The static study window shows up blank in the simulation I am running on Solidworks. Attached is the picture:
Is it a bug? What can I possibly do to run this successfully?
My initial 'guess' is that you have 50M DOFs and your computer doesn't have the horsepower to handle that much. Can you set it up with 1/4 symmetry? (Since your part appears to be cylindrical.)
Even though it is 1/2 symmetrical, it cannot be sliced because contact on both sides holds a part in place.
The specifications of my computer are:
CPU: Intel Core i7-6700 3.4 GHz
16 GB RAM
I have a hard time running something that big with twin Xenons & 132 Gb RAM. I don't think you can hope to run a model of that size on a box that small.
Aug 2, 2016 4:04 PM
Reduce your mesh density.
I do want the mesh to be of a high density. I get inaccurate results
I agree to James and Siavash, Reduce your Mesh density and apply mesh control to the area of concern
Mesh control lets me select faces. Can I select volumes? I have already split the model using Split feature
If the volumes are actually parts then it is easy & select 'mesh by part size'.
Unfortunately I am unable to do that. I need an accurate simulation
Aug 3, 2016 12:04 PM
Well, you can increase your mesh density to infinity if you like. However, you either should have a super cluster machine to solve it or being very patient!!
There is always some numerical errors as a result of many factors involved in simulations. However, there is always a trade-off between accuracy and computational cost. That's why scientists and engineers consider an acceptable margin for those numerical deviations.
On the other hand, I would again recommend to considerably reduce your mesh density, get some results, increase the mesh density, compare the results and see if its deviation is within the acceptable margin for you. In fact, you should perform "grid independency" test; until you conclude that your solution is independent of the grid resolution. In summary, you may get close results between 1 million and 1.5 million DOF. So, it would be completely pointless and wasting of your time to perform the same analysis with 50 million DOF.
I hope it helps to clarify the situation for you.
Cheers,
- Siavash
i am sure you get the most most most accurate result with 50m dof
Why did you split the model?
Use split face to the area of concern for that way you can use mesh control and select the split face then apply mesh control
or try to run a course mesh, then use h adaptive method the software will apply mesh control in areas of high stress concentration.
Running the in tire model in very fine mesh is not actually a practical way since its very long to complte and you need a machine that so much horsepower
