More Art of Diagonal Bracing

[charliealphabravo]

Hi again guys,

Another question related to the configuration of lateral braces and brace connections...

My steel detailing books do not address the connection of diagonals to column webs. I have only found one online example that uses the method I illustrate here...that is using a wide flange tension/compression brace that is the same depth as the column and has a gamma of 90 degrees so that the flanges of brace and column connect directly. I realize that if the lateral loads are light that this type of brace may be oversized and there are also practical clearance considerations for W8 columns.

I'd just like to hear the thought process you guys go through when choosing members sizes/types/configuration for these types of connections. In my case the bracing loads may be light since only the top 14' of the 64' tower is exposed. In that case I would have no problem using a WT or double angle (possibly x-brace) connected directly to a gusset that is welded to the web, with a web stiffener if needed.

Also is it recommended that the braces align with the intersection of the beam/column or can the diagonal be connected directly to the beam with a slight eccentricity which will always avoid the issue of a column web connection? What if I use only a WT brace in the configuration I show (gamma = 90 degrees). Can the brace connect to the outer flange of the column/beam 'only' or should it align with the centerline of the beam/column width.

Also the contractor has indicated a preference for mostly bolted connections so I'm pretty sure the configuration I show will not be practical.

Thanks again.

 

[rb1957]

your attachment opened but was pretty darn useless !?? .ico extension ??

Quando Omni Flunkus Moritati

 

[charliealphabravo]

Oops, sry. That was the last extension I used and forgot to change it.

 

[TLHS]

It really depends on your loads. Here is some random information.

Take your loads and examine what your options are for bracing. I normally size in this order barring other considerations:

1) single angle for low loads with low . (sometimes look into
2) double angle once you're at a size where it becomes practical. It's generally cheaper per unit of resistance than a single angle because of the better buckling resistance.
3) WT if you're at a size where double angles aren't reasonable or they don't fit. WT sections are slightly more difficult to connect
4) HSS for larger loads due to the better buckling capacities. I generally only end up here for large trusses, chevron braces that take gravity loads or seismic applications
5) Wide Flange for some types of trusses, but they're often a pain to get good connection geometry on. They're cheaper on a per weight basis than HSS, but HSS has better buckling characteristics and I've had better luck with connections.

Generally speaking you want to optimize the weight of steel in your brace, but you don't want to do that by picking members or geometry that will require difficult connections.

How you connect will depend on how you're building it. Is it bolt up stick built in the field, is it fully shop assembled or is it modular. I don't know what you're building but it looks like something that you could at least modularize and erect in pieces. In that case, it could be weld up in the shop with some field joints.

Depending on the geometry and construction strategy it sometimes makes sense to aim somewhere other than the point where the lines of action for the other members meet but you have to take the eccentricity into account when you do it.

Is this a project where you're doing the connection design, or is it something the fabricator's engineer will do.

 

[TLHS]

Oh, and whether they're single braces or cross-braces will affect your selection. For double angle (in the normal orientation), HSS and W-Sections you'll end up needing to break one of your members and run plates or otherwise connect one member through the other.

With WT and single angle braces you can easily run them back to back if there's space in your bay and just put a simple spacer plate and a bolt or two at the junction point if you want the extra restraint.

 

[FeSE]

Have you looked at the Uniform Force Method in the AISC manual? (Steel Construction Manual) That will explain how to adjust your theoretical model and connection design if you don't want to connect at the beam column intersection. UFM will also address how you can attach to just the beam. In the case of W8 columns you're not going to want to connect to the column if you can help it. If you have to connect to the column then use an extended gusset plate so that the bolted connection is outboard of the W8 flanges.

If you can put gusset plates on the outside of the brace/beam and gusset then it makes sense to use wide flanged everything of the same size and oriented in the same direction. Otherwise consider turning your brace back to gamma=0, or using double angles.

Cheers.

 

[cliff234]

I like double angle braces when the loads are small enough to allow the use of reasonably sized double angles. Connections for double angle braces are the simplest and most efficient. A gusset plate with bolts in double shear. I only use W shape diagonals when I can use a claw angle connection (angles bolted to the flanges of the brace in single shear and bolted to the gusset plate in double shear). Some of your W shape diagonals as configured in your illustration may be difficult to connect - particularly where they are connecting to the webs of the columns.

 

[charliealphabravo]

Can anyone address the other part of my question in more detail?

How do you configure the brace to beam/column connection when the web of the column faces inward? For a W8 column the clearance is pretty narrow for welding the gusset and I suppose you might need stiffeners somewhere if the load is high.

Thanks again for all the ideas.

