Hello everyone,
I am currently working on an indentation and scratchin simulation and everything works very well for a depth-controlled simulation (e.g. the indenter is pushed 100 µm deep into the sample, simply by a BC). I am working with Abaqus 2019.
However, when I want to transfer this model into a load-controlled state, I get major problems. But since indentation and scratch-testing in reality is load-controlled, I want to have it that way.
One of the things I came across is that Abaqus tells me that I would have to constrain the indenter in every degree of freedom, or assign a mass to it. The first suggestion can't be the solution as the indenter has to move, obviously. But if I assign a mass point mass to the RF of the Indenter, I have to guess the mass, because the real mass of a diamond indenter-tip can't be the correct mass here, as it would be only a few grams. That is not the only Problem I came across. Somehow my contact definitions seem to fail and the indenter intersects with the samples surface.
It seems that all tutorials and examples on the internet are depth-controlled and the load-controlled variant is not discussed at all. I found papers where a load-controlled simulation is used, but they dont go into details about their general model design.
So here my question:
Has one of you done a load-controlled simulation of indentation, scratching or a similar process and/ or could share some insight or .inp-file with me?
Hello Janis Kimm,
I would suggest to constrain the indenter in each degree of freedom except normal translation and add very soft spring element (point to ground) in this DOF.
Hello Janis,
One way of load controlled simulation is to model the indenter as a rigid body (since diamond is very hard) and constrain it at all DOF except normal direction. You can also constrain the tip of the indenter with sample at the initial point of contact to avoid any overlapping.
Hi Kimm,
If your problem is static or quasi-static, the mass of the indentor is not important, because there is no dynamic effect and zero kinematic energy. Therefore assign an approximate value for it's mass.
About the contact problem, if mesh refinement is enough, a little intersection between instances is acceptable.
- Badges
- Science topic