Research and development of automobile crash simulation dummyRelease time: 2015-08-19 09:02
The modern collision simulation software package is composed of a pre-processing / post-processing program module and a numerical calculation module. The industrial use efficiency of the collision simulation software package depends not only on the programming efficiency of the calculation program, but also to a large extent on the quality and efficiency of the pre-processing / post-processing program modules connected to it and its users and synergies. The PAMGENERIS pre-processing program module and the PAM-VIEW post-processing program module currently provided both use the PAM-SOLID numerical calculation program. At present, both program modules are undergoing large-scale improvement and reconstruction to fully ensure their applicability and success in industrial use. The PAM-SOLID calculation program includes the PAM-CRASH car collision analysis program, the PAM-SAFE passenger safety analysis program, the PAM-STAM sheet stamping program, and the PAM-SHOCK impact and high-speed response analysis program. This article focuses on the PAM-CRASH automotive crash analysis program.
Industrial car collision simulation started 12 years ago. The PAM-CRASH calculation program was used on the CRAY1 computer for the first time to successfully perform a overnight overnight simulation of the front of the VW Polo. Since then, collision simulation has been used by all major car manufacturers to help design and manufacture competitive collision-resistant cars. Moreover, collision simulation has been extended to other industries, such as truck collisions, train collisions, container drops, packaging structures, and aircraft collisions.
The logical second step is to extend the simulation method to the simulation of passenger safety. The object of this kind of simulation is passive safety devices, such as airbags, safety belts and shock-absorbing cushions that can protect the passenger model (dummy) in the event of a collision.
The development of dummy test technology has led to the emergence of the fourth generation of dummy, and the fifth generation of dummy (biological dummy) is about to appear.
In addition, explicit finite element technology has also been used in related processing technologies, such as sheet metal forming, thermoplastic composite shell forming, pressure forming, and hydroforming. These processes can use a large number of thin shell finite elements (up to more than 100,000) to dynamically analyze and simulate large deformation / large displacement and nonlinear materials of the structure.
The remarkable success of explicit thin shell simulation technology is being further enhanced. The first is that computing programs are currently input into massively parallel computers, which provides huge computing power to design departments. The second is to enter the calculation program into a cheap microcomputer-level workstation, so that small manufacturers can also afford it.
The following developments have made the calculation program more effective and effective for the collision and impact simulation of many industrial problems: ① directly read CAD data for grid production. ② Macro beam element for conceptual design. ③ New material models, strain rate models and damage sensitive models such as foam materials, composite materials and plastics.
The next development of computer simulation includes the following main areas: ① Linking simulation software packages to CAD systems. ② Develop more material models, more accurate element equations and contact algorithms. ③ Develop larger computing models. ④ Multi-physical quantity model (thermal effect and solid-liquid interaction). ⑤ Continue to input computing programs to massively parallel computers and group workstations.
I. New progress of the algorithm Although the industrial efficiency and stability of modern explicit finite element collision simulation calculation programs have reached a considerable level, there is still much room for improvement and development of the algorithm. The work done in this regard is described below.
1.1 Contact simulation It should be pointed out that in collision simulation, correct and complete simulation of contact between objects is a prerequisite for its success. However, it is worth noting that, whether using a single-processor (or shared memory parallel) program or a large-scale distributed memory parallel program, contact simulation is always the most challenging and complex programming task. In the first case, in order to make the calculation both efficient and without losing its accuracy and reliability, the handling of the "rigid wall" limits of contact or the handling of "sliding contact surfaces" must be as reasonable as possible. For contact algorithms of the sliding contact surface type, penalty compensation methods have proven to be the most widely used. The current development trend is to strive to make this method as expressive as possible for end users, including the need to overcome difficulties encountered in meeting computing efficiency and accuracy requirements. A comparison of airbag contact treatments using the original and improved penalty-compensated contact algorithms.
