Beam w/ Pinned Ends = Instabilities?

[BigTank]

Huh?

I've just received an older version of RISA 3-D from my boss to play with. Admittedly, I burned through the 'users guide' (tutorials), and have glanced through the 'general reference', so by no means have I read all of the documentation.

I decided to try to model a beam just to see if I could get results. It's a beam with a distributed load and 2 point loads. The ends are pinned. The results I get are nil. Nothing. No moments, no stresses. It gives me an error that reads 'Instabilities were detected and locked for this model'.

I must say that, in my opinion, the included documentation doesn't clearly define what is required of the model and its setup regarding loads in order to successfully represent and solve a design problem. It appears that there is much more required than simply drawing and defining a member, its section and material props, and its supports, but exactly what that is isn't clearly defined in either manual.

Don't get me wrong, the software seems very robust and user-friendly. My problem is with the documentation...which does make an honest attempt at being complete. I think they might have missed the trees for the forest, however. Can anyone point me in the right direction? I could really use a general dumbed-down numbered order of operations from blank sheet to fully-defined model.

--------------------------------
Fitter, happier, more productive

 

[JoshPlumSE]

The tutorial is probably the best place for you to start. Specifically, Tutorial #3 (loading). I believe you will find some things about Load CATEGORIES there that you are missing. I'm looking at page 80 of the version 7.1 users guide. But, your pages might be different.

Based on your description, it sounds like you have probably defined your load combinations using load categories (DL, LL, et cetera), but have failed to use them in the Basic Load Cases. When that happens, the program doesn't know which load are defined as DL or such.... Hence, the fact that you get all zeros in the results after you solve.

 

[JoshPlumSE]

The instabilities issue could be more complex.... There is a Stability section in the Help File that will discuss the possible causes the problem more completely.

 

[msquared48]

First support pin pin, second support pin roller.
Release moments at both supports.

Check your member end conditions specified too as well as the support descriptions.

Should work as I use 3D all the time for some more complex 2D models.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[miecz]

When RISA gives the message 'Instabilities were detected and locked for this model', go to the Joint Reactions sheet. There you'll find which joint was locked and how it was locked. Imagine your model with this lock removed.

 

[BigTank]

My apologies, folks. I jumped the gun a bit on the questions. I figured out that the end releases are as important to check as the joint type. This seems to be a very nice robust piece of software. After fiddling with it for a few days, I'm beginning to find a method for model input, but I would like to know if anyone has developed their own method or if there is a sort of 'RISA for Dummies' methodology anywhere out there.

I have a working model, but it appears that the stresses and deflections I'm finding are around a factor of 4 higher than my rough hand calculations. A model definition methodology from some other source might instill some confidence in my methods...

--------------------------------
Fitter, happier, more productive

 

[JAE]

RISA will include shear based deflections in addition to flexural calcs. So your typical delta = wl^4/384EI won't always match the RISA output.

As far as a method for model input - there are various ways to input data in RISA without any specific order but this is what I do:

1. Input/check the global data (see the globe icon)
2. Set up section types for all anticipated members
3. Import or input joint coordinate data.
4. Input members
5. Input Boundary Conditions
6. Set up Basic Load Cases
6. Input loads onto the model
7. Set up Load Combinations.
8. Run a test of the model to see if it runs...correct any problems.....look at the deflected shape to see if it makes sense.
9. Input member design parameters (unbraced lengths, etc)
10. Begin running the model and developing the design...this might take multiple runs as you fine tune the member sizes, model geometry, bracing, etc.

Note that two of the most important parts of ANY model input are items 8 and 9. Many times you find folks running models and "getting answers" before even considering entering unbraced lengths, etc.

The deflected shape is also a very important aspect as it is a significant step in debugging or checking whether your model is making sense.

Finally, some type of separate efforts to verify your model output is important - hand calculations, spot checks of member designs, etc. Don't trust anything.

 

[miecz]

The deflected shape...checking whether your model is making sense.

And if not, you can print the deflected shape to decorate your office.

 

[BadgerPE]

A little new to RISA myself. When I was developing a model in 3-D the other day, I couldn't for the life of me figure out why I was getting the same error. Turns out one of my members was grossly undersized. I fixed that issue and my model worked fine. Not sure why that happened, but I thought I would pass it along.

 

[rwpe]

Back to your first question about the instability issue...For RISA 3-D, I rarely use the "pinned" boundary condition when you are creating a model that could be analyzed as a 2-D model, i.e. a single beam span in one direction. With the 2-D model, it is always assumed that your member(s) are fixed in rotation along the axis of the member. Using the 3-D program for a 2-D problem, you will have to fix that rotation.
Another issue (without knowing your boundary conditions) that would cause problems is if you pin the end of the member at a pinned boundary condition (yes...I have done this...).

 

[msquared48]

Another point in the modeling here to get results and not a null matrix. Not really applicable in your case, nevertheless...

At any pinned joint, you must have one of the intersecting member ends fixed, but the other member ends can be pinned. If you don't do this, the member will just spin in space and no solution, a zero matrix, will result.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask