Frame modelling

Hi there, please I am working on a steel frame design that includes OWSJ in the model. I have tried various forms of releases for the joist members like assigning truss parameter to the diagonals and verticals, releasing moments at the ends etc. but when I run the analysis I get the following errors – member design failed for member xxx, Load case xxx has infinite or nan values and it doesn’t show utilization results too

I have looked through the forum and it looks like it has to do with the member releases but I’m still not able to solve this. Can you please look through my attached model to see if I’m doing anything wrong? Your response will be very much appreciated. Thanks.

The attached model has not been run yet as it takes some time to run but maybe you could take a look at my code file (loading, releases etc.) prior to running to see if I missed anything.

Frame model release variation1.STD

  • What are you trying to model where members 4322 4820 connect? As it is, it is unstable. Just get rid of the release at the ends of the members that connect to node 2329.

    Also, you can't have X bracing where all four members are truss members. You have that all over the place. Even fixing those two things I'm still getting hundreds of instabilities. I don't have time to investigate further. Go to the output file. Find instability warning like this: "***WARNING -  INSTABILITY AT JOINT  2334   DIRECTION = MZ ", then examine that node in your model to see what could be causing it.

    Oh wait, this is your lucky day. I just took a look at the one I mentioned and I think I see what's causing the rest of the instabilities. On your roof trusses, don't make the two members I've highlighted truss members. I just removed them from the TRUSS command and it fixed the instabilities. It's because when these are truss members the entire bottom chord can rotate about it's long axis. You'll have to think through how you want to release this member.

  • Thanks for your response. 
    You pointed out a huge error on my part. Node 2329 is not useful and should have been deleted. It was my master node when I used master-slave specification to model diaphragm but apparently I forgot to delete the node or merge the members.

    When you say I can't have x bracing where all four members are truss members what would be an alternative?

    I agree with you those members need not be truss members. I will release the moments at the top ends of the member and remove the truss specification for those members. Let me know if you disagree or have any other suggestions. Thank you!

  • That sounds okay. If it makes you sense you might want to consider going with 2 members that cross without connecting.

Reply Children