Hello everyone
I turn to you to understand the wear model used in Liggghts. I know it is qualitative, but anyway I would like to have a numerical approximation to the wear that occurs.
My model shows values between 0 and 1000, which according to my understanding are measured in kg/m², which is a very big wear in a short time
My first question is if the area corresponds to a triangle of the mesh or something else.
I would also like to know if there is any special way to convert kg/m² to mm/year and finally I would like to know if any of you have worked with wear and how you have done.
I leave my script that corresponds to the wear of a canoe through which circulates concrete modeled with spheres the size of gravel, with which I get that the wear is 1 mm per second
I thank you very much for your time and availability to help
| Attachment | Size |
|---|---|
 Files wear model | 5.56 MB |
jsams | Tue, 02/13/2018 - 07:30
re: modelling wear in liggghts
The wear model implemented in liggghts is based on the Finnie wear model (I believe there is a section about it somewhere in the liggghts doc on github). If I remember correctly, the model considers the contact angle and particle velocities. The wear values you get out of the simulation are calculated per mesh element and depend very much on the wear coefficient that you define.
There exist other wear models too, e.g. Archard wear model which I believe is used for modelling abrasive wear. You could, for example, extract the relevant information (force, particle velocities at mesh contact) from your simulations to calculate the wear in order to get a quantitative wear result. I believe you should try to find out the appropriate wear model for your kind of simulations and try to apply it to your case. The challenge I see is that the models often require material specific wear coefficients (depending on the materials in contact), however, if you want to compare different setups/configurations it might not matter that much.