modification of base plates

[nuche1973]

I researched the topic of base plates in this site and found some interesting discussions but not what I am lookiing for. I am reviewing a steel structure that has already been fabricated and erected. The structure supports a tank and at this time the tank is empty. The end user wants to use this tank, per its original design, but needs verification of the support steel design. Everything checks out with the exception to the base plates. The existing plate is 3/8" thick, with (4) 3/4"dia bolts, set at 1.5" from the corners. The column is a W4x13 w/ an axial load of 25kip. The structure sits on a concrete pad. I calculate the allowable bending stress (Fb) of the plate to be 35.91ksi the actual fb is 48.326ksi. Does this look right? and if so, how does one remedy the base plate being to thin? There is one catch (there always is) the tank is in an extremely rural location (read desert) and the only repair methods available is welding plates to the structure, no crane to alleviate loads, no large machinery. (I thought about welding triangular plates to the side of the colmn down to the base plate for a larger bearing area.)

 

[JoshPlumSE]

What was the basis of your calculation for the "actual" bending stress in the plate?

Usually, hand calculations are based on a "rigid" plate assumption which results in a triangular distribution of the bearing stresses. That's not necessarily what really happens to the plate.

There are hand calc methods that attempt to distribute this load in other ways (yield line theory?). But, I'm not too familiar with them. Instead, I've always just used an FEM solution (i.e. RISABase) whenever the hand calc method gives me results that I don't trust. The FEM solution won't always give me results that I like better, but it will give me a good rational basis for accepting or rejecting the Rigid Plate assumption hand calcs.

I think welding stiffeners to the base plate should work fine to reinforce that plate. It's just a question of whether or not a retrofit is truly necessary.

Josh

Disclaimer: I'm an SE that works for RISA Technologies. Though, I'm trying not to overly advocate here.

Any FEM solution with a good plate element can be made to work for this situation. The only reason I mention RISABase is that it is the one I am familiar with...that and it is specifically designed for base plate analysis.

 

[kslee1000]

Assume you have done cal correctly, then the simplest way is to enlarge the column foot print by adding welded plates, thus reduce the critical length of the base plate. Make sure the add'l plate shars the column load, and is fully bear on the base plate.

 

[nuche1973]

kslee: Thanks that makes sense.
Josh: I currently learning how to use RISA. So, I'll try the plate thing later. Thanks for the advise.

 

[DaveAtkins]

I would figure out how large the base plate needs to be, assuming the concrete below is stressed to the maximum allowed. The size required may be smaller than the actual size--ignore the rest of the base plate. Then recheck the bending stress in the plate. It may work without being modified.

DaveAtkins

 

[csd72]

Dave beat me to it,

what he has suggested is the quickest way forward on this one and is a very valid approach.

 

[kslee1000]

While above method is quite straight forward, it is to be noted that without cutting the base plate to meet assumption, the ignored but still attached abnormal thin base plate (3/8") would see service life significant reduced, even under gravity load alone, caused by creep & relaxation.

 

[nuche1973]

kslee1000:
"creep & relaxation"? I not familiar with these terms used for steel design. Are you talking about settlement of the supporting soil?

 

[kslee1000]

The stress are concentrated under column flanges, so as stress on the concrete. That are where the concerns coming from - concrete creep, steel plate relax (deform) under high concentrated, sustained compressive load with time.

 

[nuche1973]

kslee1000, thanks.

 

[BAretired]

The load is 25 k. A 5" x 5" baseplate would result in a bearing pressure of 1000 psi which should be acceptable for 3000 psi concrete and would result in negligible bending stress. The base plates were made larger in order to accommodate anchor bolts. I agree with Dave Atkins. Leave them alone.

BA

 

[kslee1000]

You welcome.
I have always stated with 1/2" base plate as the minimum. add another 1/8 - 1/4" for tears & wears, which are common place in the industrial settings. Also, future addition in column load without checking base plate is a frequent phenomenon.

 

[lkjh345]

Throwing in with BAretired and DaveAtkins.

Existing plates are sized correctly. No need for modifications assuming at least 3,000 psi concrete footing.

 

[JoshPlumSE]

I just remembered that a couple of my former colleagues published an alternate design procedure that was published in the AISC structural journal.

The authors are Richard M. Drake and Sharon J. Elkin. I just checked and the paper can be downloaded for $10 from the AISC website.

I don't think the procedure ever really caught on. It's not a bad concept, but I don't think any labratory testing was done to verify the behavior. Even so, you may investigate using it as an alternate procdure to justify an existing design.

Josh