Bracing for steel platform design

[mfstructural]

Hey everyone, I'm designing some platforms for an industrial plant and have a couple questions. I've attached a picture of the platform in STAAD. The platform has a monorail (for a 2.2k load) at the top (36' high) (the longitudinal beam along the center all the way at the top). At the middle elevation a hopper (13.3 kips full) is being supported by the framing. The floor will be diamond plating at this elevation because bags of powder will be stored on the floor and this way powder can fall down below. My question is also about bracing for a frame like this. Since there is not diaphragm at the top does each side of the frame have to be braced? I currently have some bracing to start analysis with because I want to see how it behaves. It's essentially self weight and seismic effect of full hoppers right now.
Thanks

 

[JAE]

Just a few comments on the graphic.

If you are in a substantial seismic area - the knee braces and diagonal braces ending near mid-height of the columns is probably not a good idea.

Without a diaphragm you either have to brace all sides or add a horizontal system of bracing to establish a stable framework.

Not sure how STAAD works these days graphically (I hated it back in the 1980's)....but your X bracing doesn't appear to be pinned at the ends of the diagonals.


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

 

[mfstructural]

I'm in SDC B. I've designed for R=1 (Cs=.316250) to design to AISC 360. That was another question I had there is another platform that is narrow in both directions. 10'x8' and 35' tall. Their preliminary drawings show these little kickers and that's not going to be enough. I modeled it just to see what happens and the deflection is .6" at the top in one horizontal direction. SRSS of two horizontal directions you get about 1", but you can definitely reduce moments and member sizes by using full bracing. I haven't modeled any horizontal braces yet to see how it behaves. I'm still getting input on where the braces can be installed. The bracing members are pinned I just have them defined as truss members in STAAD so they can only resist axial loads.

 

[mfstructural]

I have another question regarding these platforms and what would be considered overkill. I don't particularly like the idea of having kickers at all the corners so I put in full bracing. The platform attached is 19' x 3.5' wide (excluding landing, 6' total at landing), and 9.5' high. I was getting a displacement of 4" at the right side of the platform under DL and LL because of the load at the landing. I added these braces and am now getting a resultant disp of .04" under seismic loading. I know this is very small and I can probably get away with even smaller braces (L3x3's now), but there is a line between being economical and being overdesigned. Would you consider this bracing as overdesigned? It's an industrial plant application so I prefer to be a little more conservative, but even the W8x28 beams I used have interaction coefficients of less than 0.1, some even 0.01 to 0.05.

 

[KootK]

L3x3 sounds fine to me. Sticking to a slenderness ratio less than 300 usually leads to reasonable sizes. For indoor applications, I often go with back to back 6" single channels. Way overkill but:

1) Commonly available.
2) I'll use the same shape everywhere to simplify the mill order.
3) Vertically stiff so no issues with sag/draw.
4) The two side by side bolts that can be installed at the ends of the channels are usually enough. Nice compact gussets.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[mfstructural]

That sounds reasonable. I thought about the sag with the angles, so channels would definitely help that.

 

[KootK]

The only thing that I dislike about it is that I sometimes end up with braces that are wider than my columns. It bothers me aesthetically. In those cases, sometimes I'll go down to a C3 or C4 and one line of bolts.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[mfstructural]

Hey all, back with an update. I have some questions on the amount of bracing I'm using here. I've attached a picture of this platform. It has a tank on the second level (can't see it here) and also a monorail along the top. Since there is no diaphragm to transfer the loading, I have horizontal braces at intermediate framing and at the top to keep deflections down. The deflection of this platform is 0.7" under seismic loading (SDC B). Just as a side note, I could not get channels to work, they were failing. But Double angles did work. Just wanted to get your opinion on the bracing configuration. I'm also under constraints on where I can have bracing on the first level. I know in the model there are interferences with horizontal bracing and framing, but the intermediate framing will be shallower so the brace can pass below.

 

[KootK]

Look as though you've pretty much got it covered mfstructural. Like JAE, the mid-height column bracing makes me a bit nervous. However, if it works, it works. It sounds as though your loads are pretty heavy so make sure that you've properly accounted for P-delta effects and such.

If it were my design, I would eliminate:

1) Six of the eight knee braces for the very lowest platform. For most of the lateral degrees of freedom in that platform, I'd let the platform ride along with the two columns that go full height.

