Hydroforming”, Tube and Pipe Journal, September, 2002, pp.76
[Patrick, 2004] Patrick, A., T. and Ghassan, T.L., 2004 “Warm forming of aluminum alloys; an overview and future directions”, International Journal of Materials & Product Technology, vol. 21, issue 1/2/3, pp. 24-40
[Palaniswamy, Palaniswamy, H., Ngaile, G. and Altan, T., 2004 “Finite element 2004] simulation of magnesium alloy sheet forming at elevated temperatures”, Journal of Materials Processing Technology, vol. 146, pp. 52-60
[Pfaffman, 2003] Pfaffman, G. and Dykstra, W., 2003, “The Next Generation Process for Manufacturing Vehicle Structural Components”, Recent Developments in Tube & Sheet Hydroforming - A Short Course and International Workshop, Columbus, Ohio, June 18- 19, 2003
209 [Pietrzyk, 2000] Pietrzyak, M., 2000 “Finite-element simulation of large plastic deformation”, Journal of Materials Processing Technology, vol. 106, pp. 223-229
[Rohsenow, 1998] Rohsenow Warren M., Hartnett J.P. and Cho Young I., 1998, “Handbook of heat transfer”, 3rd Edition, McGraw-Hill
[Rodrigues, 2002] Rodrigues, J.M.C. and Martins, P.A.F., 2002 “Finite element modelling of the initial stages of a hot forging cycle”, Finite Elements in Analysis and Design, vol. 38, pp. 295-305
[Rauscher, 2005] Rauscher, B., Gosling, M., Homberg, W. and Kliner, M., 2005 “Gas-pressure forming of an AlMg- alloy sheet at elevated temperatures”, Steel Research International, vol. 76, issue 12, pp. 925-930
[Redecker, 2005] Redecker, M., Roll and K., Haubinger, S., 2005 “Magnesium sheet metal forming considering its specific yield behavior”, Advanced Materials Research, vols. 6-8, pp. 771-778
[Samekto, 2004] Samekto, H. and Roll, K., 2004 “Finite element simulation for superplastic forming process for aluminum 5083”, Materials Science Forum, vol. 447-448, pp. 117-122
[Schuster, 2005] Schuster, C., Loretz, C. and Klaas, (2005) “Potential and limits with hydroforming of aluminum alloys”, Proceedings of the International Conference on “Hydroforming of tubes, extrusions and sheet metals” in Leinfelden-Echterdingen (near Stuttgart) – Germany, vol. 4, pp. 113-135
[Sillekens, 2003] Sillekens, W.H. and Bohlen, J., 2003 “The MAGNEXTRUSCO project: European Community Research on hydrostatic extrusion of magnesium”, Magnesium –Proceedings of the 6th International Conference Magnesium Alloys and Their Applications, pp. 1046- 1051 [Siegert, 2001] Seigert, K. and Vulcan, M., 2001 “Supperplastic Forming of Sheet Metals”, Innovation in Processing and Manufacturing of Sheet Materials, pp. 41-51
[Siegert, 2003] Siegert, K., Jager, S. and Vulcan, M., 2003“Pneumatic Bulging of Magnesium AZ31 Sheet Metals at Elevated Temperatures”, Annals of the CIRP
210 [Shah, 2003] Shah, K., Aue-u-lan, Y., and Altan, T., “Hydraulic Bulge Tooling to Determine the Properties of Tubular Materials”, ERC/NSM Internal Report THF/ERC/NSM-03-R-27
[Shirayori, 2003] Shirayori, A. Fuchizawa, S., Ishigure, H., Narazaki, M., (2003) “Deformation behavior of tubes with thickness deviation in circumferential direction during hydraulic free bulging”, Journal of Materials Processing Technology, vol. 139, pp. 58-63
[Schuler, 2005] Schuler Hydroforming, (2005) “How to measure tube quality? Result of burst test study”, presented at the Hydroforum meeting on June 2005 by Schuler Hydroforming Inc.
