04 September 2022 0 9K Report

Hi guys.

I am currently working on writing VUMAT subroutine to develop new damage model.

To check if it works well, I ran a simulation with a simple model. Detailed model conditions are as follows:

- dynamic/explicit solver

- single element (refer to the image '1. single')

- C3D8R-type element (10mm*10mm*10mm)

- uniaxial tension boundary condition (lower nodes are fixed, and upper nodes are assigned tensile velocity) (refer to the image '3. boundary condition')

- boundary condition was assigned to the upper/lower surface(geometry), not vertices or edges.

Single-element model resulted in a SUCCESSFUL analysis; the job was completed and the output physically made sense.

Then, I re-ran the identical simulation with multiple elements (refer to image '2. multiple'). All model conditions were the exactly same but the element number. This time each element has the 2mm*2mm*2mm size, and therefore the total number of elements is 5*5*5=125.

However, simulations with this multiple-element model kept aborted. Almost right after I submitted the job, the job is immediately aborted (refer to the image '4. aborted')

I checked the code hundreds times but it does not contain any terms which is divided by zero.

I literally have no idea why this happens. Why does this error happen only when multiple-element models are used?

Please share your thoughts and experiences.

Thank you in advance.

More Youngseo Lee's questions See All
Similar questions and discussions