Learning Forum

[karun00162]

I am using Ansys workbench 19.2 and I am getting repeated errors, or unexpected results in static structural model. I clear the generated data from model and restart Ansys as well but problem doesn't go away. I reinstalled Ansys 19.2 from 19.1 (student version) as well. Comments would be really appreciated. Thanks,

Karun

[karun00162]

In addition to the problem above: 

I am using fracture tool and trying to find Stress intensity factor and J-integral values using pre-meshed crack in DCB samples. Whenever I change load, K and J-integral values turn out to very unreasonable. Before changing load I clear generated data from model too. Are there any limitations/drawbacks in fracture tool model, which generates the k and J-integral once only (with certain load) unless and until a new model is made? 

[Sandeep Medikonda]

Karun,

  Please elaborate and explain with pictures, what you mean by unreasonable or unexpected results? I would suggest you to take a simple geometry, set it up the exact same way in 19.1 and 19.2. Look at the deformation results, solution output and compare. Post images and maybe that will give some insight into what is going wrong?

Regards,
Sandeep

[karun00162]

Hello Sandeep,

Thank you for the comment. Attached pictures describe two loading conditions; one at 15N and other at 20N. For 15N load, the stresses generated in the specimen are shown with the probe and the J-integral is 2144 J/m^2. After generating this data, I clear the generated data from the model and update the project. Then I changed the load from 15N to 20N, the generated data shows that, stresses in the specimen are too low which is unreasonable and the J-integral value is also in-correct. However, I further tried to build new model and applied the load 20N. The stresses generated were reasonable and J-integral was close to 3800 J/m^2.

After generating the data, I use unit tool to see values in different units (MPa or mJ/mm^2). I assume there is some bug which messes the simulation with units or something. I tried on different computers as well using Ansys 17.0, 19.1 and 19.2. 

Thank you,

Karun

[karun00162]

Hello Sandeep,

Is there anything I can do to fix this problem ? Eventually I have to find fracture toughness and J-integral values in between two different materials. I still face the similar problem in the multi-material model too. I do not get any reasonable J-integral values in the model. 

Karun

[David Weed]

Hi Karun,

I tested a premeshed DCB example in R19.2:

using loads of 15KN (first pic) and 20KN (second pic) and the J-integral results do not widely differ:

That is, after I cleared generated data, specified a new load and re-solved, the result change looked reasonable. Can you give us a few more details about the problem setup? When you say that you updated the model before changing the load, what did that entail? Also, is this for interface delamination, employing the VCCT method, etc? Lastly, are you solving in distributed memory processing mode or shared memory? You can check this by going to Tools -> Solve Process Settings -> Advanced and if the box is checked then you are running in distributed mode and if the box is blank you are in shared memory mode. Please let me know. Thank you.

[karun00162]

Thank you for your help David.