Perforated plates to analyse using Finite element or simple check

[e104909]

Hi,

I have two cases of issues with the same problem but different connection. 1st is curved roof beam to column. 2nd is column to base plate.

I have a lot of things in my mind which makes my problem complicated. I want to make a simple approach but it seems that Im uncomfortable with it and im not 100% convinced. Please give me some opinions.

The problem is a connection of a curved roof beam (double angle back to back) to a CHS column which is bolted connection through a plate. In this connection is also a CHS runner beam connecting all the frames (curved beam and column) through welded connection. The diameter of the CHS is 88.9mm. So, it is like a base plate connection (pipe connected to a plate through welds and the plate to the angle roof beam through bolts). The main thing is the plate has a hole inside the CHS for the purpose of electrical wires. The diameter is 32mm. How should this be done? Is the portion of the plate outside the effective bearing area is ineffective?

2nd case is a connection of column to base plate which has opening of 40mm inside the CHS column.

I was thinking that the portion of the plate inside these pipes could be neglected provided the effective bearing area is met.

The code requirements is BS code.

Please see attached

Thanks in advance

 

[hokie66]

The holes inside the members would have little effect on the capacity. Architect's joy, engineer's cringe.

 

[Ron]

Agree. Neglect the holes.

 

[e104909]

Thanks for immediate response.

I am wondering. Up to what extent can we ignore this hole? Is the bearing net area could be the basis for shear and tension? As far as i know the bolts and bolt holes might experience the tensile, shear and bearing but the hole inside the fixed connected hollow section cannot be known unless finite element is done.

Im worried that the plate where the chs pipe is welded could easily shear off but i confused on how to check

Regards,
E104909

 

[ishvaaag]

The way to get the utmost insight on any nontypical connection of some complicated characteristics it is through FEM. It needs be however combined with common sense and engineering judgement respect the biggest stresses discovered, if likely to constitute an actual limit state or not.

If ready to get with FEM, typically one would have before some estimated worse loadings. You normally won't want to model the whole structure 3D in FEM, and then apply an extense array of hypotheses. Normally you investigate the joint. But you may identify a short number of worse loadings coming from other structural analyses that may deal with the more worrying or thought more unknown aspects of the structural response.

For example, a circular fillet weld against P,M,V,T is sizable with some normality by just using Pilkey's approach (you may need more if you enter the realm of fatigue, at least some limited solicitation or ranging values). Hence, confided in that you will be solving such issue, you need not to enter the fillet details in your FEM model, where you will be more than anything dealing with overall shape and loading position issues affecting the strength.

The same with small details, as bolts. You put the relative restraint between parts there, not a very close representation of the material thing.

Your connection seems to be fixed

-column to base
-base to vertical plates
-vertical plates to arch beams
-vertical plates to round members
-vertical plates to stiffeners

in spite of the fact of that maybe a hinge has been assumed there. Since the connection seems relatively small, the actual displacements considering the actual fixity may mean that some moment or top of column displacement may potentially damage your connection at some worse point. This means that, if not already done, I would consider also a fixed connection at the point in the analyses that produce the input data for the study of the connection, just to cover the issue, or alternatively modify the connection to ensure a hinge is formed.

 

[bgatto]

Maybe you can stiffen the L's a bit more. Maybe you can run a short piece of CHS with internal diameter slightly bigger than 88.9 mm through all the node. And then connect the runner beam to it - away (say, 100 mm) from the angles. Also you could add some stiffeners to the CHS-angle connection. In general the holes are really not wanted where they are now, but the magnitude of the internal forces is important for further considering. I would try to find another route for the cables - over or under the roof beams.

 

[csd72]

If you think of the base plate as a beam with 2 supports(the walls of the tube) you have two cantilevers that are simple plates but between the two supports you actually have the plates plus a contribution from the side walls of the tubes. Obviously this middle span is much stronger and can therefore take the loss of part of the plate which makes it into two t-beams at mid span.

Using this philosophy and simple calculations I have been able to justify base paltes to tubes to to about 1.5m diameter with holes up to 80% of the tube diameter.

If you cannot do simple calculations to justify the feasibility of your design then you probably shouldnt be using FEM.

 

[e104909]

csd72,

Thanks for your bright idea. I am interested to know further.

Can you elaborate more?

Your idea is for the connection of the column to base plate right? Would this applies to the connection detail I have attached?

FEM using software is quite easy to use. Modelling and observing the behavior of stresses is what I just do to know if I am doing it correctly. Knowing the proper way of meshing then comparing the absolute values to allowable steel stress and base pressure to concrete bearing is what I just did and the computer will do the rest. But going beyond what you want to imply is what I don't know which I am interested to. I am quite new to FEM and haven't taken my master's degree and I believe anyone who have taken any problem at first time would remain a question mark until you prove and get satisfied with the answer. It takes experience and exposure to get the nearest possible answer. So I think I should take this opportunity to step in and find for the correct answer.

Could you give me insight with these? sketch or any rough estimate on how you do so I can learn from you?

any preferred books to read on?

Thank you very much.





Regards,
E104909

 

[csd72]

There is an old CIDECT guide on tubular structures which should help.

 

[e104909]

Fellows,

I have reposted my concern here:

http://www.eng-tips.com/viewthread.cfm?qid=305365&page=1

I am encountering such strange results in my model.

Can you spot my mistake?



Regards,
E104909