For the preparation of large-scale parallel programs, it is easy to parallelize explicit finite element programs with effective automatic region partitioning technology. Each parallel processor needs only a small amount of communication with other processors to complete the work of a large number of similar models. Because the determined contact area is not static (the original disconnected parts are connected to each other when the structure is deformed), the original area division will soon be unsuitable. If the contact calculation is still performed as originally, it will lead to information transmission Large increase in volume and extremely uneven workload distribution. Therefore, the important work of large-scale parallel interface for collision simulation programs is the effective parallelization of contact algorithms.
1.2 Grid self-varying algorithm In the collision simulation of thin plate components and structures, the main goal of the grid self-varying algorithm technology is to automatically refine the local finite element mesh of the thin shell near the large deformation area. This problem has been solved in industrial sheet metal stamping simulation. The PAM-STAMP thin plate stamping simulation program uses a program that enables uniform and automatic selection of thin shell mesh refinement and coarsening.
The first reason that the grid self-varying algorithm is difficult to apply to collision simulation is that the adaptability of the grid must be compatible with a wide range of contact treatments and rigid wall contact / impact selection. Although contact is the predominant mechanism for sheet metal stamping, it is relatively simple to handle because of the tool geometry and the free-standing contact of the sheet metal. In collision simulation, self-contacts often appear in large numbers, and it is difficult to modify them to apply the grid self-varying algorithm.
The second reason that the mesh self-variant algorithm is not very suitable for collision simulation is even dangerous. It is due to the limitation of plastic corners or collision areas. In other words, in the actual calculation, the refined mesh area has less rigidity than the original coarse mesh, and the plastic corners and collision areas of the unrefined mesh may be increased in softness or reduced. It is covered by the detailed area for calculating the resistance. Therefore, in order to avoid this limitation, the refinement criteria must be carefully selected, and the original mesh should not be too coarse.
Examples of application of the grid self-varying algorithm include the axial impact of thin-walled box columns (Figure 2a), the study of large-deformation plastic corner zones in an S-shaped frame (Figure 2b), the front collision of a passenger car (Figure 2c), and metal Sheet metal stamping (Figure 2d).
Designing, calibrating, and validating various material models for the description of aluminum alloy, plastic, foam, rubber, and composite behavior has received increasing attention. The PAM-SOLID program includes models for describing the behavior of these materials.
Second, the following highlights some of them.
2.1 Elastoplastic / strain rate / damage model
The finite element method in the PAM-SOLID program uses a variety of isotropic and anisotropic elastoplastic materials and damaged material models containing strain rates. The general law of material stress-strain is shown in Figure 3. Figure 3a shows the response and damage curves. In principle, the law can be applied to a variety of materials. Basic undamaged stress-strain laws. Number reduces material strength. The general description of damage indicated here is due to Lemaitre-Chboche, which has been applied to any material, such as metals, plastics, composites, and foams. Currently, in cooperation with the University of Valenciennes, we have adopted the Gurson damage mode for ductile metals to describe the effects of plastic strain and nucleation, growth and coalescence of material microvoids on material behavior. The effect of voids or damage on stress in different scale models is shown in Figure 3b.
2.2 The use of Hill (1990) non-quadratic yield function for anisotropic materials can describe the plastic behavior of aluminum alloys, as shown in Figure 4. The PAM-STAMP program has adopted this model and has been used for deep drawing simulation of aluminum plates.
2.3 Plastics During the simulation of plane stress tensile tests on plastics, plastics generally show an elastoplastic behavior of softening and then hardening. First, the finite element model is used to simulate the softening phase on the stress-strain curve in the tensile test (Figure 5a). Once the plastic stress σ drops below the tangent modulus Et, plastic instability (necking) occurs. When the plastic strain reaches a certain fracture limit, this instability quickly leads to the concentration of the plastic strain and the fracture of the test piece. Secondly, the original plastic hardening curve is modified (Figure 5b) in order to strengthen the formation of a higher plastic strain force, thereby limiting the plane stress condition σ = Et of the necking within the range of ε1 <ε <ε2. The initial necking still develops as before, but now when the finite element of the necking region is stretched above the plastic strain ε2, the hardening of the material inhibits the necking. Then, the necking expands towards the surrounding finite element and gradually invades the entire test piece.
