The Corus-Vegter Material Model - Validation case study with Renault
To demonstrate the increased accuracy of the Corus-Vegter material model compared to standard models, Renault and Corus chose the rear door inner of the Renault Modus as a test case. The validation trials were performed during the pilot production phase of this part in Valladolid, Spain.
Introduction
With ever increasing demands on reducing tool build, tryout lead times and improved panel quality, Renault are constantly seeking out better modelling methods for formability analysis. As an innovator in new technical services and ‘intelligent materials supplier’, Corus has been developing the Vegter materials model in an effort to improve the accuracy of forming simulation. In order to deploy this technology effectively, Corus has been working with software developers for several years to enable the benefits of this innovation to automotive customers through an industry-standard software package such as ESI's PAM-STAMP™. Renault and Corus collaborated to prove this new tool on a production tailgate part of the new Renault Modus (‘the small car with a big heart’), built in Vallidolid, Spain.
The Technical Challenge
In the automotive industry formability analyses using PAM-STAMP™ can help to reduce the cost and time of new vehicle development. It can also help to optimise manufacturing processes. Nevertheless, accurate analyses of the feasibility and processing of an automotive part can only be done when an accurate material model is used. It is known that for a wide selection of materials the standard plasticity models (e.g. Hill'48 and Hill'90) seldom coincide completely with measurements. The Corus-Vegter material model was developed to alleviate this. This model consists of an improved yield locus description and an improved strain hardening description. ESI have worked with Corus to implement the new model in PAM-STAMP™ as an option of AUTOSTAMP™.
Validation Studies
The Corus-Vegter model was first successfully validated on simple products like cup drawing, stretch forming and the other experimental tool shapes. To validate the model on a full scale complex automotive part, a joint project was set-up with Renault using the Modus tailgate part (see figure 1 below).
In order to be able to compare two material models with the actual current situation in the pressing of the part at Renault, Corus did not use CAD data to build the simulation. Instead, a laser scanner was used on-site at Renault, to obtain the actual current geometry of the press-tools used. For correlation to test during production of this part, the strains resulting from the first draw were measured using the PHAST ™ measurement system. The actual press forces, drawbead geometries, spacer locations, blank geometry and blank positioning were modelled to recreate the situation in the press as faithfully as possible.
Three simulations were performed, varying only the material model: one simulation used the standard Hill'48 model, one used the Hill'90 model and one used the Corus-Vegter model. The areas of the part that the PHAST™ system identified as the most critical, were also identified as critical areas using the Corus-Vegter model based simulation. On the other hand the Hill'48 model based simulations considered those areas less critical, while Hill'90 considered them as over critical.
As an example, the rupture risk is shown in (figure 1) below, for one of the most critical areas of this part. When the contour plots are compared, it can be clearly seen that the Hill'48 model underestimates the risk of rupture compared to the PHASTTM measurements. This means that simulations with the Hill'48 model are inclined to predict that a part is feasible whilst in reality it may split. The Hill'90 simulation predicted splitting in this area whilst in reality the part is being produced successfully. The Corus-Vegter model correctly predicted the risk of rupture of this area of the part. The results of the other critical areas showed similar differences. Furthermore, the thinning, strain distribution and draw-in that were predicted by the Corus-Vegter model were closer to the PHASTTM measurements than the predictions using the Hill'48 or Hill'90 model.
Conclusion
For the Renault Modus rear door inner case, the prediction of the rupture risk, thinning, strain distribution and draw-in for the Corus-Vegter material model were a lot closer to reality than the predictions of Hill'48 or Hill'90. This work provided a good basis for recommending that the cost and time needed to design and modify the tooling and checking the feasibility for automotive parts can be reduced significantly if the Corus-Vegter model is used.
