The circular yield surface can be achieved by employing “plastic-moment-y” and “plastic-moment-z” commands for beam or pile elements in FLAC3D. It is clearly declared in the following newsletter published in 2019.
I want to define a plastic hinge not for the whole element but at the intersection of two pile elements. Such an interface plastic hinge definition is given in one of your examples, named “Plastic hinge formation in beam (double-node method)”, by structure link elements. I am sharing as follows:
When I used this method in two orthogonal directions, I got an uncoupled response, in other words, square yield surface. My code is as follows:
Is there anything I can do to achieve the circular yield surface for the structural links defined in two orthogonal directions?
Hello @cgulenc,
This question is a duplicate of 2D Plastic hinges in 3D model? Was your question answered in the linked forum post? I believe you had a meeting with @dpotyondy and @musicinquark, but I could be mistaken.
Hi Derrick,
Yes, it was my question but this new question is not the same with that 
.
Previous Question:
When we use “plastic-moment-y” and “plastic-moment-z” commands, the yield surface is circle. This is the answer of previous question. It’s OK.
Also, we discussed similar issues in the meeting. I know that if I use plastic-constitutive model, it is going to work. I know that if I use 3d zone elements, it is going to work. But those methods are not feasible to be used in my model. Therefore, I am specifically asking the following question.
New Question:
I don’t want to assign the plastic moment capacity to the whole element. I want to define this capacity only at the first node of the element. Therefore, I am using the “double-node method” that is given in your example. In this method, plastic moment capacity is defined via structural link element. When I use this method, I see that the yield surface is square. In other words, no coupled behavior between structural links defined in orthogonal directions is observed. Is it possible to provide the coupling by using structural link elements (with double-node method given in example above) specifically?
My code is as follows:
Kind Regards,
CGulenc
Hello @cgulenc,
As far as I can tell, the issue is that when assigning an attach condition to a rotational dof of a link, you are limited to the free, linear, normal-yield and rigid attach conditions (see structure link attach command — Itasca Software 9.0 documentation).
In your code you specify the rotation-y and -z attach conditions of the new link to be normal-yield which does not create a dependent model like the pile-yield attach condition does. The feature improvement you show above dated July 11, 2019 only applies to beam or pile elements that have their plastic-moments (plastic-moment-y or -z) defined.
Would specifying the plastic-moment to the sub-elements on either side of the node you wish to double be an adequate solution? Something like this:
Keep in mind elements with zero plastic-moment do not undergo plastic-correction, they remain elastic.
Thank you very much your reply.
It would work if the plastic-moment capacity could be assigned to the one side of an element.
In your example, it is OK because the moment at the other node of middle sub-elements will not increase (no other hinge potential exist) when the plastic hinge occurs at the middle of a simply supported beam. However, in case of existing the possibility that the other hinge can occur at the other side of element it doesn’t work.
Anyway, thank you. I will try to reduce the size of element having plastic-moment capacity as time-step permits me. I will try to minimize the effect of both sides’ hinge occurrence.