[Singh, 2003] Singh, H., (2003) “Fundamentals of Hydroforming”, AFFT - Association for Forming & Fabrication Technologies- and SME – Society of Manufacturing Engineers-
[Schmoeckel, 1992] Schmoeckel, D., Hessier, C. and Engel, B., 1992 “Pressure Control in Hydraulic Tube Forming”, Annals of the CIRPv. 41/1/1992, pp. 311-314
[Shehta, 1978] Shehta, F., Painter, M. J. and Pearce, R., 1978 "Warm forming of aluminum/magnesium alloy sheet", Journal of mechanical working technology, Vol 2, pp279-290
[Sokolowski, 2000] Sokolowski, T., Aue-u-lan, Y. and Altan, T., 2000 "Evaluation of Tube Formability and Material Characteristics" Journal of Materials Processing Technology, vol. 98, No. 1, January 15, pp. 34.
[Takuda, 2003] Takuda, H., Fujimoto, H. and Inoue, D., 2003 “Effect of warm forming on mechanical properties of an AZ31 magnesium alloy sheet”, Magnesium –Proceedings of the 6th International Conference Magnesium Alloys and Their Applications, pp. 463- 468
[Takuda, 2002] Takuda, H., Mori, K., Masuda, I., Abe, Y. and Matsuo, M., 2002 “Finite element simulation of warm deep drawing of aluminum alloy sheet when accounting for heat conduction”, Journal of Materials Processing Technology, vol. 120, pp. 412-418
[Takuda, 1999] Takuda, H., Yoshii, T. and Hatta, N., 1999 “Finite-element analysis of the formability of a magnesium-based alloy AZ31 sheet”, Journal of Materials Processing Technology, vol. 89-90, pp. 135-140
211 [Taylor, 1980] Taylor, B. and Lanning, H.W., 1980 "Warm forming of aluminum - production systems", 25th National SAMPE symposium and Exhibition, pp 471-480.
[Wang, 2001] Wang, H., Martin, P., and Houghland, E., “Evaluation of Tube Materials for Tube Hydroforming”, Proceedings of 43rd Mechanical Working and Steel Processing Conference, Charlotte, NC, October 28-31, 2001, pp.251-261
[Wagner, 2004] Wagner, W., “Waermeuebertragung”, 2004, 6. Auflage, Vogel Fachbuch Kamprath Reihe P. 63- 116
[Wood, 1996] Wood, R.D. and Bonet, J., 1996 “A review of the numerical analysis of superplastic forming”, Journal of Materials Processing Technology, vol. 60, pp. 45-53
[Woo, 1978] Woo, D. M. and Hawkes, P.J., 1978 “Determination of stress/strain characteristics of tubular materials”, Journal Inst. Metals Sciences. pp. 357 - 359.
[Wu, 2001] Wu, X., Liu, Y. and Wang, S., 2002 "Superplastic Deformation and Forming of Commercial Magnesium and Aluminum Alloys With Initial Coarse-Grains", Proceedings of the 2002 NSF Design, Service and Manufacturing Grantees and Research Conference, pg 2072- 2080.
[Xing, 2002] Xing, H.L. and Makinouchi, A., 2002 “FE modeling of thermo- elastic-plastic finite deformation and its application in sheet warm forming”, Engineering Computations, vol. 19, no. 4, pp. 392-410
[Zhang, 2004] Zhang, B. and Baker, T.N., 2004 “Effect of the heat treatment on the hot deformation behavior of AA6082 alloy”, Journal of Materials Processing Technology, vol. 153-154, pp. 881-885
[www.Omega.com] http://www.omega.com/temperature/tsc.html
[www.castingsourc http://www.castingsource.com/tech_art_Magnesium.asp e.com] [www.magnesium- http://www.magnesium-elektron.com/about-magnesium.asp elektron.com] [www.matweb.com http://www.matweb.com ] [www.dynalene.co http://dynalene.com/dyn600.html m]
212
APPENDIX A
ANALYTICAL MODELS TO DETERMINE FLOW STRESS BASED ON DEFORMATION THEORY
A.1 Analytical to determine flow stress
1. Effective stress calculation
Longitudinal stress:
Prθ σ Z = 2t1
Hoop stress:
⎛ ⎛ Pr ⎞ ⎞ ⎜ θ ⎟ ⎜ ⎟ ⎜ P ⎝ 2t1 ⎠ ⎟ σ θ = ⎜ − ⎟ * rθ t r ⎜ 1 Z ⎟ ⎜ ⎟ ⎝ ⎠
Effective stress:
2 2 σ = σ θ −σ θ σ Z + σ Z
213 2. Effective strain calculation
Hoop strain:
⎛ r ⎞ ⎜ θ ⎟ εθ = ln⎜ ⎟ ⎝ r0 ⎠
Thickness strain:
⎛ t ⎞ ⎜ 1 ⎟ ε t = ln⎜ ⎟ ⎝ t0 ⎠
Longitudinal strain:
ε Z = −(εθ + ε t )
Effective strain:
2 ε = ()ε 2 + ε 2 + ε 2 3 θ Z t
Figure A. 2: Geometry of the deformed tube: nomenclature used in calculations
214 A.2 Thickness prediction based on the deformation theory
Stress and strain State at the center of the bulge needs to be known under various process conditions (i.e., various internal pressures and elongations). For this purpose, we can measure the bulge height at the center of the bulge and knowing the tool geometry we can calculate the final thickness at the same location. This chapter will explain the detailed procedure to calculate the thickness of the tube in the middle of the bulge.
