Instability Issue

[BadgerPE]

I am modeling an existing heavy timber tension type truss with sag rod at the middle panel point. Due to current loading on the truss, I need to add exterior "flitch-like" plates to the bottom chord. I solved for the properties of the "flitch" plate and entered them as a generic section set in RISA 3-D (v5.1). I then split the member bottom chord at the locations I would like to add the plates. I change the section set to use the built up member. This of course doesn't calc any properties for the section which is alright.

However, I have zero releases for the end of the wood only section and the built up section, but I am getting instabilities at the nodes where these two sections meet. If I fix them for x-direction rotation, the instabilities go away and there is no reaction at those points. I don't like to do that though as it looks goofy for our reviewer. Does anyone have any ideas why this would be occuring?

 

[JoshPlumSE]

Crackerjack -

If you look at the Joint Reactions spreadsheet, the program will tell you exactly what direction is locked for that joint. This is important.

I suspect all that is happening is that all the members that frame into that joint have a "pinned" end release. If that is the case, then the joint will be technically unstable (because nothing that frames into it can accept moment or rotation at that joint).

Most of the time, RISA is smart enough to auto-correct that issue without throwing up an instability setting. However, since you are running such an old version (released circa 2004) this may not be the case.

 

[BadgerPE]

Josh,

Both joints are set up for full moment transfer. There are no pinned end releases on either member. Maybe this is an issue because of the date of the program. I downloaded the v9.1 demo and I don't have any instablility issues in that version.

 

[JoshPlumSE]

If memory serves me correct, there may have been in issue with the calculation of shear deformation for certain members back in the 5.1 days. I believe certain arbitrary members could be interpreted to have a shear area of zero.

If this is indeed the problem, then running the model in the new version is probably your best bet. But, you may be able to get away with running it with the old version if you go to the global parameters and un-check the shear deformation box. Because if the program ignores all shear deformation, then it will not care about those members having a shear area of zero.

 

[BadgerPE]

Hmmmm....tried turning off shear deformation and it is still producing instabilities.