Rapid Parametric Process Design Using FEM ...

0 downloads 0 Views 2MB Size Report
planning and process optimization and die design using sophisticated FEM packages. ... time” eliminating the cost- and time-expensive „trial and error” methods which ... This approach involves stamping evaluation using simulation at every.
Advanced Materials Research Vols. 6-8 (2005) pp 235-240 Online available since 2005/May/15 at www.scientific.net © (2005) Trans Tech Publications, Switzerland doi:10.4028/www.scientific.net/AMR.6-8.235

Rapid Parametric Process Design Using FEM Analysis M. Tisza1,a, Z. Lukács2,b and M. Tisza jr.3,c 1

University of Miskolc, 3515 Miskolc-Egyetemváros, Hungary

2

MTA ME MTT Research Group, 3515 Miskolc-Egyetemváros, Hungary 3

GFT Hungary Ltd., Budapest-Hungary

a

[email protected], [email protected], [email protected]

Keywords: computer aided process planning, FEM simulation, process optimization

Abstract. During the recent years, due to the rapid development in Finite Element modeling, as well as the rapid evolution of computer techniques, numerical modeling has become an important everyday tool not only for process analysis and process optimization but for die design and rapid die manufacturing, as well. The Department of Mechanical Technology at the University of Miskolc is regarded as the leading research centre for Computer Aided Engineering of sheet metal forming processes in Hungary. The activity of the department covers the computer aided technological and tool design, the development and application of knowledge based expert systems for process planning and process optimization and die design using sophisticated FEM packages. In this latter activity, the AutoForm Engineering GmbH is a strategic partner of the department for several years. In this paper, the concept for an integrated computer aided die design system specialized for the rapid design of parametric die faces using finite element simulation technique will be presented. Introduction Due to the global competition, today’s competitive market imposes continuously increasing demands on the products and product development. Styling, initial quality and reliable long-term design performance are important factors that are critical from the viewpoint of success in the market. Industrial companies are very keen on reducing development and manufacturing costs, as well as total lead time-to-market to ensure profitability. The challenge to successfully execute a product program is to develop and design products and manufacturing processes “just right the first time” eliminating the cost- and time-expensive „trial and error” methods which were very characteristic particularly in sheet metal forming for many years. In order to meet the above challenges, significant developments in sheet metal forming simulation were performed. This approach involves stamping evaluation using simulation at every stage of product and process development from conceptual design through product and process design up to the manufacturing phase leading to continuous improvement in the quality of part design, as well as to significant reduction in total cost and time in producing sheet metal parts. Simulation software is a strategic component of the above engineering approach. The use of simulation software in sheet metal forming has increased radically in the recent years, due to its benefits both in optimization and trouble-shooting. Traditional way of die-design in sheet metal forming using CAD systems In traditional die-design approach, most die surfaces are created in general purpose CAD systems, e.g. ProEngineer, I-DEAS, Solid Works, etc [1]. Though, it is much faster and more economical than the old, engineering design practice, when part and assembly drawings were made manually, but it is still very time-consuming and nearly impossible to fully integrate with computer aided process optimization to rapidly modify die design alternatives even in CAD systems using parametric design principles. However, designing die faces in a conventional CAD system is still very time-consuming: even for small parts, several days are required, and for a complicated sheet All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of TTP, www.ttp.net. (ID: 193.6.9.66-27/08/12,11:40:18)

236

Sheet Metal 2005

part – as typical automotive panels – the average time required to complete draw-die development, ranges between several weeks. A further disadvantage of this approach that die designer must wait until the die surfaces are completed – several days or weeks – before carrying out any real test to check the design for cracks, wrinkles and other stamping criteria. Simulation and knowledge based design procedures At the Department of Mechanical Engineering (University of Miskolc) several CAD/CAM packages were elaborated to support technological process planning and tool design tasks in sheet metal forming processes [2, 3]. In these packages, first, knowledge based engineering approaches were used to generate alternative process plans for various sheet forming operations (e.g. blanking, piercing, bending, deep-drawing, flanging, etc.). These knowledge based expert systems have modular structure including geometric description, technological process planning and die-design modules. The geometric description module is capable to create 3D part models and also to import them from various CAD systems. The elaborated program packages apply so-called two-step process planning strategy. First, using a special feature recognition technique, the system is capable to generate automatically a so-called geometrical process plan just taking the geometry of the part into consideration. Then, starting out of the geometric process plan and applying technological-formability rules, the system is capable to elaborate technologically feasible process plans. This second step is Step_2: mainly based on simple plasticity theory and Step_1: Initial + Redrawing Reverse redrawing empirical rules. Following the computer aided technological process planning, the die-assembly and part drawings can be created in a general purpose CAD system on the basis of the strip-layout design. Though, it was a big step forward compared to the so-called “conventional CAD-based approach”, any modification in the process planning requires to re-run the die-design module, too. In further developed versions of these systems Step_4: Flanging [4], numerical integration was integrated into the Step_3: Piercing + Trimming process sequence planning of these packages. In this latter case, the feasibility of the elaborated forming processes was checked by finite element simulation, and a process optimization was performed before the die-design stage, thus saving the time of unnecessary modifications in diedesign alternatives. Since at this process planning stage, practically no sufficient information on tooling available, therefore, numerical simulation Step_5: Cutting Step_6: Piercing at this stage can only be used to assess the (Windowing) of small holes feasibility of process plans. One-step codes are ideally suited for this feasibility studies: they can Fig. 1. Simulation results of the complete predict potential cracks, wrinkles, they can give technological process plan for information on strain-, stress- and thickness a car audio loudspeaker distribution, providing valuable support for process planning engineers to make decisions how

