Discussion Forum

Hi

I'm not sure I catch really your question, but have you checked the doc/help under "Analysis of Deformation" and the related sub pages ?

I have also added a "toy model" to check that the spatial x,y,z and material X,Y,Z do link via the deformation u,v,w, but only if the geometric nonlinearity is "on" else they are equal.

This "toy-model" also illustrates the differences between 3D solid, 2D Solid in Plane Strain and Plain Stress, and how these compare to the mid plane of a 3D solid and to the outer edge. My material is artificial, just there to give some large deformations

To better understand check the equations that COMSOL uses (see the jpg below, I believe you must turn on the Equation Section via the *eye* icon in the model tree (by default COMSOL hides many features to make the model tree look simpler, that is best when you start with COMSOL otherwise one tend to get lost in the thousands of menus ;)

--
Good luck
Ivar

Hi,

thanks alot for your reply,

I red a lot about including or not geometric non-linearity, I just feel that I am still don't understand the physical meaning of displacement field components - and how do they relate in both cases, I tried to attached a toy model of what I am looking for but I don't know why I am not succeeding.
I have a 3D box and after loading it with say 1.5e4(Pa), I look at the solid.sy and see if it match my analytic calculation: sy=nu/(1-nu)*(1.5e4+rho*g*l) I do includ gravity. and I see that for the linear case it does match but with the (∇u)^T (∇u) under solid Mechanics eq. that you have mentioned, become more dominant - I should change my analyitc eq accordingly. I red abot solid.J the determinant of F - I calculated it when ignoring all displacement field components but wZ, and when I devided my equation by solid.J it matched the solid.SYY (Second Piola-Kirchhoff stress, YY component), and once more it fitted the solid.sy as for the linear case.

I am still have to understand for sure how to the displacement field components connected in both cases, in the linear case when my box move down due to gravity and load no deformation means that Z and z match and w tels us to what depth the box moved. in the non-linear case deformation must be considered z deforms from Z and w tels us to what depth the box moved including the deformation..what bothers me is in the linear case w includs movement of the box and deformation and only in the non linear case w =Z-z. I feel that I am close to work this out (I hope).

thanks very much,
Shahrazad

Hi,

The displacement (u,v,w) has the same interpretation irrespective of whether the analysis is geometrically nonlinear or not.

When you switch in geometric nonlinearity, two important things happen:

1. The Spatial Frame is separated from the Material frame so that x = X+u etc.
2. Nonlinear stress and strain measures are used.

The fact that z=Z for a geometrically linear problem is consistent with that it should be linear (infinitesimal deformation). You do not want a problem to become nonlinear just because for example a load is a function of z.

Regards,
Henrik

Hi,

thanks for your reply,

yes indeed I understand that. it's just that I am modeling a small crack within this 3D loaded box, and trying to connect between the resulting mode I SIF at the crack front in both cases (linear and non-linear). in my caculation (the crack front is one of box xz plian and crack opening is into y), I hoped that by understanding how these u,v,w in both cases relate I could modify my equations and get good matching. so far I did succeed with my previous sigma-y equation and connected it to solid.sy and solid.SYY.
hop that soon it will work out with my othe theoretically predicted physical parameters.

thanks,
Shahrazad