圓筒件沖壓成形工藝與模具設(shè)計(jì)【拉深?！?/h1>
圓筒件沖壓成形工藝與模具設(shè)計(jì)【拉深?！?拉深模,圓筒件沖壓成形工藝與模具設(shè)計(jì)【拉深?！?圓筒,沖壓,成形,工藝,模具設(shè)計(jì)
Annals of the CIRP Vol. 56/1/2007 -269- doi:10.1016/j.cirp.2007.05.062 Design of Hot Stamping Tools with Cooling System H. Hoffmann 1 (2), H. So 1 , H. Steinbeiss 1 1 Institute of Metal Forming and Casting, Technische Universitt Mnchen, Garching, Germany Abstract Hot stamping with high strength steel is becoming more popular in automotive industry. In hot stamping, blanks are hot formed and press hardened in a water-cooled tool to achieve high strength. Hence, design of the tool with necessary cooling significantly influences the final properties of the blank and the process time. In this paper a new method based on systematic optimization to design cooling ducts in tool is introduced. The optimization procedure was coupled with FE analysis and a specific evolutionary algorithm. Through this procedure each tool component was separately optimized. Subsequently, the hot stamping process was simulated both thermally and thermo-mechanically with the combination of optimized solutions. Keywords: Hot Stamping, Finite element method (FEM), Optimization 1 INTRODUCTION In recent years, weight reduction while maintaining safety standards has been strongly emphasized in the automotive industry for building new models. Hot stamping of high strength steels for automotive inner body panels offers the possibility of fuel saving by weight reduction and enhances passenger safety due to its higher strength. In order to achieve high strength by hot stamping with high strength steels, blanks should be heated above austenitic temperature and then cooled rapidly such that the martensitic transformation will occur. Normally, the tools are heated up to 200C without active cooling systems in serial production 1. However, in hot forming processes, the tool temperature must maintain below 200C to achieve high strength. So far, very few studies have been conducted regarding the design of cooling systems in a hot stamping tool. This paper presents a systematic method to design hot stamping tools with cooling systems in optimal and fast manners, in which the cooling system is optimized with the help of FE analysis and a specific evolutionary algorithm. Subsequently, a series of hot forming processes was simulated thermally as well as thermo-mechanically to observe the heat transfer and the cooling rate through the optimized cooling system. In the hot stamping process the tool motion requires relatively short time compared to the whole process time. Therefore, thermal analysis of a series of hot stamping processes without considering the tool motion could be sufficient with reasonable accuracy and shorter computation time for quick design of the hot stamping tools with cooling system. However, thermo- mechanical analyses that include the motion of the punch and the forming process are necessary to enhance the accuracy of the predictions. In this paper, a crash relevant hot stamped component of a vehicle and its corresponding prototype of hot stamping tool are introduced in chapter 2. And the optimization procedure with FE analysis and evolutionary algorithm is introduced in chapter 3. Subsequently, the results of thermal and thermo-mechanical analyses with the optimized hot stamping tool are presented. 2 COOLING OF HOT STAMPING TOOL 2.1 Motivation To enhance the economical production procedure and good characteristics of the formed parts, hot stamping tools need to be designed optimally. Therefore, the main objective of this study is the optimal designing of an economical cooling system in hot stamping tools to obtain efficient cooling rate in the tool. So far, very few researches have been conducted regarding the design of cooling systems in hot stamping tools. Therefore, an advanced design method is required. Also, an adequate simulation model is required to perform the optimization and investigation of tools and products as fast and accurate as possible. 2.2 Characteristics of 22MnB5 In direct hot forming process, the quenchable boron- manganese alloyed steel 22MnB5 is commonly used. Also, 22MnB5 is one of the representative materials of ultra high strength steels. Therefore, in this study, aluminium pre-coated 22MnB5 sheet (Arcelors USIBOR) was considered as the blank material. The material 22MnB5 has a tensile strength of 600MPa approximately at the delivery state, and the tensile strength can be significantly increased by hot stamping process. Higher tensile strength is achieved in the hot stamping process by a rapid cooling at least at the rate of 27C/s 2. The initial sheet of 22MnB5 consisting of ferritic-perlitic microstructure must be austenitized before forming process in order to achieve a ductility of blank sheet. As the austenite cools very fast during quenching process martensite transformation will occur. This microstructure with martensite provides the hardened final product with a high tensile strength up to 1500 MPa. 2.3 Tool