Tensile Bond Strength (Minimum embedment length) of a Plunge Column 1

[EricHimawan]

Dear all;

I am designing a Tower Crane foundation that (due to construction method reasons) should stand on a steel frame, and the steel columns (I profile) would be embedded into a bored pile, making it a plunge column. Since this is a Tower Crane, we are expecting considerable uplift force (the value from my analysis is around 500 kN base reaction), which then raises the question: is there a formula/code that regulates the minimum embedment length of a steel profile inside concrete?

I have read a source that stated tensile strength (ft) = 0.5 sqrt (fc'), and then bond strength should be taken as 10% of ft. Is this process reasonable? Is there any legitimate reference from papers I can read? My thought process is

[Bond strength] *[skin area of profile inside concrete] > Factored Uplift Load
[10%ft] * [perimeter of profile*required embedment length] > Factored Uplift Load
am I mistaken? Please correct if I am.

Thank you for reading, hoping for answers.

 

[KootK]

I've seen similar recommendations for pile embedment into pile caps when the piles are in compression. I wouldn't be trusting that for the tension imposed by a tower crane though. Instead, I'd want something like a bunch of studs on the embedded parts of the columns and appropriate rebar detailing in that area to transfer the tension down into the lower regions of the piles.

 

[Enable]

This is something I've never done but seems neat.

KootK so is the idea here that we are using the weight of the concrete to resist the uplifting force (plus whatever skin friction you have of the outside of the concrete, which to be conservative I imagine we take as 0) and the entire enterprise is to transfer the tension in our steel columns to the concrete mass via some sort of shear connection?

Seems simple. But how do you get comfortable with the distribution of that large tension force making its way to anchors/shear studs lower down? And say, not just consecutively failing them? Are you thinking that vertical bars in the caisson should be detailed to help that transfer (e.g. load transferred from first set of studs higher up -> vertical bars -> transfer to studs/concrete lower down)? Also, what are your concerns about partial failure of the concrete + failure of lower shear studs, similar to an end tear in a steel angle?

EDIT - I suppose you design the concrete to be able to transfer the equivalent force all the way down. Makes sense. But I guess I am trying to wrap my head around the studs higher up not yielding thus introducing the possibility of a cascading failure of the studs. I mean if we design each stud pretty robustly it's possible. I guess a 2" stud welded all way around with a beefy fillet would do ~ 300 kN). But that just seems so uneconomic to design every stud as if it's taking the equivalent of 1/2 to full load. Plus local concrete failure has to be a concern even with adjacent rebar.

 

[KootK]

KootK so is the idea here that we are using the weight of the concrete to resist the uplifting force (plus whatever skin friction you have of the outside of the concrete, which to be conservative I imagine we take as 0) and the entire enterprise is to transfer the tension in our steel columns to the concrete mass via some sort of shear connection?

That's how I'm seeing it.

Are you thinking that vertical bars in the caisson should be detailed to help that transfer (e.g. load transferred from first set of studs higher up -> vertical bars -> transfer to studs/concrete lower down)?

Yup. I see it as, effectively, a lap splice between the embedded steel section and the pile reinforcing. In my head I see little concrete struts emanating from the studs over to the rebar at a reasonable angle. That, then, creates a demand for localized ties to keep the thing from bursting apart laterally.

Regarding a potential unzipping failure, my feeling is that "failure" is really crushing of the concrete in front of the studs. We assume that mechanism is ductile enough to allow redistribution in composite beam studs so I'm comfortable doing the same here with adequate end distances in play.

 

[KootK]

Fundamentally, I suppose that I'm also assuming that the steel section will be stocky enough that it's axial strain over the distance of the load transfer will be very small. If we we doing this with an straightened out coat hanger, it would be a different story.

 

[Enable]

Fundamentally, I suppose that I'm also assuming that the steel section will be stocky enough that it's axial strain over the distance of the load transfer will be very small. If we we doing this with an straightened out coat hanger, it would be a different story.

Had the same thought!

Regarding a potential unzipping failure, my feeling is that "failure" is really crushing of the concrete in front of the studs. We assume that mechanism is ductile enough to allow redistribution in composite beam studs so I'm comfortable doing the same here with adequate end distances in play.

I can appreciate the parallel and I think you may be right. But my mind is stuck on any yielding or concrete crushing resulting in minute degree changes of the base that could lead to substantial PDelta effects. Maybe a lateral stabilization configuration at grade level or similar would mitigate this concern (either with steel or grade beams).

Though, like I said, I've never done this so perhaps my concern is unjustified and I am just scared at the thought of the support for a large crane moving/yielding...at all

 

[Settingsun]

API RP-2A gives 0.14 MPa allowable for plain pipe inside pipe connections. Your I-shape with re-entrant corners means you need to interpret an effective perimeter, maybe just the outside of the flanges. However, I agree that some rougher bond is better and maybe welding deformed reinforcement would be simpler than studs.

I don't see that you need to take skin friction on the pile as zero if there's reliable skin friction available.

 

[KootK]

...minute degree changes of the base that could lead to substantial PDelta effects.

Unfortunately, connection flexibility is something that we deal with, and is difficult to predict, for pretty much all steel to concrete connections. Moreover, any foundation settlement / flexibility would only exacerbate the situation. For something like this, I'd give some thought as to whether or not 2" of base stretch over the 10' to 12' width of the crane would be enough to cause a P-delta stability issue. Were that the case, I'd seek to re-gig the setup somehow.