STEEL FRAME DEFLECTION

[LOKSTR]

Hi everybody,

I have two questions.

1. How to apply the lateral force to get the true deflection of the frame?

e.g.
ASSUMING a single storey frame with 7 bays, with tension only bracing in 1st, 3rd & 5th bay)

If I have 400kN lateral force, I get different lateral deflection values if I apply 400kN at

a. 400kn at one END ONLY
b. 200KN EACH AT BOTH ENDS


2. What would be the best location to put the brace in a frame, near the ends or centre? And why? ( assuming no obstruction for bracing).

Thanks

 

[JAE]

1) By locating the force in different joint locations, you are changing the axial compression in the horizontal members - and the deflection you "get" (I'm assuming from some software) may be due to the difference in axial deflection (PL/AE).

2) In many cases, the braces are placed in the center of the frame - especially for freezer type warehouses, where the temperature differential won't cause the separated braces to fight each other.

 

[J1D]

Assume you have strut beams along the storey height. If the beam is very stiff (for axial compression/tension), the lateral load transferred to each braced bay is the same. The deflections should be the same (or very very close). If the struts are of small sections or you have more complicated framing, the deflections might be different.

 

[pba]

It would be sensible to investigate why your deflections are different - I would expect the results to be similar for both load placements in most typical frames. Perhaps you could give some more information of the frame geometry?

As to where to put the bracing - JAE's comments are or course correct but for a different take:- Put the bracing where you expect the erector to commence the erection. They generally will want to start at a corner or end of a building so first bay would be good.

 

[LOKSTR]

pba,

Frame geometry is as follows:

ht.=8.7m, bay width=11.2m, no. of bays=7, single storey

Tension only brace in 1st,3rd & 5th bay.

Lateral force applied=400Kn at one end or 200kn at each end.

Properties of all frame members are:

Area=6452 mm², I=1.05e+08mm^4, Z=2.95e+06mm^3.

Model this in any software and see the deflection at each end. It is different.

I got the values as 8.6mm & 1.3mm when 400kn applied at one end only.

Def= 4.93mm & 6.67mm when 200kn applied at each end.

Which is correct value of deflection?

Thanks,

 

[J1D]

Lokstr,

I modeled the frame and obtained the same results as you got. To get the properties the section is either 330x330x5 (in mm) square tube or 350x6 pipe. The walls are very thin and not typically used (as I know), the local buckling should be checked or you can used only an effective sectional area.

The different deflections are caused by the different axial deformations of the strut beams and the diagonal bracings. You raised an interesting one storey frame case, the axial deformation plays a big role like a truss.

 

[LOKSTR]

JID,

The properties of section is not important. I used the default properties of my software. You can put different properties for columns, beams & brace but the deflection results will still vary.

My question remains the same. How to get the actual deflection of frame? Where should we apply the lateral force to get actual deflection?

Also, position of brace does affect the value of deflection. So what is the best location to minimizereduce lateral deflection?

Thanks,

 

[J1D]

The frame deforms like a truss, the axial deformations will therefore keep dominant.

Back to your original questions. 1) consider whether you have to use three anchor bays. If the frame is exposed, you will have big temperature stress. One anchor bay should be enough if the foundation is sufficient. 2) apply loads where they initiate. You can change the locations of bracings to adjust the deflections. But the deflection/height ratios are small in the braced frame.

 

[apsix]

The most accurate deflection is the one where the lateral loads are applied so as to most closely model the real world. If it is a 60/40 split from wind force on windward and leeward walls say, then apply the loads at that ratio.

I must say that I would be suprised too with the difference in deflection you obtained. I believe if you used sections that are actually strong enough for that geometry and loading the cross sectional area would be greater and therefore the axial deformations less.