Design of two steel columns with one inside another 2

[structure567]

I am in the process of designing a steel pipe column that will require inserting a new smaller column inside of an existing column to increase the allowable strength. The smaller column will have an outside diameter 0.2" smaller than the larger column's inner diameter.
If I were to treat the columns as two separate columns how should I calculate the force applied to both the columns?
The columns has an existing circular deep foundation. The columns will be embedded the whole length of the foundation.

 

[r13]

Have never seen it done before. Can't you fill the pipe column with concrete, what size is the column?

 

[structure567]

column is 4" and client rather have a prefabricated material and minimize the use of concrete.

 

[STrctPono]

I'm trying to imagine this....

Unless you have a shear mechanism between the two columns (i.e. grout, welded, etc.) they will not act compositely and will instead "share the load" as a percentage of their own flexural stiffness relative to the sum of the two. However, with that being said, I feel like unless detailed properly, their behavior may be much harder to predict. If you are thinking you can just drop this new column inside this existing column and call it good, then forget about it.

I had a delegated design engineer try and submit something similar to what you are talking about for a pedestrian railing post connection to a base plate and I called his bluff.

retired13's suggestion of using concrete really is the easiest and cleanest way to do this and will improve the strength considerably. Confined concrete has way higher strength properties over unconfined concrete.

 

[CANPRO]

How is the load delivered to the column? I'm not sure how you'd insert a new column into the existing, cap it (presumably), and ensure somewhat proportional load distribution. For these two columns to truly share the applied load I think you're going to end up with a very funky cap plate and a decent amount of fieldwork. Concrete filled might be the path of least resistance here. I'd be very interested to see how this column is loaded.

 

[hokie66]

The question others have implied: is this column a compression member or is it a flexural member...or some combination of both?

 

[Agent666]

As hokie is maybe eluding, filling with concrete barely increases the moment capacity (maybe 5% if you're lucky. It's primarily a means of increasing the design axial capacity of a member.

But if its already loaded then the concrete fill isn't doing anything for the existing loading, only what might come after would a column be considered composite.

If the column is existing. How are you attaching the new one to the baseplate to actually transfer any load. Might need to provide more information on the exact configuration you're looking at, constraints, etc to enable everyone to respond with specific information or point out issues.

Whole concept seems a bit of a stretch. I can't quite picture what you mean by the column will be embedded the entire length of the foundations?

 

[Agent666]

A pertinent question for OP to answer is if you are asking about how much load to design each column for, then how is, the load being introduced to each column?

 

[MIStructE_IRE]

I don’t believe that this can work. Even if you COULD detail a proper connection that would adequately transfer load to both, it will only apply load 50/50 to each and as such the smaller pipe buckles before the larger one ever reaches capacity.

I would steer clear of this detail.

 

[structure567]

Thanks to everyone with the new replies!
Agent666 the new column will go down all the way into the deep pile footing so there will be no baseplate for the new column. There will be a lateral wind force being applied to the column and the axial load is very minimal. After doing some calcs, both pipes will be able to handle at least 55% of the total load. Both pipes are also below 200 limit for slenderness ratio.
I'm having trouble trying to see how the load will be distributed. Should I apply the maximum allowable load the outer column can support and have the inner column handle the excess load? Or should I distribute the load according their stiffness as StrctPno has suggested? As of now I'm treating this as two separate columns. I'm also having to consider deflection which is a whole another issue.

 

[Agent666]

I'm still failing to see how you're going to get your bending loads into the central tube. To me even though it's relatively tight fit, it's still just rattling around inside another member. It's only going to try take load when it's in contact with the outer tube which obviously is dependant on the stiffness of each tube, straightness of each tube.

Perhaps a better approach if able to do it is to feed a reinforced pile cage down the outer tube then concrete it. My earlier comments on increasing the bending capacity were related to filling it with concrete only. However add some reinforcement and you might achieve what you are after with a reinforced concrete/steel tube composite section.

 

[r13]

structure567,

Assume the pipes are perfectly fit without gap in between, and the contact surfaces are smooth without friction, I would try to determine the load ratio by consistent displacement/deflection concept - place an unit load in the span, then write two equations for deflection based on property of each, and equate them with the known condition, that the deflections shall agree with each other, and there is no gap before and after the load applied. A little tricky, but should work.

 

[STrctPono]

Assume the pipes are perfectly fit without gap in between, and the contact surfaces are smooth without friction, I would try to determine the load ratio by consistent displacement/deflection concept - place an unit load in the span, then write two equations for deflection based on property of each, and equate them with the known condition, that the deflections shall agree with each other, and there is no gap before and after the load applied. A little tricky, but should work.

