2 Adjunct Membranes not Connected


I have been running into an issue after initial form-finding. Basically, we are taking the geometry from the form-finding step and running non-linear IGA. It appears the geometry of the two adjoining membranes is just off ever so slightly, such that two cables are created at the spine. As such, there is a discontinuity in the results.

I have tried connecting points (disabled in the attachment), but no luck. I have also tried coupling features as well.

Do you have any suggestions or advice on correcting this geometrical issue?
2022-01-06_16-28-23.pdf (57.4 KB)

5 Point Sail Surface Analysis.gh (34.1 KB)

it is not neccesary to use the correct material in the formfinding as it is not considered. Only the prestress relation of the different members have an influence on the converged result. The Young’s modulus of the material influences only the speed of convergence. So if you set “E=1”, it should work. The problem is related to the stiffness relation between membrane and coupling elements. Otherwise you have to increase the penalty factor for coupling in the AnalysisModel component menu or in the input text by ModifyModel.

Hi Anna,

Thank you, but I don’t believe this addresses my specific issue. My issue is related to geometry that is ever so slightly off at an intersection between two membranes that should be connected. I have no issue with the form- finding; that works perfectly. It’s when I take that form and perform an IGA non-linear analysis that I have the issue of the two membranes not connecting to each other. I’ve tried the connect points component and that didn’t seem to analyze properly; same with the coupling feature.

you have to avoid that the prestress in the membranes pulls the surfaces apart from each other in the form finding. So the result is without gaps and coupling works correctly. keeping the ratio of prestress but decreasing the absolute values should avoid this. Same effect is reached by increasing the penalty factor.
Connect Point component can stitch the surfaces together. However, if the points are too close to each other (Rhino Modeltolerance), there might be an issue of correct assignment.

Looks like Solid-Union and Join BREPS after disassembling the form-geometry solves the issue. Thanks for your assistance.