Method
1) This method is based on the plasticity and membrane theories.
2) As shown in Figure A.1, Bulge height (h) can be measured from the experiments,
and Bulge width (w) is known from the tooling set-up.
3) Based on the bulge height and bulge length, the radii in hoop and longitudinal
directions (Figure A.2) can be calculated as:
The radius in hoop direction is defined as:
rθ = r0 + h Equation A-1
where r0: original tube outside radius
h: bulge height
The radius in longitudinal direction:
(w/ 2) + Re(1− sinθ ) (w/ 2) 2 + h 2 r = = Equation A-2 Z sinθ 2h
where w = bulge width, re = die corner radius and θ = contact angle
215 4) Based on the membrane analysis, a relation among internal pressure (P), hoop stress
(σθ), and Longitudinal stress (σz) can be represented as shown below.
P σ θ σ z = + Equation A-3 t rθ rz
where
t is the thickness
rθ is the final bulge radius
rz is the principle radius of curvature (in axial direction)
5) The stress in the longitudinal direction can be calculated in terms of internal
pressure, the bulge radius and thickness as:
Pr σ = θ Equation A-4 Z 2t
6) From Equation (A.3) and (A.4), the stress in the hoop direction can be derived as:
Prθ σ zrθ σθ = − Equation A-5 t rz
7) From the Equation (A-4) the relationship between stress and strain on the hoop
and thickness directions can be calculated as:
216 ε ε θ = t 1 1 Equation A-6 σ − σ − (σ + σ ) θ 2 Z 2 θ Z
where
⎡ rθ ⎤ Hoop strain: ε θ = ln ⎢ ⎥ r ⎣ 0 ⎦
⎡ t1 ⎤ Thickness strain: ε t = ln ⎢ ⎥ ⎣t 0 ⎦
8) From the Equations (A.4), (A.5), and (A.6), thickness of the deformed tube in
the middle of the bulge can be determined as:
⎡ rθ ⎤⎡ 3 rθ ⎤ ln⎢ ⎥⎢ − ⎥ r 2 2r Equation A-7 t = e A t where A = ⎣ 0 ⎦⎣ z ⎦ 1 o 3 r − θ 2 rz
217 φ θ rθ σZ
rZ
t
σθ
Figure A. 3: State of stress on an element at the apex of the hydroformed tube
218
APPENDIX B
TEMPERATURE EQUIPMENT TO MEASURE THE DIE AND TUBE TEMPERATURE
Thermocouple selection
Due to the high fluid temperature (up to 260°C) and the demand for accurate measurements the experimental conditions prove to be very challenging. In order to select a suitable thermocouple, the following criteria had to be considered:
A. Temperature range – the upper temperature limit in the
experiments was 270ºC
B. Measurement Accuracy – temperature variations between
measurement locations can be very small and need to be detected
C. Wire diameter – the thermocouple wires must fit through the fluid
release channels of the die
D. Wire Insulation – the wires had to withstand the hot heat transfer
fluid (up to 260ºC)
219 The type of thermocouple chosen is an E-Type, consisting of the metals Chromel and Constantan and exhibits an operating temperature range of -100ºC to 1000ºC. This type of thermocouple is ideally suited for lower temperature measurements (below
500°C), since the milli-volt range is very large (highest among all thermocouples), which makes it useful for detecting small temperature changes and results in more accurate temperature measurements (7.8·10–2 mV/K). In the range from 0ºC to 316ºC the standard error at a reference junction of 0ºC is given with ± 2ºC, which is one of the lowest among all standard thermocouples [www.omega.com].
