Compressive buckling on a plate intended for bracing.

[Leftwow]

Good morning all,

I was wondering you opinions in this subject. I have designed an eccentrically braced inverted U braced frame on a structure. The controlling factor means of failure was the KL/R of the whitemore section in compression. I had 50 kips of compressive force on my braces. The whitemore section is displayed in AISC Fig. 9-1. In order to reduce my L factor we welded 2 transverse stiffeners on the braced plate to the column and to the braced plate. I used the distance from the transverse stiffener to the top bolt as my L. Do you guys feel that this was an effective way to reduce my buckling factor, and have you ever seen this done?

 

[KootK]

Could you supply a sketch of your detail? It's not clear to me how you've set up the stiffeners. Also, is this a seismic application where you'd need to worry about plastic hinging? It sounds promising but I'd need a clear picture of the situation in order to comment definitively.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Leftwow]

Yea Check this one, IT IS NOT TO SCALE.

 

[Leftwow]

Another thing, the work point went down the the baseplate, so the design changed a little since that sketch.... instead of 6 inches above

 

[KootK]

I agree that the low stiffeners would reduce the buckling length effectively. It strikes me as an expensive solution though. I would:

1) Get that joint a lot more compact, starting with the work point modification that you mentioned and;

2) Consider a thicker gusset. It would look insane but, up to about 2" thickness, that would probably be cheaper than all of the stiffening.

You also didn't answer my question about seismic. If there needs to be a plastic hinge here, that changes things.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[JAE]

Agree with KootK - that detail is bad for seismic.



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[Leftwow]

Job site in south Louisiana so seismic is never a controlling factor. Yea we got it more compact but the ratios were still high, interesting though that a thicker plate would be cheaper. Kootk, if I get a thicker plate for this, say the 1 to 2 inch thickness, will they need to heat stress the plates while installing?

 

[KootK]

Kootk, if I get a thicker plate for this, say the 1 to 2 inch thickness, will they need to heat stress the plates while installing?

I couldn't say. Might be worth calling a local fab shop. Although, we've got some good fab guys here that may well chime in. What plate thickness have you been considering? While I stand by the 2" comment in principal, I was really hoping that it would be more like the difference between 1/2" and 3/4"-1". Might be kinda weird looking to gap your channels a full 2".

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Leftwow]

Well, yea I read something a while back about welding to thick plates that it needed to be preheated and maintained while welding. So we avoided that lol. Also that is true that gapping the MC channels may affect the channel sizes. But it looks weird anyway

 

[dozer]

Leftwow, are you sure you are computing the unbraced length correctly? Seems like a plate of reasonable thickness should be able to handle 50 kips. I've attached a sketch of what length for checking buckling should be, assuming no stiffeners. Also, what value did you use for K? Oh, btw, fillet weld symbols should always point to the right.

 

[sbisteel]

It depends on the weld process and base material, but generally preheat would be required for plate thicknesses over 1.5".

 

[Leftwow]

I used 0.6 for k. My L I have 12 inches. The AISC guide told me L was from first bolt to edge of column. That would reduce my length by maybe an inch or two. you see the angle is a steeper than it looks in the drawing. Then I used equations AISC EQ. E3-4, AISC EQ. E3-2 AND EQUATION E4-1. My kL/r ratio was 39.9, however that isn't an applicable sketch for what I have in there because we changed it while the engineering sketch simply just held the concept for the design.

 

[SAIL3]

not sure how the horiz stiff help....use thicker pl ,as Koot pointed out, or use edge stiff on vert edge....also, if possible, reduce the size of base pl....

 

[Leftwow]

Well the general consensus of the engineers was that I should have increased the size of plate rather than added horizontal stiffeners. Thanks all

 

[dozer]

Leftwow, assuming ASD and Fy=36 ksi, the allowable compressive stress for the gusset is about 19.8 ksi (Table 4-22, kL/r=40). I figured out your thickness based on the info you gave is 5/8". I don't know your geometry but just figuring lw (See Fig 9-1 that you referred to) is 6 inches, the allowable load would be:

Pall = 19.8 x 0.625 x 6 = 74.25 kips

I laid out a geometry that the brace was 70 deg from the horizontal and the unbraced length of the gusset was 12" and I got lw=8.7", so I think the 6" I used is quite reasonable. So, like I said, I would really be surprised if the plate couldn't take 50 kips without stiffeners.

 

[Leftwow]

Yea 6 inches was right for the lw, Pn is the critical buckling stress*Area of whitemore section. So my critical buckling stress ended up being 33.1049 ksi. Aw =6*0.625, then Pn=(33.1049*6*.625)/2. Looking back I see that my safety factor that I used should have been 1.67, and not two, then we would have the same answer. THANK YOU DOZER!! At 22 inches of L I get around 61 kips of resisting force.

 

[SAIL3]

not sure if the whitemore section is entirely appropriate in this application....the load in the gusset pl will tend to accumulate @ the stiffest portion of the conn....
order of stiffeness:
base pl area of gusset
top portion of gusset @ col fla
web of col
I would expect more of the axial load in the gusset to migrate towards the free edge of the gusset pl and would check it's unsupported length as a precaution.....

 

[Leftwow]

Basically I just followed Design Guide 29. But yes I did consider those situations.