Hi,
I am interested to know that why one should use a reference point in applying say , a displacement of say 1 unit to a face of an assembly and not directly applying the displacement to the face the displacement of 1 unit . What is the use case of using the reference point in Static or dynamic analysis with reference to commercial package say, ABAQUS
Thanks for answering in advance .
Emre Sengul I observed that it improves the convergence and shockingly results also came different in both the cases. Simulation time yes,it decreases because of improved convergence i.e., less iterations are required per increment so total simulation time goes down.
The most fundamental thing in any commercial FE package is to create an accurate model of the physical phenomenon that one would like to investigate. Therefore, the way that a model is loaded should simulate as much as possible the actual load condition. There are different ways of applying a load to a structure and one of them is through using reference point. If you could provide more details about the model you are working on and its boundary and loading conditions, it would make it easier to justify which method would suit the situation best.
Kamran Tayyebi Thanks for your answer. Absolutely agree,with the model trying to simulate as much as possible the actual load condition .
I have attached the rough sketch of the simulation I am trying to do .
Description : Refer to the sketch, I am trying to simulate a bi material containing a notch . I am fixing the left edge completely and in the right edge I have shown a reference point in the sketch where I wanted to know whether constraining the eqn. with the right edge and the reference point and giving the displacement to the reference point is more appropriate or giving the displacement to the whole right edge nodes .
That's the brief description of the simulation I am trying to do .
Sorry for getting back to you so late. The boundary conditions are still unclear to me. The left edge does not seem to be fixed in your sketch. It's a simple boundary condition if the displacement loading is applied in the East-West direction. You can make the the elements and nodes of the right edge slave to the reference points in various ways. For example, you can tie them to the RP, or you can use a coupling constraint, or you can employ a rigid body constraint and etc. When you use various methods, you will notice that the response you get would vary based on the size of your mesh, analysis method and etc. To check your assumptions, you can verity your FE results against the results provided from a reliable source for a similar problem.
Kamran Tayyebi Thanks for your answer. Actually I found that the use of reference point enables to do the following :
- One need not to check the load Vs displacement curve at every nodal position (to the edge or region with which Reference point is attached) after the analysis is completed . One can simply check it at the reference point .
Also, I wanna correct that as I earlier mentioned that the use of it (reference point) might generate different results but nope after running the simulation multiple times and correcting my modeling errors I found that it doesn't generate different results.
Also, I wanna correct that as I earlier mentioned that the use of it (reference point) it might improve convergence but nope after running the simulation multiple times and correcting my modeling errors and integration algorithm I found that it doesn't generate improved convergence .
That's it from my side. Hope it might help someone in their research in the future.
- Badges
- Science topic