Calculating Load on Rebar to calculate embedment?? 3

[Brandnew]

How does one calculate the load on a rebar that will be embedded into a concrete wall??

I have the Hilti manual so I know how to calculate the embedment but I'm having trouble figuring out the forces applied to the rebar.

The element in question is a beam. It's 18" x 18" with two layers of steel (3-#6's in each layer). I have calculated the shear at the interface but I'm having trouble calculating what the pullout force on the bars is. The bars will have to be embedded on each side of the beam (for its support).

Here are my questions:

1) What is the proper way to "model" a beam that is embedded on each end?? Is simple span or fixed ends appropriate??

2) I've done it as a simple span to be conservative....is that as conservative as I think it is??

3) I know the total shear at the "interface", is it appropriate to assume that the shear is split equally to each bar??

4) How does one calculate the pull-out force? Is it appropriate to assume that I just need to achieve the yield strength of the bars??

They don't teach the "details" in school and that's what you really need to know out in the work force!! Any help is appreicated!!

 

[ishvaaag]

Where I live is common to leave "waiting" dowels (even bent) to later lay a slab only connected to the standing structure by such dowels, sometimes with some keying or shelf but most no, and then consider and have designed the thing as if an entirely monolithic structure cast at the same time was the case.

Respect your case, your structure will work as much efficiently as it can starting from the devices you have laid in for resistance. This is one very positive way to look to the minimum energy of deformation thing in structures, they are mandated by nature to find one path of equilibrium if it is existent, and find not only this, but the one that has more reserve for strength for the existing conditions. So it finds first a primary path (that needs not be coincident at all with the one we devise and use for design) and if this structural system and load path is damaged starts to search other, the more efficient it finds from the viewpoint of maintaining its present integrity, which is got by the device of self-inflicting the less energy of deformation (and normally deformation as well) it needs to. Succesive reserve secondary, tertiary and so on paths are used if required in the same self-preservation of the thing mood.

Now, you can model your end connections the way you want. However, if you want definite and determinate forces you need the notional mechanisms you devise be quite closely implemented in reality. Then, for a simply supported beam this would mean no negative reinforcement.

However if you are going to use embedment both atop and at bottom, it may be more logical to consider the fixity the supporting parts can deliver.

Seen this way, you simply have an end of beam, with maybe the particularity of that you want some check in shear-friction of the steel bars passing the interface.

For the shear friction check you can use all the steel capacity not consumed by extant tensile action according to your notional scheme. All the rest of the steel section and limit capacity you can count for shear friction transfer -following some code practice- through the interface as long its capacity is entirely developed at both sides.

 

[Qshake]

Good Gravy, if they taught you a little bit of everything you'd know nothing!

Usually applications involving dowels are noted to develop the strength of the material. There are several failure mechanisms for this application and unless you're also designing the element that the dowel is to be embedded in then its fine to use the strength. The ultimate pullout force will be the lower of the failure mechanisms. See PCA Handbook for more detailed information.

On the fixed or simple - it is simple, period. I don't know how one would quantify the fixity of that type of connection but it is seen that you can't prevent rotation especially if the beam is relatively small.

Yes, you would assume that the total shear is divided equally.

 

[JAE]

I'm assuming, from your description, the following:

You have an 18" square, horizontal beam, with three top and three bottom #6 bars running longitudinally in the beam. The beam is supported at each end by a concrete wall in which you want to dowel into to create a vertical support for the beam. Is this correct?

Assuming simple span conditions will give you a conservative design for the flexural strength of the beam at midspan (your 3 - #6 bars). This will indicate that you have zero moment at each end where there is maximum shear.

Check the shear in the beam and add stirrups if your ultimate shear Vu is greater than phi Vc / 2.

The dowelling: You have two things to check. First, you have a shear friction condition between the wall and beam end. Thus, you need to ensure that you have enough steel area to engage the interface between the beam and wall. Chapter 11 of the ACI code has the technique for shear friction and the results depend on whether you are casting this beam monolithically with the wall, casting to an intentionally roughened surface, or to a smooth surface.

The second thing to check is your main question, the embedment into the wall. The shear friction method requires you to develop the full yield in the bars - for the bars that are required to develop the shear friction. Your six #6 bars may be more than enough so possibly you can develop only four. HITLI offers the required embedment for yield in their tables.

 

[Brandnew]

Thanks for the responses.

Maybe they did teach me a little of everything then

The beam is NOT supported by the wall. The situation that I'm forced to deal with does not allow me to set the beam on a wall/support. Therefore, the dowels will be ALL that supports the "beam". As you can see, the embedment is VERY critcal. Also note, this "beam" is actually going to be vertical with a high lateral load applied (no axial load).

I've added the necessary stirrups and have checked shear-friction.

If I understand you correctly JAE, only the dowels that are required for shear-friction need to go to yield?? What about the other dowels??

This is part of the reason I came here. When I calculate the "Hilti" interaction of the yield stength along with the shear strength for embedment I'm going to get a VERY deep embedment (14"+). I'm afraid my contractor won't embed them that deep anyway and at 14" they were still failing to pass.

It's the pullout that's killing my interaction equation!! Any other attachment options (angle will only work on one end)??

Thanks again. Any additonal insight appreciated.

 

[jheidt2543]

I would be very uneasy with a connection that relied only on the shear resistance of 6 rebar, even if my calculations said it should work. Is is possible to provide a haunch detail or put an angle support under the beam?

