Flex to rigid contact modeling is giving very high peaks in forces

Question

I am modeling a flex to rigid contact. In this I have two queries. First is regarding IMPACT function, it works decently well when surfaces remain in contact. But where there is frequent attach-detach scenario, I am getting random and impractical forces.

Second question is regarding contact force recorded in result file: I downloaded a macro from simcompanion which basically collates all the nodes and then calculated damping force and stiffness force seperately using inputs and basic IMPACT function. The total force given by this macro does not match with the force shown in result file. So what is the validity of the force shown in result file.

Jesper Slattengren · Accepted Answer

Select the materials and the dimensions and contact type. For you, probably a line contact with the Length=width of the leaf spring, radius 1 rather large (50 mm?) and radius 2 infinite can come fairly close to leaf-leaf contact.

Enter a low force. I often start at 10N. Read the "Impression" field. That is the penetration. Note the unit (and the unit changes with load to whatever is suitable, between nm, micro-meter (um) and mm.

Type these values into an Excel sheet (penetration vs force).

Repeat this. I usually get good data by using 10, 20, 50, 100, 200, 500, 1000, 2000, 5000, 1000 N

Now you have a table in Excel.

Plot this data. Add a trendline and format the trendline as "Power" (k*x^e). Select to show the equation in the plot. Format the numbers in the equation to have enough significant figures to read k and exponent. Use those values.

But contact with flex bodies are always tricky to get smooth values from. From the FE mesh a Parasolid geometry is created that is tessellated and the contact is calculated with this surface. At each time, the Parasolid is adjusted for the deformation.

Have you changed the CONTACT_FACETING_TOLERANCE parameter in the preferences setting? If not, it is worth trying. Default is around 300 and you can increase this to get a "smoother" surface. But still discretized so if the contact slides from one facet to another, there will be a discontinuity in the forces. No reason to use a value over 3000.