Momment Releases 3

[anonymous119]

Dear all my question is that how to be sure that in the member where release is assigned is not transfering the moment to other member, I am dealing in steel structure with welded connections? Because sometimes if you not add release the software will give failed member result.

 

[Ron]

anonymous....I've seen several of your posts and in my opinion you need someone locally to help you understand the structural concepts about which you inquire and the software that you are using.

Modeling a structure with software does not dictate how it will perform in reality. Conversely, you need to model the structure in such a way as to reflect its expected performance. In a welded steel connection, you will get some moment transfer, ranging from a small amount all the way to full transfer relative to the stiffness of the respective members. The respective member sizes, weld sizes, geometry and stress risers will all control the moment transfer.

In short, I think you need mentoring that a forum cannot provide.

Good luck.

 

[anonymous119]

Thanks for guiding me Ron but I have asked such an amature question (ok I admit I am an amature) because I have seen a design in which moment releases are assigned all over in a welded connection by a senior designer.

Now thats why I have asked this question because I know a welded connection tranfers the moment, as its a moment connection, on the other hand a single pin doesnt because its a shear connection, but why to assign releases if its sure moment would be transfered to the other members, or it just depends simply on the design conditions.

Yeah I will work on to get mentoring I know that.

 

[rb1957]

yes a welded connection transfers moment, but is it fully fixed ? ie does it prevent relative rotation between the two pieces ? probably not; it's a good enough assumption (and maximises the moment at the connection), but if you believe that the structure can accept a small rotation then releasing the moment will give you the other end of the reaction spectrum (with no moment being transfrred between the two pieces). but this solution increases the moment in the two pieces (consider SS and cantilevered beams). and fully released (ie pinned) may imply that some of the weld has failed, and what would stop it all from failing ? maybe you have three weld segments, two carrying the moment, one the web shear, so the connection should be able to continue to transfer shear. maybe you don't want weld's to break under load ??

 

[anonymous119]

So, rb1957 do you mean that whenever the moment is released,if there is a welded connection, the capacity of the member is to be increased because it has to carry greater moment all by itself because it cant transfer it & where the welded connection is not supposed to be behaving properly because its not carrying the moment properly.

Are there possibilities that the designer may be thinking of that during construction the weld may not been done properly so in in order to take extra caution he introduced releases?

 

[rb1957]

i think what i'm saying is that you need to be very careful releasing the moment at a welded connection. sure you can analyze the structure to be the worst possible assumption (analyze the end connection as fixed (full moment capability), analyze the beam as SS). sure you might analyze as though it's a shear connection BUT if the connection does react moment, and IF that causes the weld to fail, THEN you've got a problem.

 

[anonymous119]

OK on more thing can we assign releases where extrsordinary values of moments are coming at the ends?

 

[msquared48]

You can assign anything you want, but as Ron suggested, is it realistic structurlly?

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[rb1957]

sorry, but you can do whatever you're comfortable doing.

IMHO, if you want to release a connection in the model, you should make sure that the structure will act that way.

i imagine that you're looking at FEA output (on a screen), and i imagine that it's a linear FEA. thus it doesn't understand that the sub-structure is (presumably) flexing under the loads it's predicting. if your assessment is that the internal moments it's predicting are "extraordinary", then you can adjust the model; but you can't adjust the reality !

 

[BAretired]

Thanks for guiding me Ron but I have asked such an amature question (ok I admit I am an amature) because I have seen a design in which moment releases are assigned all over in a welded connection by a senior designer.

When designing a truss by hand methods, all joints are assumed to be hinged, even though they are really not. The chord members are continuous and web members are usually welded to the chords.

Why would a senior designer (or an amateur) assign moment releases all over with welded connections? I can think of two reasons: 1) he may wish to test the accuracy of hand methods and 2) he may wish to design web members without relying on end restraints. Perhaps others can come up with more reasons.

BA

 

[msquared48]

A mistaken notion that the analysis is simpler, but conservative.

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[dhengr]

Anonymous119:
This whole thread scares the hell out of me. We can’t see your structure or details or what you are imagining, from this distance. As Ron said, you truly do need a local mentor if you are going to do this kind of engineering without hurting someone. Ask your senior engineer why he released the moments, whichever moments you are talking about. You and he can be looking at the same plans and details and he can explain what was intended and why. We are only guessing, since we can’t see your conditions, and you seem not to be getting the message, that you must have a good understand what you are doing, when you do this kind of engineering. The structure will act the way you force it to ack, by the way you design and detail it. Then you damn well better model it to act the way you have forced it to respond to the various loadings. We are not trying to be unhelpful here, but you really do need someone, right by your side, who can help you every step of the way, until you really understand what you are doing. The idea that you can use a computer program to analyze/design something, and then everything is just fine, irrespective of the fact that you don’t understand what you are doing, is really very dangerous.

 

[anonymous119]

Aright guys I give you information about that structure. It is a skylight which is supported by steel tubes It has the following specs.

a). 5.5m x 5.5m area
b). Steel tubes 100x100x3mm( Both Sides),
c). The distance between two beams (tubes) is 1.65 on an average,
d). Wind Load 0.85 KPA pressure (downwards) and uplift (upwards),
e). Dead load 4 tons glass and aluminum + Self Weight.
f). Releases are assigned in longitudnal direction.
I have attaced an autocad file pardon me for the drawing. (I am working on to get the staad file also or a one which is designed on the same basis as the orignal one)

The struture is already designed and standing and for the relief of all you guys I am not the one who has designed it.

dhengr that senior designer is not approachable otherwise I would have asked him. I have already explained that I am working on acquiring mentoring. I know that software is just a calculator its upon the designer how he takes the output from it. Also I will not take the responsibility of the work which I cant do because we are working for making the lives of people better but not to harm them.

 

[Ron]

anonymous....just because the connection is welded does not mean it is a "moment" connection. There are many conditions whereby a welded connection can act similar to a pinned condition.

As for releasing constraints, my only warning is the same as Mike and rb1957....make sure that the releases are intended to reflect the way the connection will act in reality.

Connectegr.....where are you? You're up.

 

[hokie66]

I wonder why using software is required in the design of such a small structural component. Many of us wouldn't be able to open your file. I hope those 100 x 100 tubes aren't spanning the full 5500.

 

[Ron]

Based on your drawing, you have a grid of 100mm tubes with pyramidal glass skylights attached.

Your spans are rather long so I'd be concerned about deflection with that much glass.

I've designed almost the same thing in aluminum and acrylic, not glass.

As for your modeling, I would model all of the tubes as being fully welded and fixed, but the whole structure is simply supported at the concrete and "I" beam.

 

[anonymous119]

Up til now I have got a lot of what could I get from a forum thanks for all you guys and I will be working on all of your professional advice including getting mentoring, definitely its the real structural behavior for which we have to model the design not just our own assumption .