 

[cliff234]

In answer to your question about welding a gusset to the web of a W8 column, that can be done, but for the arrangement that you have (all W shape members) how does the brace and the horizontal beam then connect to the gusset? Connections can be developed for the configuration that is shown in your illustration, however (and please consider this as constructive criticism that is being offered by someone who has made countless constructability mistakes in his career) easier (more economical) connections could have been developed if you considered the connection details prior to selecting the member sizes. As engineers we often consider the workflow like this....analyze, design, detail. Only recently did it occur to me that engineers should think like this - detail, design, analyze. First figure out how the pieces can economically be connected (detail). Second, do some elementary manual computations to select the member sizes (design). Third, model the structure (analyze)to refine the loads used in step 2, confirm the selected member sizes and to come up with the "exact" loads for which the connections should be detailed. It appears that your model is a classic case of a framing configuration that was configured without thought to how the pieces would be fitting together -and now you are trying to figure out how the pieces are going to fit together. Better to first figure out how the pieces can be easily fit together - and then do your analysis model. I would have used W12 columns (so that the beams could be bolted to the web with double angle connections). I would have shop welded the gusset plate to the flanges of the beams and I would have used double angle diagonal braces. You are correct. W8 columns present challenges in making connections to the webs. Can you make the columns W12's? Can you make the diagonal braces double angles?

 

[paddingtongreen]

"How do you configure the brace to beam/column connection when the web of the column faces inward? For a W8 column the clearance is pretty narrow for welding the gusset and I suppose you might need stiffeners somewhere if the load is high."


I can't pull this from my memory banks because it is a situation in which I would not find myself. Connections are, or should be, a consideration in member selection. You might put a gusset between the beam and column, cut a slot in the diagonal, and then use clip angles to connect them. With small sizes, this may or may not be possible. The gusset would have to go in with the diagonal, it could not go first.

Bottom line use double angles or turn the diagonal 90[sup]o[/sup].


Michael.
"Science adjusts its views based on what's observed. Faith is the denial of observation so that belief can be preserved." ~ Tim Minchin

 

[charliealphabravo]

Thanks guys.

I guess my question is not clear. The drawing I attached is just an illustration of how to solve the problem without welding to the web of the column. But this can result in oversized brace members since the brace must be the same depth as the column and beam.

I am wondering what the connection looks like when the brace is a double angle or a WT and the web of the column faces the beam. None of my steel detailing manuals (Zayat or Hayward) address this condition. Would a gusset be welded to the web of the column? My understanding is that this can be difficult weld to accomplish when the column is less than a W10 due to clearance.
Cheers.

 

[dhengr]

Charliealphabravo:
A couple thoughts about welding and welded connections....
If your idea that the brace be the same depth as the column or beam sections is so you can weld flange to flange, there isn’t anything real clean or neat about that either. Some of the flgs. will need a bevel prep. for welding, and there is no way to start and stop those welds without a lot of extra work. There is a good chance that those starts and stops will be left as pretty sensitive stress raisers; for lack of fusion, craters, how to terminate without back-up bars or run-off tabs, and clean-up, etc.

As for a rule-of-thumb about welding a gusset pl. to the column web, clearance to make the weld is the issue, but W10 isn’t the magic thumb. You need about a 45̊ line of sight to the welds from the inside tips of the flanges, so you can get a welding gun in there to make the weld. But, this is subject to the exact welding equipment being used, and varies a bit. Ask your fabricator what he can reasonably do with his equipment, and bag of tricks. Thus, those welds could probably be made on a W8 with a 6.5" flg. width, since d/2 (half depth) is about equal to bf/2 (half the flg. width); but a W10x60 might be tougher to do, since you are reaching deeper. Obviously, mat’l. thicknesses and exact gusset location must be included in these calcs.

 

[cliff234]

charliealphabravo:
Eliminate all thought of welding the gusset to the column web. That is generally not the way it's done. The gusset plate wants to come out of the shop welded to the horizontal beams. The gusset plate would have a single or double angle connection attached to the vertical edge of the gusset that then field bolts to the web (or flange) of the column.

The connection is virtually impossible to do if you weld the gusset to the column web. If you do that, then how are you going to attach the beam to the gusset? Does the gusset get shipped from the shop welded onto the columns? That is generally not done - at least in my experience.

 

[charliealphabravo]

Thank you cliff, that's exactly what I needed to know.

 

[paddingtongreen]

To be fair, I have seen the gussets shop welded to the web of the column but only when the the column flanges were so wide that the full length beams could not be juggled horizontally or nearly horizontally into place. Those connections were one sided with the flanges stripped on one side of the beam so that they could be lowered between the column flanges from the top of that section.

These were on 660 and 800 MW boiler house structures.

Michael.
"Science adjusts its views based on what's observed. Faith is the denial of observation so that belief can be preserved." ~ Tim Minchin

 

[TLHS]

I've also seen some seated connections come full attached to the column complete with brace gussets so that they can erect the columns and then just drop the beam on the seat.

You'll also weld up the gusset if it's something you can ship assembled or partially assembled.