Our Research

Finite Element (FE) Based Modeling for Part, Die Design and Formability


  • Assist automotive part and die designers, and stampers in the use of aluminum.


  • Creation of FE models from part CAD data.
  • Simulation of part forming via FE models.
  • Part and die design optimization through iterative simulation of component forming.

Results of Stamping Process FE Simulation Obtained in Dynaform

Finite Element Modeling and Simulation of Capstan Friction Test

A top half of the geometry was modeled to exploit the symmetry condition. 4 brick elements through thickness and 8 elements in width were used. The stress-strain data were selected for aluminum sheet and isotropic elastic-plastic model was used. Local axis systems were defined for each element in the strip model. These axes rotate with elements throughout the deformation. This allows the stresses in though-thickness direction to represent the contact pressure.

Investigation of contact pressure distribution during the Capstan friction test

To achieve more realistic results, the pre-bending steps in the capstan test were included in the simulation. The simulation begins with the unbent strip model. In Step 1), while fixing the inlet end, the outlet end was loosely bent around the pin by applying a circular motion to outlet end nodes. In Step 2), nodes at the outlet end moved with specified velocity until the strip reached a required tension. In Step 3), both inlet and outlet ends moved together with defined velocities. During step 3, the outlet end moved 22 mm and the inlet end moved 20 mm. Therefore the average sliding distance was 21 mm, which was 1.75 times to the quarter circumference of the pin.

The pressure distribution is quite non-uniform in both longitudinal and transverse directions. Pressure peaks are shown at the inlet and outlet regions and along the edges. The outlet peak is somewhat higher than the inlet. A simulation without friction showed the almost same pressure distribution as frictional case. Therefore, the pressure peak difference is not coming from friction, but rather it is related to bending and unbending forces.


After the transient period, contact pressure reaches the steady state condition. The real contact angle is less than the strip wrap angle, 90 degree. To check the contact condition in longitudinal direction, the penetrations of the nodes defining the pin through the strip surface were measured. The noticeable feature is that the strip edge deforms outward and there is no contact at edges. This matches with the experimental observation of non-contacted surface topography shown on the specimen after the capstan tests.


Bending and unbending behaviour of the strip during the Capstan test

Longitudinal stress is plotted in local axis system. The stress reaches the steady state condition after the transient condition. For the details of the Friction Test modeling and simulation and evaluation of Friction Coefficient estimation method see this paper.

Microstructurally Based Modeling of Formability for Automotive Aluminum Sheet

Finite element analysis models

Load-displacement curve (click above image for a movie)

Comparison between shear bands

Effect of inhomogeneous distribution

Layered inhomogeneity

Formation of 3D shear band

Effect of inhomogeneous geometry

Texture distribution and FE model (Assumed distribution of heterogeneous material regions based on measured distribution of texture component)

Shear band formation under plane stress assumption

Variation of the stress state for a point within the shear band

Correlation between strain hardening rate and instability

Modeling of trimming processes