Problem Modeling 1

[tsankar]

Hi,
I have a vertical channel section fixed at the lower end and a vertical load in space. The load is transferred to the section at two areas by some means (in the actual equipment. Here I am not interested in modeling the connecting members. I am analyzing the structure assuming the connections are rigid and transfer the load entirely to the structure). I have modeled this by distributing the vertical load equally on the two areas and applying a horizontal force on the top area so as to create an bending moment equivalent to that created by the space load.
I have used shell elements for the analysis. Am I right in applying the vertical loads on the two areas (meaning the faces of the elements) ?. I feel that applying the loads on the areas and henceforth the element faces, will not properly represent the compression caused by the load. Should I apply the load on the horizontal edge of the top most element in the section?. Can anybody help?


Thanks,

TS

 

[Qshake]

Am I to understand that you are modeling (or trying to model) a vertical cantilever beam (column) with a load applied at the end? What quantity is it that you wish to investigate with FEA? If I understand the problem, most quantities of interest, for design, are easily obtained from simple statics. Please explain further, if I have totally misinterpreted your situation.

 

[tsankar]

Hi,
Iam trying to model a vertical member(channel section) welded to a base at one end(I have simplified the problem actually.This section along with the base and other members forms a part of a frame).A vertical load is transferred to this section through two other members(belonging to the another frame which Iam not interested in.These are connected to the channel section member at two areas.). Iam not interested in modeling the two connecting members. Taking this as a case of eccentric loading , Iam trying to transfer this vertical load into a vertical load of equal magnitude applied along the member (compression in my case) and an equivalent horizontal force at the channel section member's end to cause a moment equivalent to that caused by the applied load).I wanted to know whether I have to apply the compressive load on the two areas where the connecting members meet the channel section member (which becomes a load applied on the face of the elements in those areas) or to apply it axially (along the horizontal edge of the top most element in the section ). Iam interested in knowing the stress distribution in the frame. Can you suggest whether my modeling technique is correct?. I hope I have explained the problem clearly.

Many Thanks,

TS

 

[Qshake]

I apologize for not being of much help but you situation still reminds me of a simple frame problem. You have a vertical member to which are connected to other members that transfer a load into this vertical member (channel section) which is welded to a base at one end.

If I understand further, you are then looking to apply the two loads to the vertical member and wish to account for the eccentricity. Correct? Since you have simplified the overall assembly to the vertical member I will only comment on that and leave out comments regarding lateral forces, if any, which need to be transferred to the channel.

I would, very simply, add the two vertical loads (appropriately factored, dead load, live load etc) together for an axial load on the centroid. Then I would find the resultant moment due to the vertical loads. Since these act on opposite sides of the centerline channel they will not add and may very well minimize the resultant on the vertical channel. Now you have a simple statics problem: a vertical cantilever column with a axial force and resultant moment. Since this is determinate, you can find whatever stress your looking for by using P/A +(-)My/I. And for this case it should compare very favorablay with a FEA analysis should one be run.

The FEA seems, in this case, more trouble than its worth as the actual stress will vary little and any quantity of stress that seems remotely interesting to me would be in the gusset plates at the connection of the other two members. Which really isn't saying much because I wouldn't use that to design the connection plates or fastners either.

I hope this helps, Good Luck.

 

[ishvaaag]

I think the same than Qshake, you can ascertain the elements pertinent to strength design for this kind of problem with such end conditions as a channel column fixed at bottom and free or another atop just from the AISC code or so.

Other thing is that you want to capture all the behaviour with a state of the art FEM package throughout the whole non linear range...for this whatever the way you model the end solicitations, at 1 or 2 times the depth of the channel you should be seeing more or less one passable portrait of what happens.

In any case I think have posted a Mathcad sheet for the check of channel columns under simultaneously eccentric on x and eccentric on y loads, if you have access to it, download and for LRFD check would be enough.

 

[kreative]

Hi sankar. from wut i understand this is wut i would do.

I would select the region where the members ( the ones u are not interested in modelling) join the vertical member. and since the load is acting at a distance from these regions, apply the load as if acting on a surface and at a point. (mechanica gives that option). hope i was of help.

-YM