Base plate design ignores corners of base plate?

[AggieYank]

The cantilever method for rigid base plate design per AISC seemingly ignores the areas of the base plate that aren't directly perpendicular to the column.

In other words, there is no check for the corners of the steel base plates.

I'm sure this is fine for a small base plate, but what about for a relatively large base plate, where the area of the corners of the base plate is more than the area that is directly perpendicular to the column?

Is there a rationale for ignoring the corners of the base plates?

 

[nutte]

Prior successful performance? "We never worried about it, and we never had a problem with it?" I'm not sure, so I'll let others chime in.

 

[csd72]

The line across the diagonals is usually wider than the width of the base plate. i.e. it is a greater effective width.

Also check that your base plate dimensions are not outside the recommended limits (I believe there are some somewhere)

 

[akastud]

The rigid baseplate analysis always has bothered me. RISAbase actually does a finite analysis of baseplates which gives really interesting (but sensible) contours of loads. It is interesting on the side where the baseplate is in an uplift condition, the corners past the anchor bolts go back into compression because of the stiffness at the bolt. Good program but takes some time when analyzing.

akastud

 

[Lion06]

akastud-
wouldn't you expect the basplate to back into compression on the side of uplieft on the far sde of the anchor bolts? That is where the prying forces come from.

 

[271828]

You're welcome to check that pattern. It shouldn't control because that yield line is so long and because the resultant upward force is only about 1/3 from the corner of the column to the corner of the plate.

 

[AggieYank]

271828, can you explain that?

The case I'm asking about is for axial compression only. There would be no uplift due to bending / tension in the base plate. The upward force on the base plate would be the allowable bearing pressure, and would be uniform under the base plate.


csd72, I couldn't find any limits on base plate dimensions, either generic, or related to a design theory. My guess is it's because base plates are sort of in between steel design and concrete design, and so neither ACI or AISC like to go into too much detail on them.

 

[271828]

Sure.

The proposed yield line cuts diagonally across the base plate near the flange corner. The part of the plate outside the yield line is the triangular (or close to triangular part) part near the corner.

If the pressure is uniform over this triangular part, then the resultant force is about 1/3 the way from the yield line (which passes through flange corner, or close to it) to the corner of the plate. This is more favorable than if the resultant was halfway between the yield line and the edge of plate, which is the case for the typical patterns shown in the AISC Manual.

 

[PEinc]

AggieYank, why are you worrying about this? Steel design manuals and the AISC manuals give the design method for column base plates. Why not just use what has been successfully used for years?

 

[AggieYank]

271828, correct me if I'm wrong, but doesn't the n', the yield line theory cantilever distance, refer to the distance inside the flanges, and not account for anything outside the flanges? I didn't quite follow your explanation.

PEInc, I'm trying to design a base plate for a temporary steel structure that sits on soil. The base plate is relatively large compared to if it was sitting on concrete. I don't think the traditional rigid plate cantilever method property accounts for large areas of the corners of the plate, but couldn't find a design method which addressed this issue.

Right now, I'm planning on designing based on the cantilever method, and increasing the thickness some nominal amount.

 

[Lion06]

Aggieyank-
I see your question. The formula for the thickness of the baseplate does it on a per inch basis (along the length/width of the plate). It basically assumes that every inch of B has a support inch at the critical section line.
I believe Aggieyank's question is how is this possible since the column is providing the support for the cantilever baseplate and the column does not extend all the way along the creitical section line (along either the B or N dimension). Basicaly, should the moment capacity of the plate be determined based on the acutal column dimension and the required moment capacity be determined based on the entire trib area, instead of assuming it on a per inch basis since there is a distance (B-bf) thast has no cantilever support.
Does this represent your question Aggieyank?

 

[AggieYank]

Essentially yes StructuralEIT. For typical base plate design, the "B-bf" distance is typically not a "big" distance, and so it's probably fine to ignore it. However, for a relatively "big" distance, it seems like you may want to have some kind of check for it.

It'd be extremely complicated to analyze, as it would involve two-way bending of the plate in the corners, and I'm sure the analysis would be wrong anyway. A finite element analysis in a computer program would be ballpark, but still wouldn't be perfect, because as soon as the center of the baseplate moves down, a little more load will concentrate there as opposed to the edges.

One solution I've shied away from is adding stiffeners from the edges of columns to the edge of plate, and using the distance from edge of plate to stiffener as my cantilever distance. I'd like to not add the stiffeners as it would involve field welding.

 

[IFRs]

Use a round base plate and use Roark's to find an equation for the stress?

 

[Lion06]

why not draw a trapezoidal shape in plan (one edge extending along the column flange, two edges from the corners of the columns to the corners of the baseplate and the final edge as the edge of the baseplate). Find the resultant cantilevered moment at the prescribed critical section, but only assume you have a length of baseplate = bf to resist this moment and determine the required thickness that way? That seems reasonable to me.

 

[Lion06]

Boy, now that I think about it, wouldn't you have that same issue with a typical spread footing design? You design for moment on a per foot basis, but there is really only support at the column not along the entire footing length .
Can anyone comment on this?

 

[PEinc]

I wouldn't use a large, round, steel plate to support a temporary column on dirt. If the load is large and you need a large plate, you could use a horizontal distribution beam under the column to spread the load across the plate. Then the plate could be designed for one-way bending. Why don't you install some timber cribbing under the column with a much smaller base plate? Or, pour a temporary concrete pad.

 

[AggieYank]

StructuralEIT, your trapezoid idea is exactly what I tried once I decided that the traditional cantilever method wasn't a great fit for this application. Doing this results in a significantly larger base plate, as you'd guess.

Given that test results have proven traditional cantilever method for steel on concrete as conservative, I think the best result in my application lies somewhere between the "trapezoid" cantilever method and the traditional cantilever method.

It'd still be nice if there was a reference on this.

 

[271828]

AggieYank, as you typed, the n' length doesn't apply outside the flanges.

What's the size of this column and base plate? Just trying to get a handle on just HOW far outside of normal it is.

AggieYank and StrlEIT, the answer lies in yield line analysis. The trapezoidal shape is only possible if several other yield lines form. If there's ever a doubt about a particular weird case, it should be checked using YLA, which is really pretty easy.

 

[JStephen]

"The formula for the thickness of the baseplate does it on a per inch basis (along the length/width of the plate). It basically assumes that every inch of B has a support inch at the critical section line."

"Boy, now that I think about it, wouldn't you have that same issue with a typical spread footing design?"

I think in both cases, the assumption is not that the plate is supported at every point along the bend line, but that if it fails, it will fail by hinging in a line; consequently, one part can't hinge up unless the whole thing does. You may have localized bending stresses that are higher than the average, and that is not prohibited. This approach is implied in ACI-318- 15.4.1 and 15.4.4.1.

 

[darkwing888]

I'd build an FE model of it and check nominal stresses (1 hour - done - next).