More on Base Plates 2

[jheidt2543]

I was looking at this old thread:

http://www.eng-tips.com/viewthread.cfm?qid=88803&page=1

and I’m wondering if anyone has seen any more follow up on it?

J1D notes:
“From the guide, the force and bending moment equilibrium equations at the plate-to-base interface are (page 21):

T+P=fp*A*B
P*A’+M=0.5*fp*A*B*(N’-A/3)

The dilemma is that negative (compression) values of T (anchor force) were obtained when loads (P, M) have large eccentricities.

Let’s take Example 16 (page 23) as the example, but change Mu to be 480kip.in. Then Pu=88kips, Mu=480k.in --> e=5.45in >N/6 = 2.33in (limit of full compression)
Therefore, part of the base plate will be separated from the base and the anchor bolts must be in tension.

However, the outcome is: A=4.04 --> T = -0.16kips (compression)

This can be proved more easily by a Spreadsheet or MathCAD. Keep changing Mu, you’ll find there is a theoretically tragic zone. The results fall apart and are contradictory.”

I have looked at the paper “Inconsistencies in Column Base Plate Design in the New AISC ASD Manual" and it presents an alternative equation, which I will be working through. I’m just wondering if others have run into this and what you may be doing about it.

 

[msquared48]

Blodgett has some equations to account for the prying action you are describling in designing base plates. Years ago, I had to go to an old math book to help develop the solution to some of the equations given.

Mike McCann
MMC Engineering

 

[msquared48]

I will have to post twice, but here is my solution I did some 26 years ago.

Mike McCann
MMC Engineering

 

[msquared48]

And the second page...

Mike McCann
MMC Engineering

 

[msquared48]

Let's try that again ... I obviously have the fabulous flying fingers syndrome here...

Mike McCann
MMC Engineering

 

[kslee1000]

Without refer to DG, T & fc can be solved by the static equilibrium equations:

Sum F = 0, P = T+C
Sum M = 0, M = C*d = T*d

As long as the e (M/P) falls within the plate, there is a unique solution, which can be obtained through a few iterations. (Note: the last responder on the referred link has pointed out correctly)

 

[kslee1000]

Note the second term contains mistake, the moment shall be taken about T, or P, with sum M = 0.

 

[jheidt2543]

msquared48:

Thanks Mike, very nice concise two page summary. From 1983! You're a pack-rat like me, it's hard to throw stuff away - you never know when it will come in handy again.

I'm seeing part of my problem, the columns have small vertical loads compared to their moment loads and the e=M/P
eccentricity lands outside the length of the plate. I'm wondering now about adding a horizontal brace to the frame to take the lateral load out of the base plate design.

 

[BAretired]

Mike tackled the problem from a purely elastic perspective following the method of Blodgett.

Using Limit States Design (LSD), P and M would be factored loads with M = P*e.

If e < 1/6 the load is inside the kern of the baseplate and there will be no tension in the anchor bolts.

If e > 1/6, the problem may be solved by moving the load, P to align with the anchor bolts and adding a moment P*f where f is the same as in the Blodgett solution, i.e. centerline of baseplate to bolts. The moment about the anchor bolts is then P(e + f).

Estimate a uniform stress block and, on that basis, calculate T' (tension in anchor bolts due to P(e + f)). Adjust the stress block as required to comply with the code and recalculate T'.

Finally, T = T' - P where T is the ultimate tension in the anchor bolts due to factored loads.

I think this may be the same procedure as kslee was talking about.



BA

 

[kslee1000]

Without get into details, just by a quick glance, I sense Mike's method is little deeper than the simple static equilibrium solution. Looks like it incorporates the deformations of the concrete and the anchor bolt, so on top of the simple static equations, there is a strain compatibility component to look at. Interesting is it? Kind of lazy to do the investigative work, wait till someone to validate it, and pointing out the benefits of that method.

 

[slickdeals]

Wow!!! Calculations from the day I was a year and half old .

 

[kslee1000]

It reveals engineering foundamental evolves slowly with time, sometimes stays the same. We are repeating the same problems, using the same methods, as the engineers on years ago, but situation changes each single case. That's, I assume, part of the reason why while our role in engineering is diminishing, but would not disappear

 

[jheidt2543]

Yea, there's one in every crowd! Don't take it too hard Mike, Slickdeals will have to work many, many more years which will assure us someone will be paying in for our social security!

Thanks for all the comments guys, very helpful.

 

[slickdeals]

With all due respect, I work in an office where I have about four or five 50+ year olds, and I can't thank god enough for people like that around me who have been there and done that. I have absolute respect for someone who can work a problem out on the back of an envelope and get the answer to about 90% accuracy as a super accurate finite element model that will take 2 days to create, 3 days to debug and 5 hours to run and only get you to 95% accuracy.

And yes, I hope to be around for a few years and hopefully learn enough to pass around the knowledge when I get old.

Thanks to all you guys for being around and teaching "us" the physics and mechanics of problems that we encounter.

PS: I did not mean to be offensive to anyone about their age, and no I am not an ageist. I was only trying to put things into perspective about how far I have to go and how little I know.

 

[jheidt2543]

No offense taken and none meant. Sometimes it's hard to read the humor between the lines, it's one of the problems of this form of communication. Where's a grain of salt when you need it?

 

[msquared48]

No disrespect taken here slick... Just trying to contribute where I can.

No worries.

Mike McCann
MMC Engineering

 

[msquared48]

And slickdeals, I wrote the solution to this problem for a program for my HP41C which apparently was around 2 years before you were born when I was working on a 26 story highrise in Bellevue, WA.

And jheidt2543, yes, from this solution you can see the interplay of the moment with the resisting load and how the resisting force changes location. The solution on one extreme is pure tension, and pure compression in the other.

Mike McCann
MMC Engineering