Compressive Bending of a Base Plate.. How? 6

[MegaStructures]

Question not for a current design, just interested.

I'm going through AISC DG1 - Base Plate and Anchor Design and noticed that the guide recommends designing the base plate for bending when a compression load is applied to the column. How in the world would a base plate bend from an axial compressive force if it is continuously supported by the concrete underneath it? Is this to account for voids in the concrete? I get that the DG treats the bearing force as uniform across the bottom of the base plate and this force profile would cause "upwards" bending of the base plate if it were able to "travel", but the concrete cannot act on the base plate through a distance.

“Any idiot can build a bridge that stands, but it takes an engineer to build a bridge that barely stands.”

 

[r13]

If you really want to dig into it, you can perform a FEM analysis to see how it goes. The model should consist a steel plate supported on compression springs (using concrete property for spring constant), and loaded by line loads derived from distributing the concentrated load to the flanges and web by strain compatibility method (essentially A_flg and A_web). Let us know if you find anything interesting.

 

[KootK]

Let us know if you find anything interesting.

I did something similar with the software tool RISA BASE back in 2003. The conclusion that I came to is that even very, very thick base plates won't get you anywhere close to a uniform base plate pressure with an elastic model.

In my opinion, base plate success has gotta be some combination of:

1) Local crushing and plastic redistribution.

2) Unaccounted for improvements accruing from thing such as:

a) Concrete in the hot spots being confined by the surrounding concrete to generate extra strength.

b) Competent, high strength grout pads distributing the load a bit where such pads exist.

3) Calculated loads not coming to pass.

A couple of interesting, related things that I've seen come to pass in the field:

4) A 1/4" thick warehouse column base plate that was literally curled up and exposing air gap, just as elastic theory would predict.

5) An 8" wide flange column on a narrow wall with the flanges run parallel to the wall and the concrete below the flanges spalled off, just as elastic theory would predict.

I tend to view conventional base plate design procedures as:

6) Round about, and imperfect, methods for ascertaining reasonable base plate stiffness and;

7) expedient means of arriving at a solution that satisfies equilibrium.

 

[r13]

Yeah, I guess the above points should satisfy the OP's concern. I wasn't bold enough to provide a plate less than 1/2"

 

[dik]

The concrete is a lot softer now than it was in the good old days...

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[dik]

@KootK... I think your items 1 and 3 pretty much sums it up...


That's why I don't use anything less than 3/8"

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[KootK]

Heck, I've never used less than 3/4" on anything actually supporting roof or floor. Just no profit in that from the EOR perspective. Now, if you're the steel supplier, there might enough savings across a good sized project to buy one of your kids a used compact car for college...

 

[dold]

I made a RISA3D model using solids quite a few months ago to investigate how a thick shear lug and thick baseplate interact (how shear lug moment influenced plate bending between WF column flanges) under gravity and shear force. It was a curiosity model, not a production model. I wasn't looking at OP's problem exactly but I found similar results as to what Koot describes. Using solids obviously provided a more realistic result than using plate elements (not sure what Koot used in [Koot, 2003] study), but you essentially end up seeing a projection of the column section in the pressure distribution during gravity only cases. I did not consider r13's suggestion about strain compatibility. Also I may have not used springs as a support now that I think about it. That would of course make a noticeable difference in the reactions. It was a pretty fine mesh as I recall, maybe 0.5" cubes but I'll check. Springs might provide for a nice long analysis. Or maybe I only modeled supports at anchorbolt holes for an uplift case or something...?

At any rate. I will dig it up and post some results tomorrow perhaps if I have it in me.

Happy new year everyone!

 

[dik]

For light loaded columns, you can use 1/2 that... for most columns I generally use 1/2" min...

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[r13]

MegaStructures,

why base plate bending checks are required..

Bending is required to determine base plate thickness. See derivation of equation below.

 

[MegaStructures]

KootK very complete answer as usual. My interpretation of the responses is that base plate bending occurs as a result of column axial compression due to

1) Errors in construction such as voids in grout, warped plate, etc.
2) Crushing of grout under the center of the base plate

Basically the plate is not bearing perfectly from the moment it is constructed, or fails the relatively weak grout/concrete to make room for some plate bending behavior.

A 1/4" thick warehouse column base plate that was literally curled up and exposing air gap, just as elastic theory would predict

Interesting, so here enough of those initial imperfections and concrete crushing built up to allow the middle of the base plate to deform downward and "curl" the edge up? I saw in another post that you made an FEM of this situation. Did you bother with non-linear material properties of the concrete springs?


