Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Stress derivative issues

Please login with a confirmed email address before reporting spam

I am solving a problem that relies on a gradient of hydrostatic stress.
I believe this is expressed as:

d(solid.pm,X)+d(solid.pm,Y)+d(solid.pm,Z).

I have checked the result of this expression against a program I created in MATLAB and the results do not agree (given the same hydrostatic stress profile). I am relatively certain that the profile COMSOL is calculating is incorrect, or at least is very different from what I would expect. I am not exactly sure where I am going wrong as it seems like a rather straightforward implementation. I am rather inexperienced with COMSOL, so any guidance would be appreciated.



1 Reply Last Post Jul 22, 2016, 1:29 p.m. EDT
Henrik Sönnerlind COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 8 years ago Jul 22, 2016, 1:29 p.m. EDT
Hi,

As stresses are proportional to strains, they are derivatives of the shape functions. When you try to differentiate them, you get second derivatives, which will have very low accuracy.

Suggestion 1: Use Cubic or Quartic shape functions instead of the default Quadratic.

Suggestion 2: Add an extra degree of freedom for the hydrostatic stress, which you in the weak sense map to solid.pm. In this way you can obtain something more smooth and differentiable.

By the way: A gradient is a vector, your expression is a scalar (the sum of the components of the gradient). It is more probable that you are interested in sqrt(d(solid.pm,X)^2+d(solid.pm,Y)^2+d(solid.pm,Z)^2)

Regards,
Henrik
Hi, As stresses are proportional to strains, they are derivatives of the shape functions. When you try to differentiate them, you get second derivatives, which will have very low accuracy. Suggestion 1: Use Cubic or Quartic shape functions instead of the default Quadratic. Suggestion 2: Add an extra degree of freedom for the hydrostatic stress, which you in the weak sense map to solid.pm. In this way you can obtain something more smooth and differentiable. By the way: A gradient is a vector, your expression is a scalar (the sum of the components of the gradient). It is more probable that you are interested in sqrt(d(solid.pm,X)^2+d(solid.pm,Y)^2+d(solid.pm,Z)^2) Regards, Henrik

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.