component and test part The components of the prototype hot stamping tool and its kinematics are shown in Figure 1 and the initial blank and the proposed test part in Figure 2. The initial blank has the dimension of 170mm x 430mm x 1.75mm and the draw depth of the proposed test part is 30mm. -270- faceplate counter punch blank holder punch faceplate table table blank distance bolts die barells plunger Figure 1: Schematic of a test hot stamping tool. Initial thickness: 1.75mm 4 3 0 m m1 7 0 m m 4 0 0 m m 1 0 0 m m Draw depth: 30mm Figure 2: Initial blank and drawn part. 2.4 Cooling systems in stamping tools The tool must be designed to cool efficiently in order to achieve maximum cooling rate and homogeneous temperature distribution of the hot stamped part. Hence, a cooling system needs to be integrated into the tools. The cooling system with cooling ducts near to the tool contour is currently well known as an efficient solution. However, the geometry of cooling ducts is restricted due to constraints in drilling and also the ducts should be placed as near as possible for efficient cooling but sufficiently away form the tool contour to avoid any deformation in the tool during the hot forming process. To guarantee good characteristics of the drawn part, the whole active parts of the tool (punch, die, blank holder and counter punch) need to be designed to cool sufficiently. 3 DESIGNING OF COOLING SYSTEMS 3.1 Optimization with Evolutionary Algorithm x s a boring position minimum distance between loaded contour and cooling duct (x) between unloaded contour and cooling duct (a) between cooling ducts (s) loaded contour unloaded contour coolant bore Constraints sealing plug input parameters of cooling system number of cooling channels and coolant bores diameter of cooling duct evaluation criteria cooling intensity and uniform cooling Optimization (Evolutionary Algorithm) 1 solution per given input separate optimization Solution Figure 3: Optimization procedure for each tool. The optimization procedure for design of a cooling system is presented in Figure 3. In this procedure, cooling channels can be optimized in each tool by a specific Evolutionary Algorithm (EA), which was developed at ISF (Institut fr Spannende Fertigung, Universitt Dortmund, Germany) for the optimization of injection molding tools and adapted for design of cooling systems in hot stamping tools 3,4. As constraints for optimization, the available sizes of connectors and plugs, the minimum wall thicknesses as well as the nonintersection of drill holes were considered. The admissible minimal distance between cooling duct and unloaded/loaded tool contour (a/x) and the minimal distance between cooling ducts (s) were determined through FE analyses. Parameters of the cooling system such as the number of channels (a chain of sequential drill holes), drill holes per channel and the diameter of the holes for each tool component were also provided as input parameters to the optimization. These input parameters can be obtained from existing design guidelines or through FE simulations. Based on the input parameters initial solution is generated randomly by EA or manually by the user. From the initial solution, the EA generates new solutions by recombination of current solutions and modifying them randomly. The defined constraints were subsequently used for the correction of the generated solutions and the elimination of inadmissible solutions. All the generated solutions were evaluated by optimum criteria such as efficient cooling rate and uniform cooling. Finally, the best solution was selected as optimized cooling channels for a selected tool component. 3.2 Optimized cooling channels In our research, the selected diameters of ducts were 8mm and 12mm for punch, 8mm, 12mm and 16mm for die, 8mm and 10mm for counter punch and 8mm for blank holder. EA was used to place the cooling channels optimally according to the given input and constraints for each tool component. The optimized profiles of the channels for duct diameter of 8mm are shown in Figure 4. c a b 4 0 0 m m 100mm 145 mm pu n c h cou n ter p un ch die b l an k h o ld er a b a b c a b 5 1 0 m m 260 mm a b c 70mm 510mm ab 260 mm a 110mm cooling medium plug 380mm a 70mm 250 mm b c b direction of cut view Figure 4: Optimized cooling channels with 8mm duct diameter. 4 EVALUATION OF THE OPTIMUM COOLING CHANNEL DESIGNS The design of cooling channels was generated by EA for each tool component with different bore diameters and their cooling performances were evaluated by using FE simulations. 