2.4 Other materials For the description of side impact obstacles, pads, and mechanical dummy skin foam, etc., PAM-SOLID software has a variety of models that can be used to simulate the rupture, compression, viscous deformation, and rate-dependent behavior of these materials.
There are also many models of superplastic and quasi-incompressible materials for descriptions of rubber materials and similar rubber materials such as tires, engine mounts, and dummy parts.
For the impact simulation of fiber reinforced materials, there are models of fragile and damaged materials. These models have also been extended to describe other composite materials.
Of course, all these material models need further improvement. New material models will continue to emerge to meet the growing needs of various industrial sectors to improve their specific products.
III. Conceptual collision design An interesting trend in collision simulation is that the industry is increasingly demanding simple collision simulation methods, especially in the early stages of transportation equipment design. At present, most existing crash simulation practices emphasize the final stage, that is, the use of a large number of finite element overall models (20,000 to 100,000 or more) to check the final crash performance of the designed structure. However, when designing the crashworthiness of a structure, the information and time required to make a detailed crash verification model may not be available. Therefore, the simplified method is more popular.
One example of a conceptual collision design is a box post. Currently, ESI Group is researching the graphical user interface required for this concept collision based on the PAM-SUPERFOLD software developed in cooperation with the University of Valenciennes. In this interface, the user can input the wall thickness and material properties of the box column. Using the dynamic instability software, the impact response of the component to the axial or deep camber instability is calculated. You can also use another PAM-OPT program currently under development to automatically change the set of input optimization parameters and limit variables to get the best response you need. Then, automatic finite element meshing can be performed on the parts with the determined thickness, and then input to the PAM-CRASH finite element software for calculation and verification. The subsequent collision design process can incorporate the response curve of the collision component in the previous stage into a specially designed 3D beam / nonlinear elastic or integral beam element model in the collision program. The subsequent stages form the transition from early crash design to structural nodes, whose components have actual geometry, cut-off points and connections. In this hybrid phase, the mesh of key areas can be refined, while the rest of the structure can only be roughly simulated. In the final stage, the most detailed model is used to verify the overall design.
4. In addition to the developed deformable collision obstacle (side collision and forward collision) models and various mechanical dummy models (composite III, European side collision type, American side collision type, biological side collision type, and child type) In addition, biomechanical models of the human body have also been developed. The work in this area will probably be further developed, because the simulation of the biological model of the human body is the only way to get a direct response to the collision event. ESI Group has developed models of this type (such as head and lower legs). The initial leg models include bone, ligament, tendon, and muscle models.
V. Conclusion The collision simulation technology and program can be successfully used not only in automobile collision simulation, but also in many other related fields. For example, a cushioning packaging material designed for the drop of a refrigerator, a drop test simulation of a home appliance (such as a small computer) to estimate damage to sensitive components and a collision simulation of a hard lander for space exploration. In addition, the reaction of tires when rolling over bumps, the impact response of sporting goods, and the impact response of bridges, cranes, rails, containers, trains, boats, trucks, tractors, and bicycles can be used. . In fact, the range of applications for crash and impact simulation software is growing.
More about automotive safety test equipment: http://airulm17.com/productlist/list-5-1.html
- What are the precautions for purchasing flame retardant fabrics?
- Car safety crash test dummy unemployed
- Human body 3D scanning technology attracts members into the gym
- Interpretation of the characteristics of the spine of a colliding dummy
- Black technology! 3D body scanner can be used for measuring body
- Apple vs. WeChat-Chinese Internet companies