hello, Mohan

I would only advise you to check again whether you are doing everything correctly

1) You should create and actvate local CS, so the x-axis is parallel to the crack face and the crack lays at the negative direction of the x-axis. By the way, make sure you create cartesian CS.

2) You should correctly define path, which consists of 5 nodes. The first one at the crack tip, the next two - at one face (the second is closer to the tip and the 3-rd is farther). And the last two - at the second crack face (again, 4-th is closer to the tip and 5-th is farther).

Hai Alexey,

 Thanks for your reply, I have tried it once again. I have attached the image of the crack path and a text file which contains the code of the model.

I would be glad if you could help me.

This is not a crack from LEFM point of view so you cannot apply KCALC to this domain.  The crack must be a slit with zero width, so the crack faces coincide. Therefore replace the re-entrant cotrner with a slit.

Thanks Alexey for your valuable comment. I could get the value of SIF at a crack width of .02. 

The results provide a value in which K1 is very small in the range .23 and K2 is very large with a value of 553. The crack is perpendicular to the applied pressure but why does the K1 value is very small.

Active the crack-tip coordinate system as a result output CS. Main Menu > General Postproc > Options for Outp > Local CS for result CS, or simply call RSYS APDL command. And after that call KCALC.

Thanks a lot for your timely help. I got the results perfectly :-)

The Ansys value is three times greater than the theoretical calculation. Is it normal?.

I had referred Tada book for theoretical calculation and I have attached the base model for your reference.

first of all, check, whether you are using the same plane strain/plane stress state in the both calculations.

HI, I have a similar problem. Did you find the reason of the difference between the analytical value of the SIF and the one of Ansys?