2) The knee brace in the foreground right above the lowest platform.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[mfstructural]

Thanks, I put that foreground brace in to reduce deflections more than anything else. My deflection is .7" with that brace and goes up to about 3" without it under seismic loading. I don't like the mid height column bracing either but I can't have full bracing there because of equipment/walkway interferences.

 

[KootK]

If the foreground brace makes that big a difference then definitely keep it. Heck, you might be able to get rid of the 1/2 height brace on the same line.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[mfstructural]

I'm trying different things with the two midheight braces and the kicker in the foreground. I like the kicker because it takes that mid height brace load back to the horizontal bracing. It seemed like there was not much in between. If I removed the bracing I had a high deflection, and if I kept it it was very low. And yes, the loading is pretty high. The tank at mid height is about 12 kips plus the platform will be loaded with 2.2k sacks of a material for manufacturing. They want to use diamond plates, so obviously I'll have shorter spanning beams for plate spanning, but I'm not modeling those. There is also a monorail running along the length which will be bolted to the beams in the short direction. max load for monrail is a 2.2k bag. I'll let you know if I get a configuration that has low deflection and no mid height bracing.

 

[rjenyoy12]

Regarding the orientation of the column, is it not possible to rotate it 90 degrees so that the connection of the girder carrying the loads of the equipment and hoppers could be fixed to the column flange using bolted end-plate, thus, eliminating the knee braces?

Why not make the Z-axis (transverse) as moment side and X-axis (longitudinal) braced side? Making this will also give the monorail crane load transfer to the column flange by moment connection which is same as for the equipment below. It is easier to have moment connection at the column flanges.

Also, in lieu of the mid height bracing, along X-axis, could you not make it a Chevron brace?

 

[mfstructural]

rjenyoy: Yes I will probably rotate the columns to have strong axis perpendicular to transverse direction. You may still need kickers in the other direction if seismic loads are high enough though no?

I thought about the moment connections in the short direction. Especially with the limitations on bracing. I considered the cost of moment connection vs. a tab connection and kicker and thought the gussets for the kickers would be shop welded and you just bolt up the kicker, but there are several of them. But with all the restrictions coming into play, the moment connection might be the best. Put moment connections at the two ends of the frame (or at least the one end where the fork lift pulls in) in the short direction at all levels and transfer load via horizontal braces.

I wanted to make this bracing a Chevron, but they need clearance to pull in with a forklift to load the sack (along short direction). The forklift pulls in through the front at the short direction and there is a hatch that opens and monorail lowers to pick up bag. As for the long direction, I also wanted to use a Chevron (and am still fighting for this because you need a lateral resisting system) but apparently they have a rule that if there is bracing in a bay no one can walk through that bay. Doesn't entirely make sense to me since a Chevron would give you plenty of space to walk through.

 

[KootK]

but apparently they have a rule that if there is bracing in a bay no one can walk through that bay. Doesn't entirely make sense to me since a Chevron would give you plenty of space to walk through.

Safety. Workers would bump their heads on the bracing unless you provide a system to guide them appropriately.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[mfstructural]

Yes, I think it's plant specific because we've done platforms like this before where this was not mentioned. I understand that aspect and they would have to have yellow tape and they want to avoid that. So that's why I went with the mid height bracing. There would be little area to access the stairs that are underneath because the other bay has a wall with a door opening right into it. So I may be stuck with the mid height bracing if the client doesn't want to have to secure the area for safety.

 

[rjenyoy12]

Yes, you need a bracing on the other direction after making the transverse girders of the bottom platform moment connection. You said that X & Chevron bracings are not allowed below the bottom equipment platform then you may use the knee braces on the other remaining direction of the platform. However, check requirement of AISC 341 Article C11 if it is applicable in your area.

Or alternative solution to eliminate the remaining four (4) knee braces is by using again a moment connection especially the girder near the equipment.

In the case of the mid-height bracing, if it is not possible to use Chevron as you said, then just make sure your mid-height it works in your analysis.

 

[mfstructural]

Ok, thanks I'm looking into this. Also, as a sidenote wanted to get some people's ideas on loading on the beams. The hoppers above are filled by ripping the string open on the bags hanging from the monorails. The powder then falls into the hoppers from above. I'm debating on using a multiplier for the dynamic load effect of the powder falling. We've used anywhere from 1.5 to 2.0 in the past, thinking 1.5 here.