Modeling problem for design of base plate

[Chandan321]

Hi Guys,
I am trying to analyse and design a steel 3D frame structure 9m wide x 30 m high x 52m long. The transverse frames are at 13m spacing. There are beams at 3 elevations. The end frames along the longitudinal direction are braced at all elevations. The transverse direction wind is supported by adjacent building.

The steel columns rest on concrete pedestals. In the modeling software (STAAD), the columns and the bracings meet at the common joint with the concrete pedestals. I released the moments at the bottom end of the steel column, and the bracings are defined as trusses.

---- Now for the base plate design, do I take the member-end-forces at the bottom end of the steel member, or at the upper end of the concrete pedestals. I think that if I take the bottom end forces of the steel column, then i will miss out on the forces carried by the bracings to the base plate.

---- Else, do I connect the bracings to the steel column at some distance above the connnection with the base plate, and then take the member forces at the bottom end of the steel column at the joint with the base plate for the design of the base plate.

 

[Lion06]

It sounded like the columns ends are at the top of the pedastals, so they are one in the same. Am I missing something?

 

[GalileoG]

I am not sure I understood you correctly, but, why not put a pin where the pedestal is (if it is not that way already) and design your baseplate for the reaction force of the pin?

 

[JAE]

I think Chandan321 has a diagonal brace also coming in on the same nodal joint so there are reactions from the diagonal AND the column both.

Since the model included the concrete pedestal as a "member" below this node there is no node reaction output to get since the node has no boundary constraints on it.

The concrete pedestal end reaction (at its top) should be the value to use - or simply take the brace and column output and manually combine them.

 

[edwin17th17]

or for purpose only of design of baseplate.. you may remove the concrete pedestal model and put pinned support..

not trying to be rude or anything, but
this is basic that the top member end force of pedestal should be considered.. if you're having problem with this, i wonder if you modeled correctly the footing and equivalent spring support at the footing..

 

[edwin17th17]

also this approach will overdesign the pedestal as in actual moment for pedestal design is on top of footing.. if you model in staad you go further down until bottom of footing which i assume to you modeled as finite element..

 

[Chandan321]

Thanks to all of you guys for your time and suggestions.

StructuralEIT, yes, the steel bottom end is the also the top of the concrete pedestal in the model.

Clansman, you are right and thats what i have done. the problem is that of the bracing connection, as JAE pointed out.

JAE, I think you understood my problem correctly, and yes, I agree that I should take the member-end-force at the top of the concrete pedestal. But how about: if I connect the bracings a little (say 300 mm) above concrete pedestal in the model and then take the forces at the lower end of the steel column at its joint with the concrete pedestal.

edwin17th17, I agree to your suggestion for taking the forces at the top end of the concrete pedestal. What do you think about the point in above paragraph.

 

[Lion06]

If you move the node for the brace up then you'll start to get some eccentricity. I suppose you could move it up 0.00000001', but why not just combine them manually or get the forces at the top of the pedestal, if possible?

 

[JAE]

I think StructuralEIT suggestion that you move it up just a bit to "fool the computer" and then get reactions at the bottom of the column. Not sure whether 300 mm or 0.000000001 or somewhere in between. Depends, I guess, on the programs level of precision.

 

[Chandan321]

StructuralEIT and JAE, I think its a very good practical suggestion. I will try both with 300 mm and with 0.0000001 and see the difference. If 0.0000001 works, then its good. Else will try a longer distance.

Thanks for the good suggestion.

 

[asixth]

I wouldn't try to model the pedastal.

Model as a pin base, if the bracing is eccentric then account for this eccentricity when designing the base plate or...

Model as a full moment fixity, pin the end of the column which is connected to the support and model a rigid offset for the bracing member.

 

[Chandan321]

asixth, could you please elaborate on "account for this eccentricity", "full moment fixity" and "rigid offset".

Specifically, I want to know: where to put full moment fixity, and how to model rigid offset.

 

[asixth]

Chandan321,

"full moment fixity", I was referring to restraining all degrees of freedom at the support, translational and rotational. I just tried to run this approach through my program and it was having some errors, I haven't put it through a comerical package yet.

"account for the eccentricity", take the member forces and reactions out of the anaylsis, draw yourself a nice free body diagram, look at the eccentricities (assuming the model is all done to centreline geometry), and adjust the disribution of forces as required to maintain equilbrium at that joint.

"rigid offset", the package you are using should allow you to offset elements that connect to a node, essentially this will decrease the rotational stiffness of the joint if a flexural member is offset, causing additional rotation. The offset is assumed rigid so additional displacement of the offset element will be dependent on the rotation.

 

[Chandan321]

asixth, thanks for your time and all the explanations. I will try them in my model.