“Any idiot can build a bridge that stands, but it takes an engineer to build a bridge that barely stands.”

 

[HTURKAK]

AISC design approach of column bases involves an elastic analysis based on the assumption that the plain sections remain plane.With this approach, by solving equilibrium equations,with linear stress distribution, the maximum concrete stress and the and the tension force at anchor rods can be determined. This procedure ignores the flexibility of the base plate .

However, this procedure ,IMO, satisfactory and i did not hear base plate bending failure due to concrete compression .

How in the world would a base plate bend from an axial compressive force if it is continuously supported by the concrete underneath it?[/b] Is this to account for voids in the concrete? .....]

I fully agree with your questioning the approach provided that the base plate continuously supported by grout and levelling nuts are not used under the base plate for adjustment purposes.

Nowadays, the use of levelling nuts very famous and contactors declining the use of self levelling epoxy grout in large scale projects...


The EC- 3 approach is similar to your thought. The following picture depicts the T-stub concept and stress distribution under the base plate.. Another criteria is,the bearing resistance of concrete is about 6 times higher than compression resistance of the concrete and for a square base plate is 5fcd is used.

P.S. Will you please guess the size of anchor bolt and base plate thck?

 

[MegaStructures]

That picture describes things perfectly HTURKAK! Star for you.

Put me down for 1.5" anchors and 3" on the base plate. Supporting an overhead crane?

“Any idiot can build a bridge that stands, but it takes an engineer to build a bridge that barely stands.”

 

[dik]

maybe a pipe rack or overhead conveyor...

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[HTURKAK]

Yes ..supporting O/H crane.. but regarding your guess, see the picture below to compare the sizes..

 

[r13]

I fully agree with your questioning the approach provided that....

This procedure ignores the flexibility of the base plate.

Obviously you didn't follow through the previous discussions, noting the considerations and conservatism in develop this method. I think ACI assumes the column is infinite rigid, while flexibility of the plate varies. I opine, The sketch below illustrate this school of thinking. Note that, for typical design, case 2 is the most optimum, while case 3 is overly conservative; thus case 3 assumption is conservatively adopted, with its mind set in material and construction imperfections, and combined with a realistic base pressure. The base pressure is adjusted to capture the moment would otherwise occur in a manner similar to the illustration in cases 1 and 2, due to variation in stress/strain. I hope this makes sense, and correct me for mistakes.

 

[HTURKAK]

Honestly i did not look the previous discussions ..but with your warning i screened them. Will you please tell me what posts i should read in detail ?. I assume you mean AISC . ACI does address anchor to concrete issues not the base plate thickness.

The OP has a valid point and his question ( AISC DG1 - Base Plate and Anchor Design and noticed that the guide recommends designing the base plate for bending when a compression load is applied to the column. How in the world would a base plate bend from an axial compressive force if it is continuously supported by the concrete underneath it? )

If the base plate is continuously supported by the grout and levelling nuts are not used under the base plate for adjustment purposes, he is fully right. However, in practice, pockets may develop under the base plate due to poor workmanship or the type of grout and for the use of levelling nuts which is poisoning the behavior.These may not be reasonable concerns for simple architectural bldgs.However , still concern for industrial plants and towers. Having involved in past for industrial plants , i remember one project which i involved as DE for the design of shipyard hall for two decades ago , having columns 24.0m free ht with base plate sizes 3200mm L X 1200mm W and anchor rod sizes 80,100 mm .

EC-3 approach based on T-STUB concept ,( tested with a lot of experiments ) and assumes effective area having perimeter around the column with a distance 'c' which depends on the strength of grout and thickness of base plate. That is, one can not obtain fully uniform bearing stress under the base plate by increasing the thickness within practical sizes.


On the other hand, AISC assumes uniform bearing stress distribution for concentric loading and accepts triangular bearing stress approach or uniform distribution of compressive bearing stress for eccentric loading.

IMHO, i will still prefer to check plate thickness with the AISC method for bearing stress or assuming the anchor rods at compression side will resist to compression forces....

Regarding your sketches, are you proposing to adjust the thickness to get an intermediate bearing stress distribution?

 

[r13]

HTURKAK,

My diagrams above encompass all forms of stress distribution under base plate for the concentrically loaded column, and there is always a bending in the plate outside of the perimeter of the column due to cantilever action, assuming the column is much rigid than the plate and the concrete. While all plates may pass bending check, but the magnitude of deflection varies. The thin plate in case 1 could deflect an amount that is unacceptable, and deemed it has "failed"; the third case is overly conservative (true uniform stress distribution and near zero deflection), unless necessary. Therefore, AISC adopted the second case with the concentrate load conservatively distributed to the plate uniformly, which will encompass all stress distribution variations between cases 2 and 3, yet yield satisfactory result (moderately flexible). So the claim on AISC has ignored plate flexibility in the formation of this simplified method, IMO, is false. Hope this helps in clearing up the confusions.

