Integrating Process Simulation and Performance Simulation

For some time now manufacturing process simulation has been a fairly routine activity.  Casting, forging, injection moulding and stamping simulations are all commonplace, especially at what you could call the top end of the industrial market. But what doesn’t seem to be commonplace is coupling these to structural simulation packages; at least from where I’ve been sitting for the last year or two.

Quite simply we need to simulate the performance of the components we actually end up with, rather than those we’ve optimistically designed based on geometry and matweb style material definitions.  I often try to visualize what my car might look like if I could see the residual stress fields.  Or other factors such as material directionality and compromised properties like density and material thickness. Generally these are ignored when it comes to the structural simulation of the actual part. With a little additional effort we could map the production simulation results into the FEA models we are already using.

Material data which acknowledges non-linearity is still far from generally available.  And until real material data is generally available, realistic simulation remains beyond reach for most organizations.  But this is an issue we need to take head on if we’re going to achieve maximum performance in the next generation of our designs.

Better materials data from suppliers, the internet, or from properly funded test programs which form part of managed design and development processes. If these programs are run jointly with production process simulation, the only outcome can be a better handle on what a design will actually do

We recently dealt with mapping data to models using the mapped field function within CAE.  Given that most data we’ve got, and can get, is in the form of XYZ and a value or values, this function achieves almost all of what we want to do.  And as CAE will do all the interpolation/extrapolation of this data to the mesh, this is another part of the process that is essentially pre-packaged.  Similarly simulation results can be fed back into CAD to aid the shape definition – if a part is forged or pressed the shape simply won’t be the simple form designed in CAD – in most cases the material, to a greater or lesser extent, will go where it wants. So using the shape obtained from process simulation can enhance the quality of the CAD definition.

So what are the barriers to driving this process forwards? Not appreciating that materials, and how they arrived at being in the nice shape shown by the CAD model, is a critical part of the design process, has to be the biggest issue. Its an IP issue, and therefore a critical contributor to the value of any engineering company.  Yet many organisations simply ignore all this. Which is astounding given the relative costs of what they already do and what they need to do to push their product knowledge and design by analysis forward.