Fracture Mechanics in Mecway

ivanthesi

Hi all,

Has anybody ever tried to do some Stress Intensity Factor (SIFs) study to obtain KI, KII, KII, GI, GII, GII, or anything related to J-integral calculation for fracture mechanics evaluation with Mecway?

Anything related to line integral? Python scripts?

Thanks in advance if anybody comes up with ideas.


Regards,
Ivan.

Sebastianmaklary

(Sorry for reviving an old topic)
I have done very simple fracture mechanical analysis by modelling a crack in CAD at different crack sizes and then running multiple analyses to be able to plot the potential energy versus crack size and thereof calculate the energy release rate G, which relates to the stress intensity factor by G1=K1^2/(E/(1-nu^2)). Such an analysis is very simplified as it puts a very strong assumption on the crack geometry and growth.

I am however very interested in whether it could be possible to calculate any fracture mechanical values using Mecway and Calculix but I am currently not deep enough into the maths and coding to see if it is even possible or if it requires some behind-the-scenes coding. ANSYS has put out a white paperon the use of material forces, which is a nodal vector that graphically shows the potential for and the direction of the crack growth and which is applicable to both linear and nonlinear materials including plasticity.

JohnM

In the past I have used NASCRAC which creates a crack growth rate based on a stress field and da/dn material data. However, I recall there was a method in ANSYS that used higher order elements, with the midside nodes to a "quarter point" at the base of the crack, and there was an analog for then calculating the crack stress intensities. I know we did this method in more than one FEA code, so it may have been a general method. Sorry this is vague but a bit Rusty on this,hopefully this gives you a trail marker or two.

Sebastianmaklary

I do recall learning about the crack element which creates a singularity due to the interpolation function used within it when the nodes are mapped "incorrectly" compared to the isoparametric parent element, so that it can model a sharp crack without requiring an inifinitely dense refinement. This requires you to know where the crack is and insert it as such but has the benefit that you are sure that the stresses at the crack tip is sufficiently refined. This however still requires the use of fracture mechanical methods such as the J-integral or the material force method.