Discussion Forum

This usually is due to a problematic geometry. If you can, edit your geometry to specify all the boundaries, inside and out, without any extremely thin pieces or nearly-tangent lines, etc. Consider posting your model (.mph file) to the forum if you want more specific help.

Thanks Dr Koslover, all I use is just a domain of a square. Is that a poor geometry?

Hm, a square gives you inverted elements?

at least as far as what the warning message says, though a glance of the deformed square doesn't show obvious inverted elements

I suggest you post your model (the .mph file) to the forum.

Most strangely, the warning suggest inverted elements near coordinate (1e5,1e5) while the domain size is only about 1*1.

I assume that this happens during solution, and not in the initial configuration.

It seems that you have an ill-posed problem where, for example, a solid mechanics model 'blows up'. Such problems are, for example, common with some hyperelastic models that are not unconditionally stable. For an even more common and unexpected case, see the section about linear elasticity under geometric nonlinearity in https://www.comsol.com/blogs/modeling-linear-elastic-materials-how-difficult-can-it-be/ .

I assume that this happens during solution, and not in the initial configuration.

It seems that you have an ill-posed problem where, for example, a solid mechanics model 'blows up'. Such problems are, for example, common with some hyperelastic models that are not unconditionally stable. For an even more common and unexpected case, see the section about linear elasticity under geometric nonlinearity in .

Thank you! However, though the warning suggests inverted elements near coordinate (1e5,1e5), the domain (1*1) after the computation looks normal in shape rather than sth that blows up and the whole computation is completed without error messages.