Torsion in a cylinder with crack

[YeamonKemp]

Hi,

I want to do an ANSYS analysis of a cylinder with a crack, subjected to torsion. I need to obtain the Stress intensity factor and simulate plastification.

I tried to simulate the load with MPC184 and a master-node with torsor moment. I also tried applying direct loads on the cover nodes, but in both cases the results were not good.


Which is the best way of simulating this case? Element type? How to apply loads?

 

[TERIO]

Why weren't the results good? It would probably help if you provided more detail about what you tried (e.g., element types, material properties, orientation of crack, etc.).

 

[YeamonKemp]

I used Solid95 and MPC184 for connect the master-node.

I attach the macro. It contains other stuff in order to calculate the limit load, but the most important for me right now is the SIF.

The results are not good because they don't correspond with the ETM analytical solution...

 

[rb1957]

i suspect there's an ANSY add-on to calc stress intensity (probably from a J-integral)

else use a handbook, like Rooke & Cartwright

 

[YeamonKemp]

There`s no problem to calculate SIF, with KCALC command. But I don't get good results. It might be because of the element type or how I generate the torsor moment. I suspect the stress is not well transmitted...

However, I can't use a handbook. The purpose is to simulate the SIF.

 

[rb1957]

the way i read your input yield = 249MPa = 36ksi ... looks reasonable, but E = 3720 MPa = 0.5E6psi ... looks low

if you're got Al E = 10E6psi = 69GPa

 

[YeamonKemp]

The problem can't be there. I tried the same E (s-e) with a cracked prism and it worked perfectly.

Furthemore, the non-linear analysis is not necessary to calculate the Stress Intensity Factor.

Thanks anyway, I continue with my research...