1.1 Stamping process
Stamping refers to a variety of sheet metal forming processes, e.g. blanking, emboss-ing, bendemboss-ing, anging and coining [1, p. 393]. In the context of this thesis, stamping refers to a bending-dominated cold forming process where a blank is formed into a specic shape using the tooling: a punch and a die. Stamping is usually performed on sheet metal pieces especially in the automotive industry, see e.g. [2]. The term sheet metal refers to pieces that are less than 6 millimeters in thickness, a thicker piece is considered plate [1, p. 320]. This thesis is focused on a bending-dominated stamping process of a steel plate with a thickness of 30 mm. The plate to be formed is referred to as blank throughout the majority of this thesis. The goal of this thesis is to study the theoretical background of metal cold forming and to use the theo-retical knowledge to study this specic forming process and compare the solution methods and modelling considerations.
Illustrative gure demonstrating the stamping process in the context of this thesis is presented in gure 1.1. On the left side of the gure the undeformed blank is placed on top of the die with the punch above both before the forming process. On the right side of the gure the punch has been moved down and the blank has been formed by the punch by forcing it into the die cavity.
Figure 1.1: Stamping process
This problem is dominated by bending deformation. Sheet metal forming
pro-cesses usually involve stretching and bending deformation of the blank. However, the stretching part is not easily applicable to this forming process as it would require extremely high forces to stretch a plate 30 mm thick. The stretching deformation of the blank during the forming process is usually applied by blank holders in a deep drawing process, see e.g. [2]. No blank holders are present in this thesis.
If the geometry of the die cavity would present the desired shape of the nal product, problems may be introduced because of the elastic properties of the mate-rial. After the blank is formed and the tooling is removed, the stresses present in the blank, caused by the tools forming it, will cause the blank to deform from the desired geometry to a state in which the internal stresses are in static equilibrium.
This undesired deformation caused by the relaxation of the stresses after the removal of tools is called the springback eect. The springback has to be accounted for in the geometry design of the tooling and process parameter selection to optimise the shape of the nal product.
1.2 Stamping simulation
A stamping process can be simulated using nite element analysis software to cal-culate the deformation of the blank and to study the eect of changing forming parameters. This is called stamping simulation. It provides an economic alterna-tive for optimising the stamping process without the expense of manufacturing an actual physical tool. The nite element model used in the simulation has to be ac-curate in describing the actual physical phenomena involved in a stamping process for obtaining reliable simulation results.
The stamping process involves geometrically large and materially plastic defor-mations as well as discontinuous contacts between the tools and the work piece.
Therefore, the problem involves high nonlinearity and it has to be solved by using a nonlinear solution method. The solution for the nite element analysis problem has to be obtained incrementally in a large number of time steps. This introduces signicant amount of computational eort into the simulation. The methods for obtaining the simulation results are discussed in the theory part of this thesis.
High tooling velocities cause eects that are more dicult to control in the form-ing process. Such eects are, for instance, dynamic impact forces and rate-dependent plasticity. Therefore, the stamping process is usually performed at low enough tool velocities so that the dynamic and inertial eects are neglible. These kind of low velocity processes should be simulated as a quasistatic process, in which the velocity and acceleration terms are not of importance. The simulation can then be performed with a truly static solution procedure although the dynamic solution methods may provide some advantages in the simulation as will be discussed in this thesis.
To minimize unwanted surface wearage and surface traction, lubrication between
the tooling and the blank will have to be used. The lubrication also transfers heat caused by friction and plastic dissipation away from the blank. For quasistatic processes, the heat formed in a stamping process is usually assumed to not have a signicant eect on the material properties of the tooling and the blank. For this reason, this thesis does not discuss any thermomechanical eects on the simulation.
A small exception to this is the study on the preheating of the blank, where the temperature eects are only included in the material parameters, to see the eects on the forming parameters.
1.3 Notes on the thesis structure
This thesis was performed for the purpose of studying nite element simulation of a specic thick steel plate stamping process. The study is performed by trying out dierent modelling considerations and solution procedures and comparing their eciency and accuracy. The thesis can roughly be divided into two parts, the theory part and the practical simulation part.
The chapters 2 to 4, after this introduction chapter, is a literature study on the governing theory of metal cold forming nite element simulation. This includes the nonlinearities involved in the nite element model, to which metal plasticity and contact modelling are devoted their own chapters, and the nonlinear solution methods including the explicit and implicit procedures. It also involves a short discussion on the choice of the elements.
The 5th chapter introduces the simulation model for the practical simulation part which is performed with the Abaqus FEA (nite element analysis) software suite.
The implicit Abaqus/Standard code and the explicit Abaqus/Explicit code require dierent modelling considerations and both of them will be discussed in this chapter.
The preprocessing of the model and postprocessing of the results were performed with Abaqus/CAE (CAE = Complete Abaqus Environment) version 6.10-1.
The 6th chapter follows the study on the actual simulation part. The results of the simulation performed with the model introduced in chapter 5 are presented.
Some adjustments that the simulation results suggested are applied to the simulation model and a parametrical study on the material properties at dierent temperatures is performed.
The results will be further analysed in chapter 7. This includes the comparison of the eciency and accuracy of the solution methods and the simplied models. Some of the initial simulation results with dierent material model will also be compared to the nal simulation results.
Chapter 8 includes the conclusions based on the chapters 6 and 7 and presents the true outcome of this thesis in a summary form.