Advanced Materials Research Vols. 6-8

237

to change process parameters to avoid major faults [5]. As an industrial example for using this system, the simulation results of the complete process chain of a car audio loudspeaker frame are shown in Fig. 1. As we could see from the former example, too, this type of integrated solution can provide valuable information on the process planning of the total forming chain using a one-step forming simulation. On the basis of the process sequence shown in Fig. 1. the tool design can be performed in the tool design module with quite good efficiency and sufficient accuracy. However, for more reliable process analysis, even more detailed tooling information is required, which is still not available at this stage. Therefore, it may happen that after the tool design and performing a complete – incremental process simulation – further modifications are needed which requires to repeat the process planning and tool design stage again. Parametric die-face design As a result of the above mentioned drawbacks, software developers together with several automotive companies started to develop a faster, more efficient and less costly method for die face design. Some simulation programs now include integrated die-design tools that provide more efficient, fast and more reliable solutions. This concept is involved in the development of the AutoForm DieDesigner package developed specifically for die designers and tool makers to support rapid, parametric die faces within the simulation software itself. AutoForm-DieDesigner is the result of a development effort led by AutoForm Engineering GmbH in cooperation with tooling departments of leading car manufacturers with regular technical feedback. The Department of Mechanical Technology at the University of Miskolc is co-operating with the AutoForm Engineering GmbH, Deutschland in bilateral projects for several years. The software reduces tooling development time through rapid parametric design of die faces, and their immediate verification and optimization with integrated stamping simulation modules. It is specialized for generating binder and addendum surfaces, and for evaluating the feasibility of drawdie developments and prototype tools. The most important innovations in this parametric die-face design are its fully parametric features, the complete integration with virtual try-out software modules, and an automatic optimizer. The parametric approach of the software is based on analytical engineering principles. It also conforms to the die engineering practice of using surface profiles to design die faces and is compatible with various CAD data formats. This new concept makes possible to create complete die-faces in hours starting from the CAD surface data of the part [6]. Since the generated die surfaces are parametric, subsequent modifications can be implemented very quickly. Binder and addendum surface generation is regarded as one of the most critical part concerning the successful simulation and try-out. The main objective of the binder and addendum surfaces is to assure a safe forming process with sufficient stretching, too. For a complicated part – as most of the car body panels – to create a smooth and efficient binder and addendum surface is a very time consuming process and in conventional CAD systems requires good practical experiences from the tool design engineer. This new, integrated solution was developed to overcome these problems: due to its highly specialized parameterization, it makes possible to generate complex binder and addendum surfaces very quickly. In this integrated solution, the first step is the determination of the binder surface. For complex parts, it is a curved, smooth surface determined from the part CAD data using geometrical rules and can be easily modified by the user. The second step is the addendum surface generation connecting the part boundary with the binder surface. For the parameterization of the addendum surface complex profiles based on arcs and lines are used. One profile is fully described by a set of parameters including arcs, lengths, radii and angles as shown in Fig. 2. The profiles and the transition between the part and the profile, as well as