Due to the high number of thermocouples attached on the die surface, several fluid release channels have to fit two thermocouple wires. Consequently, a small wire diameter was selected (0.01 inches). However, with a decreasing thermocouple wire size, the internal resistance increases drastically. As a rule of thumb, the resistance for the wire should not exceed 100 Ohms. With the considered diameter and a total wire length of 9ft
(3m) the total resistance is 31.5 Ω. This value is well below the upper limit, thus the wire dimensions can be used. It is worth to mention, that very small wire diameters (0.01 inches e.g.) are more expensive and since some of the thermocouples can run individually through the fluid release channels, a thicker and less expensive wire size was also selected. Ultimately, 20 thermocouples with a wire diameter of 0.01 inches and 5 thermocouples with a wire diameter of 0.02 inches were purchased. Table gives an overview about the specifications of the deployed thermocouples.
220 Thermocouple code Description Twenty thermocouples, PFA Teflon® 20 × TC-TT-E-30-108 insulated wire, E-Type, insulation diameter = 0.01 in, wire length of 108 in Five thermocouples, PFA Teflon® insulated 5 × TC-TT-E-24-108 wire, E-Type, insulation diameter = 0.02 in, wire length of 108 in
Table B.1: Specifications of the applied thermocouples [www.omega.com]
To withstand the chemical and thermal exposure in the heat transfer oil a special type of PFA Teflon® thermocouple insulation was selected making it possible to expose the thermocouples to 260ºC heat transfer fluid for up to 20 000 h.
A method for attachment was found in a special high temperature cement that can resist maximum surface temperatures of 871°C. The material properties of the cement are shown in Table B.2. It combines a low thermal conductivity to minimize temperature interference from the surrounding heat transfer oil as well as a high thermal expansion coefficient to reduce stresses between the die surface and the cement. Moreover, the cement is capable of resisting the chemical abrasion that is induced by the heat transfer oil.
221 Description Value Maximum service temperature [ºC/ºF] 871/1600 Thermal conductivity [W/m⋅K] ~ 9 Thermal expansion [µin/in⋅ºF] 12.4
Table B.2: Material properties of high temperature cement “Omega-bond 700” [www.omega.com]
In order to assure a direct contact of the thermocouple junction with the die surface, the surface was cleaned and a small piece of high temperature tape was used to attach the thermocouple on the die surface. The cement then covered the tape and sealed the thermocouple junction (Figure B.1).
Attaching the Cement covers thermocouple junction the thermocouple with high temperature tape junction and tape
Thermo- couple wire Clean and polished die surface
Figure B.1: Process of attachment for the thermocouple junction on the die surface
222 Data Acquisition System
Since a number of 26 thermocouples need to be processed simultaneously, an appropriate Data Acquisition System (DAS) is necessary. The ERC owns a sophisticated
DAS chassis from “National Instruments“ that is capable of holding a variety of 4 “Signal
Conditioning eXtensions for Instrumentation” (SCXI) modules. The various analog input modules can multiplex, amplify, filter, and isolate voltage and current signals. For the particular temperature measurements a thermocouple module with several analog input channels is required. The ERC owns one SCXI 1112 8-Channel thermocouple module, which can multiplex 8 analog input channels to a single output that drives a single DAS device channel.
However, with only 8 input channels available, the 26 thermocouple input signals must be switched manually, making a simultaneous and real time temperature monitoring impossible. Therefore, an additional 8-Channel thermocouple module was purchased, making it now possible of multiplexing two modules with 8 inputs and obtaining 16 thermocouple signals simultaneously.
In order to configure the modules and to specify the input channels, “National
Instruments” provides the “Measurement and Automation Explorer (MAX)”.
As seen in Figure B.2, it can be selected to acquire a regular voltage signal or to assign a specific thermocouple for a particular channel.
