flambter
New member
- Jul 21, 2009
- 24
Hi,
I have been trying to find the trivial relation between stress and strain when post processing the results in Patran.
I use the fringe plot with labels at the gauss points, without averaging and with shape function extrapolation.
This gives me labeled values at the four gauss points of my CQUADs. I read the stress tensors values. If I get the x, y and xy component of stress and manually calculate the von mises stress and compare to what Patran gives me at the four gauss points, it matches exactly.
Now, I read the same results for strain tensors but can`t match the stress values with the strain. Basically I would have expected, for linear analysis to read E = stress/strain. But for plain stress, one needs to account for the poisson effect. Even then, the strain tensors I read in Patran is really far from what I calculate.
One difference I am noting is that stress values are given at the two extreme fiber of the thickness of the shell while the strain values are "non-layered". This might give a difference, but the results are really far from each other.
For example, i read a sigma x = 5.69 psi, sigma y = 0.926 psi, xy = -0.874, E = 350000 psi, v =0.42. I calculate a strain of 1.18115E-05 with the following formula : ((1+v)/E)*(sigmax(1-v)-v*sigmay)
Patran gives me 2.49 E-06.
Any idea why and how can the two be related?
Thanks.
So the question that arises is do you know how stress tensors results in Patran are related to strains?
I have been trying to find the trivial relation between stress and strain when post processing the results in Patran.
I use the fringe plot with labels at the gauss points, without averaging and with shape function extrapolation.
This gives me labeled values at the four gauss points of my CQUADs. I read the stress tensors values. If I get the x, y and xy component of stress and manually calculate the von mises stress and compare to what Patran gives me at the four gauss points, it matches exactly.
Now, I read the same results for strain tensors but can`t match the stress values with the strain. Basically I would have expected, for linear analysis to read E = stress/strain. But for plain stress, one needs to account for the poisson effect. Even then, the strain tensors I read in Patran is really far from what I calculate.
One difference I am noting is that stress values are given at the two extreme fiber of the thickness of the shell while the strain values are "non-layered". This might give a difference, but the results are really far from each other.
For example, i read a sigma x = 5.69 psi, sigma y = 0.926 psi, xy = -0.874, E = 350000 psi, v =0.42. I calculate a strain of 1.18115E-05 with the following formula : ((1+v)/E)*(sigmax(1-v)-v*sigmay)
Patran gives me 2.49 E-06.
Any idea why and how can the two be related?
Thanks.
So the question that arises is do you know how stress tensors results in Patran are related to strains?