Frequency of Industrial Structure

[Jatfuentes]

I am doing an estimate of the weight of an Industrial Structure, with cranes in all spans.
As usual I modeled the tranverse frames as moment frames, and the longitudinal frames as braced frames, with the end of connections hinged.
The dimensions of the building are approximately 540 Ft width, 1000ft lengtht, and 66 ft height to the eave. The Isometric of the building is in attach.
My question is about the influence of the hinges in the period/frequency of the structure response.When I considered everything continuous, including angles of the Truss, I got T=3.29 sec, but if I considered the structural normal. ( rigid in the transverse direction) , braced with hinges in the longitudianl direction and also the diagonals of the trusses hinged, I got T =30 sec. !!
I am using RISA 3D. Obviously do not exist perfect hinges in a truss, the value of the restriction is between fixed and hinged.
Some one can give me an idea?. there is any paper talking about this issue?
I will thanks your comments.

jatfuentes

 

[GregLocock]

Adding stiffness or constraints to a structure will almost always (there is at least one exception, of no interest here) increase its resonant frequencies, ie reduce its response time. Replacing the rotational DOF by an elastic spring (or assuming it is rigid or continuous) will stiffen the structure. So far that agrees with your findings.


Typically in a linear FEA rather than modelling a hinge with kr=0 we use a small non zero positive value. Unfortunately this can be a crucial tuning parameter aka fudge factor in the model. If you perform an experimenatl modal analysis on a real assembly you can see this, as your frequencies shift with different excitation levels. As you turn the force level up hinges unstick and things start to flap.

So, if you are interested in earthquaake response, you'd better assume that the hinges are indeed free to rotate, whereas in normal use they might be jammed solid.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Elite7]

I couldn't find your attachment, but it seems as if some portion of your structure is becoming unstable on introducing hinges. Check out your mode shapes, and you ought to get a handle on it.

 

[Jatfuentes]

Thanks for your reply Elite7,
I dont know, because I sent the attach. Any way I send it again.
Really the modeling the bolt connections as hinges is a problem , because you cant use your same static model for a dynamic model.
I dont know if there is any research or PHD tesis for this problem. I have a book written by W. F chen about partially restrained connections, but is not focsed for seismic cases.

 

[JoshPlumSE]

Be sure to check your mass participation to tell you what your "fundamental" periods really are for each direction.

There are times when a localized stiffness issue will a mode with very little mass participation to have a very large period. When this happens, you can look at the mode shape to see the "near instability", so it should be easy to correct.... as long as you have only one or two.

 

[LOKSTR]

Jatfuentes,

In your model, I see a hinge at the intersection of X-B. Try making your XB as tension only member( truss member) and make it continuous from one end to other.
Try this and it may solve your problem.

 

[ToadJones]

I realize this model is probably not showing the entire building and this probably has nothing to do with your problem, but I would be hesitant to brace the end bays of a building of this nature...especially if there is any high heat applications or large temperature variations in the building. This restricts expansion of the building and can cause all kinds of problems on your runways and connections.

 

[271828]

Greg wrote: "...there is at least one exception, of no interest here..."

I am interested. What is an example?

 

[GregLocock]

A simple beam floating in space has a first flexural mode at a frequency. Adding pinned constraints to the ends reduces the frequency of the first flexural mode.

I don't have Blevins to hand but I think even with clamped ends the frequency is lower than free free.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Jatfuentes]

271828
Thanks for your atention. Yes, this is only a part of the . The building is Huge !. But now, the client is interested only in an estimate of weight, yhe most closer possible to the final weight. really I´m working a lot because, i have to buildings more, that allready I finish.
Yes, I will put bracings at end of the building.

LOKSTR,
I will take a look to the model.
You know, to analise concrete structures is very diferent that analise steel structures. Concrete structures are 99% monolitic. Steel structures, depend on if the structures is braced or not, if the connection is flexible or not, etc.
I think, i dont know, but I didnt reading nothing about this issue.
If some body have something talking about this problem, please let me know
Maybe a PHD thesis ???

thanks.

jatfuentes