223
Figure B.2: National Instrument’s “Channel Wizard” used to select the thermocouple signal
When setting an E-Type thermocouple the DAS converts the voltage signal automatically into temperature and performs the necessary thermocouple calibration internally. Tests revealed however, that it is important to warm up the DAS for at least 20 minutes before obtaining measurements. If this was not carried out, the temperature test results fluctuated strongly.
The software LabView® is used to acquire the measurement data during experiments. Figure B.3 gives an overview about the LabView® environment. With specifying all channels in the channel box it is possible to read the data of 16 thermocouples synchronously. A scan rate of 1 measurement every 10 seconds is chosen, since rapid temperature changes are not expected. This data series is then saved as a txt-
224 file. When the data file is opened in EXCEL, the signals follow the order in which they were specified in the channel box. In other words, column one will list thermocouple one data and column two thermocouple two
Channel Box
Figure B.3: LabView® environment to acquire measurement data
225
APPENDIX C
HEATING MEDIA EVALUATION FOR WARM TUBE HYDROFORMING PROCESS
C.1 Objectives
The objective of this study is to evaluate and select the heat transfer fluid that satisfies the desirable characteristics as stated below.
Desirable characteristics of the heat transfer fluid are listed below:
1) High flash point, fire point and auto ignition temperature.
2) Non-toxic, non-hazardous and satisfy OSHA requirements.
3) High oxidation temperature (temperature at which smoke formation begins due to
oxidation) when heated in open system.
4) Capable of being repeatedly heated to 300 °C after cooling (so as to check the
thermal degradation due to oxidation).
5) High convection ability (so as to minimize non-uniformity of temperature within
the fluid due to natural convection). 226 C.2 Oxidation temperature determination
The candidate heat transfer fluids that were tested are as follows: a) Dynalene™
600, b) Calflo™ HTF, and c) Dow Corning 550. The experimental set-up in Figure C.1 was used to determine the temperature at which the fluid exhibits a high degree of oxidation and starts to smoke. Samples of heat transfer fluid were placed in a 500 ml glass beaker. The samples were heated by means of an electrical heater coil powered using variable voltage transformer (output 10 A, 0-120 V or 10 A, 0-140 V). The temperature of the sample fluid was measured and recorded using a thermocouple at regular intervals of time. The minimum threshold temperature at which the fluid gives off smoke due to oxidation was noted. Figure C.2 shows the photograph of the experimental setup whereas Figure C.3 shows a close-up picture of the fluid and heating coil placed in the beaker. Table C.1 provides the summary and observations obtained from the experimental results. Figure C.4, Figure C.5, and Figure C.6 show the temperature vs. time curves and observation during heating of Calflo-HTF, Dynalene™ 600, and Dow
Corning 550, respectively.
227
Figure C. 7: Schematic of experimental set-up to test the fluid
228
Figure C.8: Photograph of the experimental set-up
229
Figure C.9: Photograph of the fluid and heating coil in the beaker
230 Fluid Sample Input Threshold Other observations volume voltage to Smoking (ml) heating coil temperature (deg C) Calflo™ HTF 500 10 A, 24 V 170 • Started smoking (transparent, significantly at 220 light yellow in C. color) • Volume increased significantly (from 500 ml to 600 ml) due to thermal expansion. Dynalene™ 500 10 A, 48 V 210 • Started smoking 600 significantly at 270 (dark maroon C. in color) • Black colored smoke. • Volume increased (from 500 ml to 600 ml) due to thermal expansion Dow Corning 500 10 A, 56 V 215 • Started smoking 550 heavily at 250 C. (transparent) • White colored smoke. • Volume significantly (from 500 ml to 600 ml) increased due to thermal expansion. • Fluid becomes dirty white after being allowed to cool.
Table C. 2: Experimental results and observations
231
Figure C.10: Temperature vs. Time plot for Calflo™ HTF
232
Figure C.11: Temperature vs. Time plot for Dynalene™ 600
233
Figure C. 12: Temperature vs. Time plot for Dow Corning 550
234 C.3 Determine if the fluid can be repeatedly heated to 300 °C after being cooled
The above-mentioned experimental set-up was used to heat the fluid samples to
300 °C and then allowed to cool. Calflo-HTF was not tested in this experiment As seen in Table C.1, Calflo-HTF has the lowest smoking temperature among fluids tested.
