Two questions about Ansys. Help needed!!!

[mual]

Hi all,

I have two questions about ansys, anyone can help me?

1- When opening the mesh with a ".ans" file and the mesh is very complex (about 4 million elements) ANSYS takes so much time (over 10 min) for the load. We tried to load it with Abaqus in the same computer and the load is very fast (2 min). Could you give us an advice for opening .ans files in the minimum time possible?

2- In the other hand, we need to buy a new computer to solve a contact problem (rigid – flexible) of a 3-tissue model (about 10 millions DOF). We are wondering if our problem could be solved over GPU (for buying a great GPU) or it is necessary a good CPU (we know that ANSYS could be launched in several cores). For a single PC (not a Cluster) what would be the best option to solve this problem as fast as it can?

I hope you could help us with this decision, thank you.

 

[admincae]

Hi mual,

I suggest you to buy some branded computer with ANSYS collaborate, like DELL or HP. I used with succesfull at work an computer with i7 quadcore and 12Gb memory of ram and 2 harddisk link in raid 0. I running an model with more than 3milion elements, with very much contacts betweens components, in couple 8 hours. For your study I think is useful to choose a configuration with minimum two processors (quad core XEON or Opteron) and minimum 128 Gb memory of RAM, plus some harddisk with minimum 10000rpm linked in raid0+1.

I waiting news from you about your aquisition.

Best regards.

 

[mual]

Admincae, thank you very much for your fast reply, could you please tell me if you use any speed up method to solve your contact problem. When i tried to solve a smaller problem “with less than 1 million of elements”, it is take me about 3 days, by using a good PC (16 GB de ram) y 8 processors "i can just use 6 of them".

Another question you recommended me to use a 128G of ram, but actually when i try to solve my problem it just use 8G of the available 16G. Could you please tell me how to use the maximum of it.

Thank you again.