Negative load factor in Collapse run


Product(s):SACS
Version(s):23
Area:Collapse

Problem

SACS collapse is showing negative load factors which were not provided in the loading sequence. Why is such odd result and how to tackle this?

Solution

When you review the Collapse log file, first, you need to check the INC column (increment column). Suppose the increment column remains the same but the load factor changes. In that case, the analysis hasn't converged for the load factor, and the Collapse Advanced uses automatic sub-incrementation (SI option on CLPOPT line).

 

In Collapse Advanced, when you see the load factor doesn't match the load increment input on the LDSEQ line, it means Collapse Advanced uses a sub-incrementation or Arc-length method. Check these reports in the Collapse Advanced Troubleshoot log file (clptrbl.* file). 

 

A negative load factor can happen, but it usually means the structure or part of it has buckled under the load - the negative load factor is a result of the arc-length analysis. 

 

The model may have many small and non-structural members. Those members may be buckled under the load and leading to unloading. We recommend cleaning up the model or setting the non-structural members to elastic with only one sub-segment in the Collapse Advanced input file. 

While the analysis runs, please check the log file and look at the Increment Number (INC#) and Load Factor. If the number of increments remains the same but the load factor changes, then convergence doesn't achieve that load factor. The Collapse Advanced uses sub-incrementation with a new load factor. 

 

Also, you should monitor the Collapse Advanced Troubleshooting log file (clptrbl.*), which has more information. You can see what is going on in the model. For example, the lack of convergence may be is because of a member’s too much plasticity or just the load (residual) is too large. Collapse Advanced user manual has a Troubleshooting section explaining all this information. 

 

While monitoring the log files, if you notice something interesting, like a reduction in the load factor, please copy the clpdb folder and open it in the Precede/CollapseView. Then, review the results (for example, plotting base shear or load factor vs load steps). In the case of a negative load factor, it means the load factor or base shear reaches its peak somewhere in the analysis. Then, look at the members around the peak load to see if any member experiences elastic or elastoplastic buckling. 


Finally, you can check the load factor vs. time step curves and the member's status in the Precede to see what happened. 

 

This way, you don't need to wait for a negative load factor and can review the results before that.