Moment release(s) define 2

[m00001]

I am a mechanical engineer, trying to use a simple piece of structural analysis soft. I got stuck in defining the MOMENT RELEASES for my frame:
- six columns in two parallel rows of three each, having all fixed bottom support;
- one beam is running on top of each row of the three columns. The beams' ends are hanging in cantilever beyond the first and the last column top;
- running @ 90º about the previous 2 beams, and resting on top of them, I have 8 other equal spaced beams. These beams' ends also hang beyond their supporting beams.
The whole assembly is welded.
All loads are linear:
- vertical down, on the whole length of each of the upper 8 beams.
- horizontal on the first row of 3 columns, parallel to the upper 8 beams.
My model doesn't work and the area I have doubts about is the definition of Moment Releases, where the manual doesn't explain anything.
I need help (Please understand I am a novice in this field).
Thank you.

 

[umit]

Please be more specific,
and explain why your model did not work?

 

[m00001]

Thank you for paying attention, umit.
I was under the impression I must release of moments the end of the beams hanging beyond the support columns (cantilever). I figured it myself in the mean time, based upon a lot of trial and error and a couple of short nights.
Anyway, thanks again!

 

[PEStructural]

Actually, the base of the columns should be pinned. Unless you can quantify a way that there is going to be zero rotation at the column base, then pinned would be more appropriate. What type of column base is being used?

Also, the moments do not need to be released at the ends of the cantilevered beam because they are zero anyway. The location of the releases, on the other hand, should be the column/beam connections. Unless you have a moment connection in which you weld the top and bottom flanges to the column flanges, and stiffen the column web, then it is appropriate to release the moments at the column/beam connections. Having said that, if you release the moments at the column/beam connection and change the column base to pinned, then you need some kind of lateral bracing between the column and beams. I usually use 3'-0" (horiz) x 3'-0" (vert) knee braces bolted to gusset plates which are welded to the underside of the beam and the side of the column (45 degrees).

I know I probably confused you even more, but please respond if you have any questions.

Hope this helps,

Mike

 

[UcfSE]

Is it giving you some sort of stability error? Are you modeling the whole system with a 3D model or more than one 2D model? It's kind of hard to really help you out with just a little bit of information. If its a stability error, then I'd say you are right, that you have probably made a mistake for which moments to release. Try to go through the model and look at each member and node and decide which way it is supposed to rotate or twist, and which way it is not, and keep in mind your connections so that your model matches the physical reality of what you're doing. For instance, if you plan to make it a simple bolted connection, don't model it as fixed, and so on. I don't think there's a problem making the base fixed as long as you do it that way in your details. You will never achieve true fixity but you can probably get close enough with a good design. Of course, making it pinned and using bracing works too. It all depends on where you want to spend money, e.g. bigger foundations and fixed bases or bracing and more details.

 

[m00001]

Thank you, mijflorio and UcfSE.
My columns are welded to a heavy channel imbedded in the concrete foundation (J hooks and all).
I have gussets to take the wind loads.
The whole construction is welded.
Am I wrong considering all connections fixed?
With regards to Moment Release I figured it out and the analysis shows it right.

 

[PEStructural]

It sounds like your column bases are close to being fixed, so that's OK. However, the column to beam connection might not be fixed unless you weld plates on top and on bottom of the beam flanges to the column and stiffen the column web with plates. Also, you would need to cut a "dog bone" in the flange of the beam a couple feet away from the column to create a plastic hinge (per FEMA 350). You can also weld the top and bottom flanges of the beam to the column flanges with FULL PENETRATION WELDS. If they aren't full pen welds, then it's not considered fixed. It's all about the detailing of the connection. Are you just using a gusset plate welded to the beam and the flange of the column?

 

[m00001]

Thank you much.
I have everything welded directly, with gussets on top, between column and beam. The welds will have a full penetration.
Thanks again,

 

[DRDearth]

Your post is complicated... I would guess your problems are related to the Beam local coordinate systems. Beams have their own coordiate system. X, Y & Z, or i, j & k.. depending on how you want to call it.. experiment using a few sample problems to make sure you are not mixing up your coordinate systems.

David R. Dearth, P.E.
Applied Analysis & Technology