4.1 Thermal analysis In the design and development phase of hot stamping tools, it is important to estimate the hot stamping process qualitatively and quantitatively within a short time for -271- economic manufacturing of tools. For this purpose, two transient thermal simulations were carried out with ABAQUS/standard, which uses an implicit method. In this analysis steel 1.2379 was selected as a tool material. The simulation model comprises 4 tool components: punch, die, blank holder and counter punch. In Table 1, the selected combinations of tool components with optimized cooling channels are presented. The variant V1 is the combination of optimized tools with small cooling duct diameters, whereas variant V2 with large cooling duct diameters. V1 V2 punch counter punch blank holder 8mm 8mm 8mm 8mm 12mm 10mm 16mm 8mm diameter of cooling duct die Table 1: Combinations of designed tools for FE analysis. In order to represent a series of production processes, a number of cycles of the hot stamping processes were simulated as a cycle heat transfer analysis. The Figure 5 shows the FE model including boundary conditions. cooling duct (c) T c = 20C h c = 4700W/m 2 C tool (t) T t,0 = 20C environment (e) T e = 20C h e = 3.6W/m 2 C counter punch blank holder punch blank die blank (b) T b,0 = 850C blank - tool D c = f (d,P) Figure 5: FE model and boundary conditions. This hot forming process for the prototype part was designed such that the cycle time is 30 sec. In a cycle, the punch movement for forming requires 3 sec, the tool is closed for 17 sec for quenching the blank and it takes another 10 sec for opening the tool and locating the next blank on the tool. However, in this thermal analysis, the tool motion and deformation of the blank was not considered to reduce the computation time. Hence, only heat transfer analysis was performed in a closed tool. In thermal analysis, the quenching process takes places 20 sec instead of 17 sec, because the motion of punch was not considered. It was assumed that the blank has an initial homogeneous temperature (T b,0 ) of 850C due to free cooling from 950C during the transfer in environment. The initial tool temperature (T t,0 ) was assumed as 20C at the first cycle and changes from cycle to cycle. The temperature of the cooling medium (T c ) was assumed as room temperature. Beside the boundary conditions, the required material properties of 22MnB5 were obtained from hot tensile test conducted at LFT (Lehrstuhl fr Fertigungstechnologie, Universitt Erlangen-Nrnberg, Germany), with whom a joint research on hot stamping is being conducted 2. In this analysis, convection from blank and tools to the environment (h e ), conduction within each tool, convection from tool into cooling channels (h c ) and heat transfer from hot blank to tool (D c ) were considered. Here, D c , is the contact heat transfer coefficient (CHTC) which describes the amount of heat flux from blank into tools. This coefficient usually depends on the gap d between tool and blank and the contact pressure P. It increases usually as the contact pressure increases. However, in thermal analysis the pressure dependent CHTC was not available, but a gap dependent coefficient was used. CHTC was assumed as 5000W/m 2 C at zero distance between blank and tool (gap) and keeps constant until the gap increases beyond critical value. 4.2 Thermo-mechanical analysis Simulation of hot forming is different from conventional sheet metal forming simulation, in which the distribution of temperatures or stresses in tools is neglected. For fast and easy way to analyze the hot forming process the tool and the blank were modelled with shell elements as in other studies 5,6. In these studies, the temperatures could be distributed along the thickness of the shell element with the user-defined function of temperature, but the temperature within the tool was not considered. Also, in this simulation model the heating of tools in a series of hot stamping processes were not considered. Furthermore, the shell model for thermal contact problems is just adequate for relatively short contact time 6. Therefore, in our studies the tools and the blank were modelled with volume elements to simulate the sequential heat transfer in a series of processes. The thermo- mechanical simulation was conducted with ABAQUS/explicit. In comparison to the thermal analysis, the whole forming and quenching process were modelled and the dynamic temperature and stress responses of tools in contact with hot blank were simulated by using time-temperature dependent flow stress curves. The heat transfer could be more accurately expressed using pressure dependent CHTC at contact places which change during forming process. In addition, temperature dependent thermal conductivity and specific heat were also considered. However, in thermo-mechanical analysis, as the number of elements increases, the complexity of the FE problem significantly increases. In conventional forming simulation an adaptive mesh can be normally used to spare the simulation time and to obtain a more accurate solution in the contact area. However, adaptive mesh refinement causes instability during computation in thermo- mechanical analysis. Therefore, a refined mesh with higher punch velocity was considered to reduce the simulation time. The heat transfer coefficients were scaled accordingly to obtain the same heat flux 7. 5 SIMULATION RESULTS AND DISCUSSION 5.1 Thermal analysis Figure 6 shows the temperature changes in the tool components for 10 cycles at tool combination V1 and V2. T C 400 300 100 0 030100 0 300100 die punch t s t s V1 V2 Figure 6: Temperature changes in heat transfer analysis. The results show that the hottest temperatures of the tools at the end of each cycle do not change almost after some cycles. The obtained cooling rates of the blank at the hottest point from 850C to 170C are 40C/s with V1 and 33C/s with V2 at 10th cycle and these are greater than the required minimum cooling rate of 27C/s. Furthermore, V1 leads to a more efficient cooling performance than V2. Better cooling performance for V1 compared to V2 can be explained with the geometric restrictions and the minimal wall thickness. A cooling duct with small diameter can be placed closer to the tool surface in a convex area and the amount of the cooling channels can be increased additionally. Usually, the heat dissipation in the convex area is slower than in concave area 6. The result shows also that the temperature of convex area in the punch -272- cools down slower than the concave areas in the die. Due to this fact, it can be concluded that the efficient cooling is most desired at convex area. 5.2 Thermo-mechanical analysis The heat transfer with optimized tool components was simulated thermally at first. However, there was a simplification of a hot stamping process in thermal analysis. Therefore, a thermo-mechanical analysis for V1 was performed to observe the differences and the significance of modelling the punch movement. Temperature change curves at the hottest point from the end of the first cycle in the blank, die and punch are shown in Figure 7. The tool cooled further 10 sec after quenching and the temperature changes in the die and punch were presented for 30 sec. A coupled thermo- mechanical analysis was done using gap-pressure dependent CHTC. The results from thermal analysis shows a cooling rate of 92C/s from 850C to 170C in comparison to 75C/s from thermo-mechanical analysis. 400 300 100 0 die punch 05 20 1000 800 400 T C 200 Thermal analysis Thermo-mechanical analysis t s 15 blank 0 0 5 30 0 5 25 30t s10 202510 20t s T C Figure 7: Temperature changes in thermal and thermo- mechanical analysis (1th cycle). To verify the accuracy of a thermal analysis or to predict a serial production process more accurately a series of thermo-mechanical analysis was done. For this analysis the punch velocity was increased 10 times and 10 hot stamping processes were simulated. In Figure 8, the temperature change curves at the hottest point of the die and punch from a thermal and thermo-mechanical analysis are compared for 10 cycles. Finally, the temperature distributions in the blank at the end of the 10th cycle are shown in Figure 9. 400 300 100 0 TC 030ts100 030ts100 die punch thermal thermo-mechanical Figure 8: Temperature changes for 10 cycles. (b) T C (a) 130 60 102 74 88 116 T C 140 70 112 84 98 126 Figure 9: Temperature fields of blanks at the end of 10th cycle: (a) thermal and (b) thermo-mechanical analysis. In Figure 8, the temperature differences at the end of 10th cycle between the thermal and thermo-mechanical analyses were 7C in the die and 3C in the punch. Subsequently, the Figure 9 indicates that the maximum temperature of the blank from the thermal analysis is slightly greater than that of the thermo-mechanical about 10C. Nonetheless, the temperature fields of blanks from both analyses are very similar. As a consequence, the thermal analysis for a series of hot stamping processes is relatively accurate compared to the thermo-mechanical analysis. Furthermore, a thermal heat transfer analysis could be used to design and develop the hot stamping tools in the early phase due to its timesaving computation. 6 CONCLUSION AND FUTURE WORK A systematic method has been developed for optimizing the geometrical design of the cooling systems of hot stamping tools. This methodology was successfully applied to design of cooling channels in a prototype tool for efficient cooling performance. This indicates that the method can be used for designing cooling systems in other stamping tools as well. This paper presented both thermal and thermo- mechanical simulations to represent a series of hot stamping processes. The thermal analysis could be used for an optimization and investigation of hot stamping processes especially in the developing stage. However, a thermo-mechanical analysis is needed to predict more accurately but it is still time consuming to analyze the processes within adequate time period. To resolve this problem, an alternative simulation model will be further studied. Also, a more accurate contact condition for thermo-mechanical analysis remains to be studied. To validate this proposed method and its corresponding FE model, a prototype tool is currently being built and experiments will be carried out for validation. 7 ACKNOWLEDGMENTS We extend our sincere thanks to all joint project researchers of LFT and ISF. 8 REFERENCES 1 Sik