238

Sheet Metal 2005

between the binder surface and the profile are tangentially continuous. Since part designs are often subject to revisions and improvements, the software allows easy part design modifications. Any profile can be changed, new profiles can be inserted, or existing profiles may be deleted. After the parameterized binder and addendum surfaces have been created for a specific part, it can be replaced with another (similar) modified design and the software will automatically adjust the binder and addendum accordingly. It results in significant timesaving, since it allows the original addendum and binder geometry that is created, to be re-used for future design revisions of the part. Applying this methodology, it is possible to design die surfaces an Fig. 2. Input screen for addendum generation of a order of magnitude faster comparing complex profile with the parameter set to conventional CAD systems. An addendum surface is shown for a car body panel in Fig. 3. This surface has been generated automatically by the Die-designer based on default profile parameters. As a result of the integrated system approach and parametric linking, die face designs can be immediately evaluated with one-step or incremental simulation modules. The incremental module is used for high accuracy and virtual tryouts of the complete stamping process. Results include predictions of wrinkles, cracks, skid and impact lines, surface quality, etc. Incremental try-out results are typically available within some hours, Part depending on the size and complexity of the part and the desired accuracy. Due to the parametric approach, additional try-outs can be performed very fast. For example, the user can make modifications to the addendum geometry, and the various tool geometries required for the next simulations are automatically updated. Similarly, if the user modifies the punch opening line, drawbead positions are automatically adjusted. To further improve the die face designs, the Addendum Binder program includes an integrated optimizer module. The determination of even a limited number of Fig. 3. Automatically generated addendum optimal process or tool parameters requires a surface for a car body panel large number of simulations. It takes a lot of time, therefore, it is very important to have the capability to run automatic optimization. In the optimizer module, the user should specify certain parameters, e.g. the allowable ranges for tool geometry and stamping parameters. Tool geometry parameters include part, die and drawbar radii, drawbar height,

Advanced Materials Research Vols. 6-8

239

wall angles, etc. Stamping parameters include binder forces, drawbead strength, blank outline, etc. The optimizer then automatically determines the "best" design within these ranges through multiple simulations, to achieve the target function: for example, elimination of cracks and wrinkles, uniform surface stretching, no excessive thinning. After selecting the optimum solution, it is possible to make modifications in the die designer to the original die concept to eliminate wrinkles and cracks, for example by changing drawbar and wall heights, radii, etc. Furthermore, results such as wrinkling, splitting, skid mark progression, etc. can be visualized at each step of the drawing process. In Fig. 4., an example for process parameter optimization is shown for an automotive component analyzing the effect of blankholder force changes.

Fig. 4. An example for process parameter optimization Using the above described die design philosophy and optimization technique, significant benefits can be achieved: binder and addendum development time can be radically reduced and thus the total die development time as well, leading to an enormous reduction in die try-out times and overall tooling costs. Besides these quantitative benefits, improved part quality and stamping reliability is also expected due to the optimization of the design. Automatic optimization can help the die designer to find the best parameters for his design. It can also help the process engineer to determine the best stamping conditions. Summary In this paper, some recent developments in computer aided die designer systems were overlooked starting from the conventional CAD design of dies through knowledge based solutions up to specialized parametric die face design. It may be stated, that the concept of rapid, parametric die face design results in significant benefits and it is a big step forward in computer aided die design solutions. This concept for sheet metal parts conforms to the die engineering practice. The unique advantage of this approach is its full integration and parametric linking with numerical simulation modules making possible further design modifications (die engineering

240

Sheet Metal 2005

changes) to the part in very short time. Due to its parametric approach tooling is updated automatically for additional try-outs. As a result, several different draw die developments and tooling concepts can be evaluated very quickly, drastically reducing tooling development time and the overall process and tool development costs. As a results significant reduction of die development, tooling cost and lead-time can be achieved providing more economical production in sheet metal forming. References [1] Zeid, I. CAD/CAM: Theroy and Practice, McGraw-Hill Co., New York, 1991. pp.1-1052. [2] Tisza, M.: A CAD/CAM system for deep-drawing processes, Advanced Technology of Plasticity, 1987. v.1. p. 145-154. Springer, Stuttgart. [3] Romvári, P.-Tisza, M.-Rácz, P.: A complete CAD/CAM package for Sheet Metal Forming, 26th MTDR Conference, Manchester, 1986. pp. 33-39. [4] Tisza, M.: Integration of Numerical Modelling and Knowledge Based Systems in Metal Forming, 6th ICTP Conference, Nürnberg, 19-24. September 1999. p.117-128 [5] Kubli, W.: Prozessoptimierte implizite FEM-Formulierung für die Umformsimulation, ETH Zürich, VDI Verlag, Reihe 20. No. 204. 1996. [6] Schenk, O.-Hillmann, M.: Optimal design of metal forming die surfaces with evolution strategies, Journal of Computers & Structures, 2004. v.82. p.1695-1705. [7] Tisza, M.: Some Recent Developments in Modeling and Simulation in Manufacturing Technologies, Production Processes and Systems, 2002. v. 1. No.1. p. 99-106 [8] AutoForm Engineering Inc. AutoForm Reference Manual, http:www.autoform.com

Sheet Metal 2005 10.4028/www.scientific.net/AMR.6-8

Rapid Parametric Process Design Using FEM Analysis 10.4028/www.scientific.net/AMR.6-8.235 DOI References [6] Schenk, O.-Hillmann, M.: Optimal design of metal forming die surfaces with evolution strategies, Journal of Computers & Structures, 2004. v.82. p.1695-1705. doi:10.1016/S0045-7949(04)00109-9