 

[Brandnew]

Looking at the area again, I could provide an angle on each side for support. Would this be a "safer" design?? Any thoughts on this??

I will run the numbers on this and see how that shakes out.

 

[JAE]

I guess I'm having difficulty envisioning your situation. The beam is vertical, yes? With lateral load inducing shear and moment with no axial. The beam is sitting on a concrete surface on the bottom end?...and you need to dowel into the concrete surface to transfer the lateral shear into the concrete surface? Is this correct?

 

[Brandnew]

You got it JAE!! This is going to be a support for a flood gate. The flood gate will have to hold back 17' of head. Maybe that will make more sense?? The top and bottom will be "doweled" into existing concrete.

However, from the suggestions on this board it seems as though it will be easier to anchor some angles for the lateral support. If this is done I have more questions (since I haven't done this either):

1) What is a good bolt spacing for an angle loaded in thi manner?? I can run the angle the whole width (18&quot but how close should I space the bolts....what is generally accepted??
2) How does one attach the beam to the angle (use anchors here too)?
3) What is the "best" way to terminate the reinforcing in the "beam"?? Hooks?

Doweling just doesn't seem practical from a design standpoint. All suggestions appreciated!!

 

[breaks]

Brandnew -

I'm not 100% certain of the existing conditions, but it may make sense to remove the existing concrete from the area in question (the footing, I assume), clean the existing rebar and then tie in the proposed rebar from the vertical beam using hooks. This would provide a more-than-adequate connection. I did something similar to this in a Finished Water Pumping Station expansion where we added a wall next to sluice gates. The wall/footing interface was completely redone to ensure the integrity of the wall/footing system.

Hope this helps, Matt

 

[jeg]

This appears to be a very onerous application. If this has to last for many years holding back 17 feet head of water you should not try to save pennies by dowelling and bracketing. Don't be afraid to consider cutting back the existing concrete and casting your new beam monolithic with the existing concrete. Make it big, make it strong - no-one will thank you if it fails.

 

[JAE]

Well, I would tend to agree with jeg and breaks.

Using angles in an exposed condition like that - always wet - doesn't sound so great as the steel would certainly deteriorate over time.

If you are truly reluctant to cut into the existing concrete, then what I posted above still applies - shear friction would definitely work - just have the contractor intentionally roughen the surface of contact between beam and concrete face and drill/dowel into the concrete using the HILTI HSE 2421 system. Simply develop the yield of the rebar that you need for shear friction. A #6 bar only requires 5" to develop yield in 2000 psi concrete.

Another detail you could include as well: sawcut the face of the concrete in a square 18" x 18" pattern and chip out about 1 1/2" of concrete to "socket" the beam into the face of the concrete. I would still count on the shear friction so this, to me, would simply be an added comfort. (We engineers always need our sleep!)

 

[Brandnew]

Great ideas!!

Unfortunaley there is not a lot of concrete that I could remove in this situation. I could create a socket and I could go deeper than 1-1/2"...probably about 6" top and bottom. Would it be worth while to go 6" deep in your opinion??

If angles were used they would be mounted on the "dry" side so deterioration would be limited. As long as the pumps never fail WHILE the river level is high the door will see NO water. It is quite likely that this door will never see any water.....but I still want it stout "just in case"!!

One last thing. JAE, you state the if doweling I would just need to go to the yield depth to meet the shear-friction requirement. Why can I dicount the shear force for embedment?? Don't I need to consider the shear applied at the embed as well as the tensile??

Keep the great ideas coming!!

 

[ishvaaag]

My concern is the safety factor. There will be some dynamic impact and either mechanical or through adhesion, some deterioration of the bonding interface may occur. In all, I see the gates relying in a mere column as flawed. I would suppert the gates in stable wingwalls.

 

The embedment length is a function of yield strength and stresses in the rebar.
go to AS3600 chapter 9 (as I recall) Australian Standards.
hope that helps

 

[Brandnew]

Okay, I think I see how I'm going to attempt to set this "beam" up. On top I'm going to cut out a 6" area for the "beam" to bear on. On the bottom I will make about a 1" socket for "additional insurance".

This brings up a more complex engineering question, at least for me. One side of the support will be supported in the R/C ceiling and the other side in a R/C floor. The ceiling is 9" thick holding up 6-7' of soil. The floor is 12" thick and supports (very little) pedestrain traffic. There is a battered wall (9" to 12" thick) about 6' away, at it's closest point, that goes away from the support at a 45 deg angle (this is actually one wall of a 6.5' wide hallway)!! There is also an 18" parallel wall right next to the support that the angled wall frames in to. (A horrible "plan view" picture can be found below)

Here are my "newbie" quesitons:

1) Is there a VERY conservative way to check what the ceiling and the floor can support?? I don't have access to any computer programs so all calculation will be by hand. I know what actually takes place is very complex but I'm curious if there is a super conservative approach I could use. What would the tributary areas "probabgly" be, etc..??

2) What is the "most likley" load path that would be taken?? Would a lot of load be taken up by the wall next to it, would it go to the angled wall, or??

!
! Another Wall
!
!
________ "Next to" wall
S> _ \ Angled wall
\ \
\ Other side of hallway

S = Support, > shows the direction of the force.
There is a large open room "behind" (off screen) the support.

All comments appreciated!!

 

[tjh]

your problem seems to be complicated and I advice you to consult with a P.E.

Good luck