I think the OP's question/concerns was adequately addressed by these quotes below, theoretically and practically.

dauwerda (Structural)31 Dec 20 15:01
If you are going to assume a rigid base plate that evenly distributes the bearing stress you need to ensure that this is a valid assumption. This is validated by checking the bending caused by the uniform force. If the force causes bending (or I should say, if it yields due to bending), you know your initial assumption is no longer valid.

r13 (Civil/Environmental)31 Dec 20 15:02
Because it is assumed that the column is infinite rigid within the rectangular area that encloses both flanges.

Celt83 (Structural)31 Dec 20 15:52
You are assuming concrete/grout is infinitely rigid which is not the case. While the concrete is significantly rigid it still experiences small strains from the bearing pressure, and once you have any movement at all then you develop curvature in the bearing plate. The rigid plate assumption while in theory not correct has been proven to be safe, in reality the pressure distribution will be more localized around the column profile.

MegaStructures (Structural)(OP)31 Dec 20 15:56
aha now that makes sense! I made this much more complicated than it is. I can say I learned something today now. Thanks Celt83

r13 (Civil/Environmental)31 Dec 20 16:44
At the end, don't forget this is a conservative practice to ensure the plate have adequate strength to resist incidental bending, as there is no such prefect column exists.

HTURKAK (Structural)1 Jan 21 15:43
AISC design approach of column bases involves an elastic analysis based on the assumption that the plain sections remain plane.With this approach, by solving equilibrium equations,with linear stress distribution, the maximum concrete stress and the and the tension force at anchor rods can be determined.

 

[HTURKAK]

I have copied and pasted the relevant page of AISC Design Guide ( DG-1) below,

If your assumption is, the bearing stress under the base plate uniformly distributed , and if your analysis are solving
equilibrium equations based on the assumption that the sections remain plane , this is literally means ignoring the flexibility of base plate...I will suggest you to look EC-3 to see different approaches..

GOOD LUCK...

P.S. if all you have is a hammer, everything looks like a nail..

 

[r13]

HTURKAK,

It is not intuitive. The development of this method is not random, though consideration of flexibility is not directly shown in the equations or write ups. I tried to explain the process of thinking, but seems failed. In the end,, I would agree that if you say "this procedure does not "include" the flexibility...", instead of the word "ignores", which, IMO, has no base.

 

[JLNJ]

Consider a 10x10x0.500 square tube column with P=160 kips.

The required bearing area is just 160,000/(.7*4000)= 57 square inches. So under the 40 inch perimeter of the column you need to activate just 1.5" of bearing width, 1/2" each side of the column sidewall. Any more plate than that is for anchoring or erecting convenience. (One could easily make the case for allowing a higher bearing stress, as we do for anchor rod nut bearing or h-pile embedment.)

Now, if you have the usual case of a 16" square base plate, the uniform pressure under the plate is 160/16^2 or 0.625 psi. The moment in the plate is 2.815"k/". A 3/4" plate yields a plate stress of 30 ksi. This is a little high for A36 plate, so you might select a 7/8" plate. This is not unreasonably thick, but seems like a lot of plate if you believe it's being provided merely for convenience.

Now, say you want a little more play in your anchor bolt setting and go to an 18" square plate. The bearing stress reduces to 0.494 ksi, but the moment increases to 3.95"k/". Now your 7/8" doesn't work and you bump it up to 1". Making the plate larger in plan has made it worse. If I recall correctly, a similar condition happens for some masonry bearing plates. The smaller they are, the thinner they can be.

This is, of course, counter-intuitive and any sane person begins to question the methodology, especially when we established at the beginning that the pier would be just fine with virtually no plate at all. Explain that to the eager recent graduate who has diligently created her own excel spreadsheet to design base plates, happily chose the 1" plate, and is now answering a question from the fabricator about whether or not they can put a 6" diameter hole in the base plate center to allow for the galvanizing liquid to exit. How will the base plate work without the back span fixity of the plate under the column? Make it even thicker?

I was given the stinkeye one time from the Owner of a mid-sized fabricator who was burning up bits drilling hugely oversized holes in a base plate of mine which was probably twice as thick as it needed to be. The bits were more expensive than the base plate material. Now, he was much more upset with the provider of the bits than my design, but watching his fabrication guys struggle in their shop has stuck with me.