topic Re: Linear release issues in Robot Structural Analysis Forum

Linear release issues

MRCH94D9W — Mon, 20 May 2019 06:20:24 GMT

I have a slab supported by columns. In the interim stage, the slab consists of prefab concrete elements, which would not transfer loads between each other.

I have tried to model it with linear releases, as shown in the figure below (only released one element, should be all).

However, when I release an edge, the release works not only between the other elements, but also applies to the columns and the structure becomes unstable.

How would I work around this issue?

Kind regards

Martin

Re: Linear release issues

Rafacascudo — Mon, 20 May 2019 12:52:54 GMT

Do you really want to support each isolated slab only on 4 nodes(top nodes of pile cap bars)?

Re: Linear release issues

MRCH94D9W — Mon, 20 May 2019 13:23:19 GMT

Yes, I'm using the "Reduction of Forces above columns and walls" features, which gives some reasonable results even with just the one supporting node.

I have considered using rigid links to "average" the results, but the Reduction feature provides better results.

Re: Linear release issues

MRCH94D9W — Tue, 21 May 2019 06:55:34 GMT

In addition to my last reply, please find this picture of the bending moment distribution of the slab element supported in just one node in each corner.

Re: Linear release issues

Rafacascudo — Wed, 22 May 2019 14:39:09 GMT

Based on your last picture ( one isolated slab) I built a more detailed model using compression only bars to simulate the contact between the pile cap slab/bar with the pre-fabricated slab 50cm thickness slab

I was expecting that the support would come from the inside nodes as the slab is "landed" over the pile cap slab.

But no . Due to the columms flexibility (almost 15 m height) , the pile caps rotate and support is given by the back nodes near the columms.

In a model with 2 slabs ,the center columm gets balanced and doesn´t rotate ,and as a consequence the 50cm slabs gets supported by the pile cap external edge nodes

Comparing a panel cut showing moment across the slab is easy to see the difference

Also panel maps showing bending moment map on global X direction(XZ plane) for both cases shows the difference

Of course if you can get more stable columms through bracing , stronger sections , the results would be different.

For the whole model ,The way it is now , inside columms will have small rotation on the top and the outer columms will rotate.

Sorry for the delay . Spent some time studying this case

both models attached

Re: Linear release issues

MRCH94D9W — Thu, 23 May 2019 06:31:37 GMT

Thanks for the effort! It quickly becomes quite a complex model 🙂

I have worked on it myself and have come up with a fairly simple solution which provides reasonable results.

I have divided each edge on all slabs according to the mesh, so the supported node above the columns have an edge of its own. That way I'm able to release the remainder of the edge.

I was concerned that it would yield some unfortunate force flows through thesingle point but the slab seems to be behaving as intended.

Released bending moment