Therefore, it was not necessary to this fluid. Only Dynalene™ 600 and Dow Corning 550 were reheated to the same temperature (300°C). According to the results, both the fluid could be heated to 300 °C within the amount of time used for the first heat after being cooled.
C.4 Summary and conclusions
Three different heat transfer fluids (Calflo™ HTF, Dynalene™ 600, and Dow
Corning 550) were tested to determine the temperature at which the fluid gives off significant amount of smoke (oxidation resistance). Calflo™ HTF started smoking at
170 °C, while the other two began to give off smoke at 210 °C. This indicates that
Dynalene™ 600 and Dow Corning 550 resist oxidation better than Calflo™ HTF at high temperatures and hence are more suitable for this project application. The smoke generated from Dynalene™ 600 is black in color, while the Corning 550 gives off a white smoke. Due to the significantly higher cost of the Dow Corning 550, Dynalene™ 600 was selected as the fluid used in this effort.
235
APPENDIX D
PROCESS SEQUENCE FOR SUBMERGED DESIGN CONCEPT
SEQ# 0: INITIAL STAGE
• The dies are opened and the axial punches are all the way out.
• The tooling has room temperature and no fluid circulates through the dies.
• The tank is empty and no tube is inserted.
236 SEQ# 1: HEATING THE DIES
. . p p m m e e t t Die Die
t
• Same conditions as in Seq# 0.
• Heating fluid circulates through the internal die heating channels.
• The fluid temperature at the MOKON heating unit will be set in steps of 50ºC,
100ºC, 150ºC, 200ºC, 250ºC in order not the shock the dies and the fluid. The
fluid will be circulated until a steady stage tooling temperature for each stage is
reached.
237 SEQ# 2: FILLING THE TANK
Starting Stage
• Same conditions as in Seq# 1.
• The fluid temperature set at the MOKON unit is 250ºC.
• Hot liquid of 250ºC will be pumped through the docking rod to fill the tank until
both dies are submerged.
• It will be waited until the fluid and the tooling reach a steady stage temperature.
• Hot liquid of 250 oC from MOKON unit will be pumped through the docking rod
to fill the tank (as seen in below figure) until both dies are submerged. Then the
filling will stop
238
Final State
SEQ# 3: HEATING THE TUBE
3.1 INSERTING THE TUBE
• The dies are open and the fluid level submerges the lower die only.
• The tube is inserted into the dies.
• The heating system continues working with a fluid temperature of 250ºC.
239 3.2 SUBMERGING THE TUBE
• The upper die moves down and the tooling closes.
• Upon closure of the upper die the fluid level in the tank rises and both, dies and
tube are completely submerged.
• The temperature set at the MOKON unit corresponds to the temperature that is
required to heat up the die surface to the designed temperature of app. 250°C.
3.3 FLUSHING THE TUBE
240 • The axial punches move into the die to seal the tube ends.
• Valves at the axial punches will be opened and heating fluid with the temperature
set at the MOKON heating unit circulates through the tube until the designed tube
temperature is reached.
• The dies are still heated by the fluid that is running constantly through the heating
channels.
SEQ# 4: FORMING PROCESS
• When the temperature of the tube reaches the designed temperature the valves at
the axial punches will be closed to stop the fluid circulation throughout the tube.
• The pressure intensifier as well as the axial punches are then activated based on
the input loading path (axial feed vs. internal pressure). The forming process starts
here.
241 SEQ# 5: PRESSURE RELEASE AND REMOVAL OF THE FINAL PART
• After the forming process is completed, the submerged axial punches will be
withdrawn, in order to release the pressure into the fluid bath.
• The upper die moves up and the fluid level decreases, making it possible to
remove the formed part manually.
• The pressure intensifier will be retracted and is filled externally with new pressure
medium.
242
APPENDIX E
SYSTEM FOR MEASURING THE BULGE HEIGHTS IN THE FORMING DIE
Potentiometer
Cantilever beam
Forming die
Figure E.1: Mechanism of measurement components used to measure the bulge height as a function of time
243
Figure E.2: Picture to show the fluid release channels
a) Initial state b) During deforming
Figure E.3: Schematic to demonstrate the measurement system used to measure the bulge height in the forming die
244