Gusset plate with large hole 1

[TehMightyEngineer]

So, here's an odd question. Working on a design-build turbine and generator replacement project. Client needed to move some cross-bracing up slightly for clearance. Modified the bracing and columns for the adjusted working points. There was a fire suppression valve that went through the new brace location but I was assured that the valve could be rotated to miss. Turns out that was wrong, the valve can't be touched.

So, we're past the point of no return and the new braces have to go in. The plan is to construct an over-sized gusset plate out of two halves with a 10" diameter hole when the halves are CJP welded together. This will allow the valve stem to penetrate the gusset. All my bolting connections, work points, welds, and such remain the same as my originally designed gusset, it just now got very fat in the middle to keep the cross section equivalent to the original gusset without the hole. Brace is tension only.

Other than tension failure of the gusset at the hole and making sure I have enough distance between the ends of the gusset and the hole (Whitmore section) I can't think of anything else to check. There's gotta be something I'm missing, it seems like having a 10" hole in a gusset plate should be a much bigger deal.

Maine EIT, Civil/Structural.

 

[jayrod12]

it's relative. if your gusset is 24" then it makes sense that you could work around it. if your plate was only 14", then I'd be a lot more worried.

 

[TehMightyEngineer]

Oh, good point. Yes, the gusset plate is large relative to the hole. Gusset plate was sized for 110 kip load and was 2'6" on the vertical dimension, 2'1" on the horizontal.

I've attached a picture showing the original and a rough sketch of the new gusset plate geometry.

Maine EIT, Civil/Structural.

 

[mike20793]

Is the gusset plate at a beam-column connection or just at the column? I seem to remember a stress distribution lab from school where we pulled on a plate with a large hole from both sides to check the stress distribution around the hole. I can't recall the results off the top of my head, but I would double check the stress distribution in the plate. A quick internet search should give you some references. I wonder if the hole will affect how the load is distributed to the column and beam (if one is present). Does the uniform force method still apply? Will a moment develop at the connection? I feel like FEA modeling may be appropriate here.

 

[mike20793]

D'oh, I posted as you were posting your sketch...

 

[TehMightyEngineer]

In order of your questions Mike:

Column only. This is at the baseplate (see previous attached picture).

As best I can tell the uniform force method will apply for a case similar to special case 3 in the manual. The brace, column, and gusset to column weld should see the exact same forces as they did in my non-doughnut gusset plate.

A moment shouldn't develop at the connections as the centeroid of the weld and bolt group still pass through the working point axis of the brace. You did remind me that some small moment will develop in the middle of the gusset as the net cross-section is no longer aligned to the working axis of the brace but I imagine this will be negligible.

Yeah, I can quickly make this in a FEA program.

Maine EIT, Civil/Structural.

 

[mike20793]

I guess if a beam were present, I would be worried about the stress distribution through the plate and how it distributes the forces to the beam and column and the applicability of the uniform force method. Since it's just attaching to the column, I can see how the forces would be the same. I guess it's really just a matter of stress distribution around the hole in the plate, unless someone else thinks of a potential problem.

Just out of curiosity, what program do you use for FEA?

 

[TehMightyEngineer]

STAAD.Pro 2007. I know it's not a true FEA program and has a lot of limitations but for the work we do it's sufficient (but we do push it's boundaries sometimes).

Looking to get my boss to switch over to RISA-3D. Probably our biggest sticking point is RISA appears to be even less of a FEA program (que JoshPlum from RISA telling me it's the best).

Maine EIT, Civil/Structural.

 

[mike20793]

I like STAAD.Pro; I used it frequently in my thesis research. We have had a few engineers do some interesting stuff in RISA. My firm has been debating getting a license of SAP2000 for situations like this, but we just can't seem to justify the cost when we already have the RISA suite, RAM suite, and ETABS/SAFE.

 

[KootK]

I don't love this but I can't think of anything better. Tension only and low seismic, right? Why is your work point up so high? Delivering lateral to slab on grade? I only ask because, if you could move the work point down, it would open up some more attractive options.

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

 

[SAIL3]

I would tend to make the gusset pl symmetrical(along centerline of load) @ the hole , at least....another area I would investigate... is there enough length in the gusset pl fore & aft of the hole to get a reasonable stress flow around the hole...if not, then the majority of the load may stay in the outer regions of the gusset pl and check if this is a problem or not.....there is a limit of relative hole size to width before this smooth stress flow around the hole begins to break down..unfortunately I don't know what that limit is without doing some research....

 

[KootK]

Considering the geometry, I wonder if you could borrow some of the design procedures used for eye bars.

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

 

[TehMightyEngineer]

KootK: Tension only and low seismic. Don't love it either and fully expect any engineer who sees this gusset to go "what the hell?" We're modifying the existing work point up to make room under the cross bracing for a means of egress doorway and material passage for the replacement turbine and generator. Yes, my lateral loads are transferred to a slab on grade. My brace forces don't increase much from the geometry. Mostly all I've done is add a large weak-axis moment to the building columns from the work point moving. Thankfully these are heavy industrial columns and required only minimal reinforcement for the changed load path. I could raise my workpoint even more but that would require reinforcement of my already reinforced columns.

SAIL3: There is enough length for the typical 30 degree Whitmore section stress flow. I would imagine a 10" diameter hole in a 3 foot wide plate would be relatively small of a stress riser but you are correct. We did add a little fat because of this.

Maine EIT, Civil/Structural.

 

[TehMightyEngineer]

Oh! Good call KootK! Star for you.

Maine EIT, Civil/Structural.

 

[TehMightyEngineer]

Looking over the section some of the limitations are exceeded but I can probably make some assumptions, as these rules are written to transfer loads to a pin in bearing, that I can make some reasonable adjustments.

Maine EIT, Civil/Structural.

 

[KootK]

Are the braces four angles cruciform? Is it to late I change that?

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

 

[TehMightyEngineer]

Single WT 6x15, originally was A36, we upgraded it to A992 and kept the same size for the slightly larger brace force from the new work point.

Maine EIT, Civil/Structural.

 

[KootK]

The side plates might be better off as WT for stiffness and nominal compression capacity. Keep an eye on shear lag. Should be adaptable to WT bracing.

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

 

[TehMightyEngineer]

Interesting idea. I could probably carve out part of the web of a W24 beam or something and make a shape that would be pretty much what your sketch showed and should mate up easily to my tee and fit around the pipe with the flanges of the beam in place of the fitting plates. It would also have the advantage of being 50 ksi steel as opposed to 36 ksi.

In this job speed is worth more than money by a large amount. For example when we told the client would be expensive to do the bracing modification it was almost like they handed us a blank check and said "do what you need to do, just make it happen."

For a little background, this turbine/generator will allow the client to run 100% off grid and they need it in by January 1st. All was well until the condenser they purchased became unavailable and everyone had to do last minute redesigns to make a new condenser work in addition to all the work still remaining on the turbine and generator.

Maine EIT, Civil/Structural.

 

[SAIL3]

Koot's concept is a good alternative if not a bit busy.....another option may be to flip the gusset pl 90 degrees and lign it up with the base pl work-point as it looks like it would clear the valve....I would be concerned with the 1'-5" off the base pl...if this is causing a moment in the col then I would expect a portion of that moment to find it's way into the base pl/anchor bolt area and cause a potential problem there...there is also the option of using a tension brace other than the WT that might faciltate a more reasonable gusset pl connection....