Learning Forum

[momidor]

Hi, 

Does someone know why for the same crack coefficient KI ( MPa mm-0,5) had been calculated six times, so I got six different results ?

Thanks in advance

Jurek

[David Weed]

Hi Jurek,

 

Those are the contour results. In a fracture mechanics analysis, the fracture parameters are typically expressed as a line integral e.g., J-integral. In an FEA context, this line integral is discretized and when evaluated yields the values of the fracture parameters. You can read more about this in the help in the Fracture Analysis Guide section 1.2.2.2. Domain Integral Method for Calculating the Fracture Parameters: help/ans_frac/Hlp_G_STR10_2.html

You can specify the number of contours under the details of the crack object settings:

In theory, all contours should produce the same value since the integral is path-independent. However, in practice (within an FEA setting) these results will be affected by the mesh quality; you'd like to see contours 2-6 have a very similar result, which indicates that the crack problem itself is well-posed and the mesh is acceptable. The first contour usually has the most error since it is the contour closest to the crack tip singularity.

 

[momidor]

Hi David, 

Thanks!

Three more questions, if I can. 

Q1: I applied a pure tensile stress 180 MPa. It is  a Type I opening mode.

If such why I got possibility to obtain SIFS K1, K2 i K3, i.e. for modes II and III as well ?

Q2: Why the VCTT is zero J/mm2 ? 

Q3: I can't find the link to Fracture Analysis Guide section 1.2.2.2. I found to THIS, where is 1.2.2 only. 

 

Thanks in advanced

Jurek

 

[ammorey]

Dear momidor , 

For the six contours , this is depend on your setting , you can add as many as you want but six much sufficient . The more important now is which one you have to select out of these results ? simply I will tell you you can chose the fifth results , avoid the result of first contour because its near the crack tip and the material can shrink. For the other contours you can see the result of each are approximately equal so the most people chose the 5th . for more details you can read the most important paper which done by Rice 1968 HERE . 

good luck 

 

[Sandeep Medikonda]

Hi Jurek, Please remove any links to Sharcnet and use ANSYS Help. Find the link to Fracture Analysis Guide section 1.2.2.2 here. 

If you don't know how to access help, please see these 2 Posts:

Post 1

Post 2

Regards,
Sandeep