In other words:

and will instead "share the load" as a percentage of their own flexural stiffness relative to the sum of the two

 

[phamENG]

Question 1) How are you going to connect the new column to a deep foundation system inside of another column while everyone is standing outside of the column? I don't see how that works. There needs to be a positive connection there somewhere.

Question 2) Have you considered surface irregularties on the inside face of the existing column or the outside face of the new column? You've got 0.1" all the way around - a little dent here, a little corrosion there, and you've got a big problem. And that's not even considering mill tolerances.

Question 3) I think this was already asked, but I didn't see an answer - how is it loaded? Shear tabs on the side of the column? If so, what's your load path from the existing column to the new one? Cap plate with beams sitting on top? I'm guessing you'll be removing those beams to do this - the load transfer is a little more clear, but the detailing is a mess. Is it pure axial load, or will it experience some amount of moment as well?

I think questions 1 and 2 are the biggest hurdles to making this work. I just don't see how you fit the thing in there in the first place without damaging something. And if you do, I don't understand how you're going to connect it. Good answers to 3 are needed to help you with your question of load sharing.

 

[canwesteng]

Disagree with everyone else, if it were axial load only it would be trivial to make sure they share the load equally at ultimate limit states, since one can't buckle without the other buckling. If mostly resisting bending, through deflection combability it's not too hard to figure out how strong the new double column is. Take the deflection of the outer pipe at failure, subtract 0.4" for your fit up allowance, and figure out what the bending moment in the inner pipe is with this deflection. The real problem is that I can't imagine you're going to be able to actually fit the smaller pipe inside the bigger one, unless this is an awfully short column or you're lucky and you get real straight pipe with perfect wall thickness.

 

[phamENG]

canwesteng - yes and no. I don't think we're trying to make them share equally, but understanding the percentage going to each is important. Once one of them reaches its buckling load, it will "try" to buckle, but will be braced by the flexural rigidity of the other one. That means it has to brace it by bending, creating a combined bending and axial load situation. If the other column is not capable of doing both, they'll buckle at the lower load. If it is, then that columns new buckling load will be less than it was previously. So your lower bound critical buckling load will be the weaker column, but the higher bound will be something less than the larger column. But then, if you consider that the existing column isn't unloaded first, it already has some set amount of dead load that can't be shared without jacking and shoring, so only new dead loads and live loads will be shared and the existing column's buckling load is skewed.

All in all a very interesting if wholly impractical stability problem.

 

[r13]

If the flexural demand is not too great, another method might work - insert a 2"ø pipe into the 4"ø pipe, and fill the annular space with (non-shrink) grout with some fibers. Grout the 2" pipe too, if necessary. Now this can be considered a composite column, and designed as thus. It could get ride of the buckling headache also.

 

[phamENG]

Problem with that, retired, is that the owner doesn't want to use concrete (why is that, anyway?). I imagine grout would fall into the same category of contraband. If it could be grouted/concrete filled, all of these problems would likely go away...

 

[structure567]

Question 1) How are you going to connect the new column to a deep foundation system inside of another column while everyone is standing outside of the column? I don't see how that works. There needs to be a positive connection there somewhere.

So the existing column is hollow and it goes down the complete depth of the pile foundation. I would have placed a steel plate on top of both columns and weld it shut. But now I believe retired13's advice will work better. I'm going to see how the design will work if I place a smaller column and fill up the gap with concrete.

Question 2) Have you considered surface irregularties on the inside face of the existing column or the outside face of the new column? You've got 0.1" all the way around - a little dent here, a little corrosion there, and you've got a big problem. And that's not even considering mill tolerances.

I am also worried in regards to this but as of now I am in charge of whether or not this design will work if it did fit all the way through. I also think that it will be incredibly difficult to fit the smaller column with such a small space in between.

Question 3) I think this was already asked, but I didn't see an answer - how is it loaded? Shear tabs on the side of the column? If so, what's your load path from the existing column to the new one? Cap plate with beams sitting on top? I'm guessing you'll be removing those beams to do this - the load transfer is a little more clear, but the detailing is a mess. Is it pure axial load, or will it experience some amount of moment as well?

So I'm not sure what you meant by this question. The existing column is used as a sign pole and has a sign on top of it. The only load being applied to this column is the self weight and the wind force due to the sign. The sign was designed specifically for a 4" pipe which I believe is the reason the client wants to keep the original column. A bigger column would mean the sign would need to be replaced as well since the connection will not fit a another sized pole.

 

[r13]

The owner must affiliated with AISC and hates ACI, anything cement