Dowels to soffit of existing concrete beam 1

[ajk1]

I want to deepen a series of 2-span existing concrete beams so that their flexural strength is increased, by adding 12" depth of concrete with added rebar in the bottom of the added concrete. The existing beam is 24" x 24". When 12" concrete is added they will become 24" x 36" beams. To be conservative, I will design the deepened concrete for a positive moment of WL/8 (since I cannot add any top steel to increase negative moment resistance).

The maximum factored horizontal shear stress at the interface of the new and existing concrete is 115 psi, based on VQ /(Ib). The vertical shear stress resistance is ok based on the existing 24" x 24" beams and their stirrups.

There are about 100 such beams, each being about 55 feet long. I was thinking of 2 lines of dowels to the soffit of each beam. (Please see that attached sketch. I know the attached sketch shows deepening only to 30" but that is a slip I just noticed. My intention is to deepen it to 36").

Question:

What is the best type of dowel to use for this purpose?
I am reluctant to use epoxy type dowels in overhead applications because of the difficulty of installing overhead (the dowel tends to slip down until the epoxy sets; etc.). But mechanical type dowels generally need to be torqued to set them and there is nothing to be torqued against in this application. Perhaps there is a mechanical dowel that does not need to be torqued to set it?

 

[BAretired]

How can you be sure your drill will miss the draped tendons?

BA

 

[rapt]

Is one of these beams supporting the other? Or do they meet at a column?

Surely increasing the depth of the beam will increase the -ve moment capacity of the existing reinforcement, especially if you can lift it to take out existing deflection before connecting the new concrete. But also if not lifted as long as the existing beam is not grossly overloaded.

If you are relying on this bottom reinforcement for a lot of your capacity, I think you need more than dowels. Shear reinforcement needs to be continuous between the tension and compression faces.

 

[hokie66]

Have you considered the possibility of external tensioning? Vertical reinforcement installed overhead, relying on bonding agent which may instead debond, concreting upside down, all sounds too tricky to me.

 

[Canuck65]

I would recommend looking into a fast set epoxy for the tip of the dowels (2min - 5 min.). I have used such type of custom formulated epoxies in the past. Have the contractor inject the first/top 25 - 50mm of the hole with the fast set. Fill the rest with what you would normally specify. The tip will set off quickly and hold the dowels in place until the remainder sets. I have been involved with something similar (but not quite the same) on vertical surfaces and it worked quite well. Also, a tie wire and tapcon would also temporarily hold the dowels in place.

 

[ajk1]

to hokie66 Several different options are under consideration, including external post-tensioning of course. They are all difficult and expensive, and each has its pros and cons. It is something much more easily said than done in a structure like this, but yes we are definitely considering it. The large post-tensioning contractor around here is rather negative about the idea. It also provides birds a place to land and poop, trivial but otherwise important in a parking garage. You make a good point about vertical reinforcement being difficult to install overhead. Perhaps they can build the formwork on the floor, place the rebar in it, then lift the formwork up into place.

to BARetired. The maximum shear occurs near the supports where the tendons are up near the centroid of the beam. I am trying to use dowels that do not need large development length (perhaps 4.25" development length), so it would be only very near midspan where there may be an interference issue, as the tendons are draped so that their lowest point is near midspan.

to RAPT: since the existing beam is post-tensioned, raising it will not help when all the tendons eventually corrode away, at which point all the upward force taken by the tendons will have been transferred to the new deeper beam. The tendons were elongated 8" when installed so raising the beam has negligible effect on the force in them unless the beam is raised a very large amount.

To rapt: you make a very interesting point about the shear reinforcement. I will have to go back to my old school text books. If that is the case, I suppose may have to drill in the dowels far enough to lap with the existing stirrups for 1/2" dia stirrups, which may kill the scheme if the lap length is large. The existing negative rebar is only 4#9 and is there for seismic basically, so I don't think that relying on it for gravity will help that much.

 

[BridgeSmith]

Attempting to make the new portion of the beam composite with the existing portion using epoxied dowels seems a dubious proposition to me. You'd have to account for the creep of the epoxy around the dowels.

I would also recommend consideration of an external post-tensioned system. If you're concerned about birds nesting, etc. it can be encased in grout/concrete after it's in place. To avoid cracking of the grout, likely the post-tensioning would not be fully tensioned until after the grout is in place.

 

[hokie66]

Another approach would be to add the strengthening beside the existing beams rather than below. You could then place the concrete through cored holes in the slab.

 

[BAretired]

Cored holes in the slab would disrupt the membrane and the traffic topping. It would be preferable to find a solution which does not entail holes through the slab.

BA

 

[rapt]

I was not worried about the prestress forces, though the 8" extension in the tendons was caused by the stressing operation so I do not see how that is relevant to jacking the beam. That would only affect strain induced tensions in the tendons, and with unbonded tendons at service loadings, that would be insignificant.

I was suggesting it to try to get the induced strains out of the concrete so that there would be some compression available to form a couple with the new reinforcement and you could take fulkl advanmtage of thenew tension reinforcement.

If the concrete is already deflected and the compression stresses very large when the new reinforcement is attached, then you are not going to get a lot of extra compression in the concrete or tension in the added reinforcement to do much good. The strain you can get into the reinforcement is only the extra induced in the new steel that occurs after the steel is attached. So need to look at already induced strains in the concrete and increases in the strains after attachment. Only the extra strain after attachment affects the added reinforcement.

 

[hokie66]

BA,
I see that as a minor issue to fixing the slab. "To make an omelet, you have to break some eggs".

 

[ajk1]

I had considered drilling holes thru the slab, but that would wreck the existing membrane and mastic which is not that old...it was all replaced only a few years ago. We would have to take the membrane and mastic off the whole floor and redo it, which is expensive. That is basically what BA is saying, so it is good that he is thinking the same as I.

To Hotrod10: As to creep of the epoxy around the dowels, that seems to me to be no different than in any connection where epoxy adhesive is used. Why would epoxy creep be so much more significant in this application than in any adhesive connection?

To rapt: I do not see how the induced strains can be got out of the concrete as long as all the unbonded tendons are active. There is an upward force that the tendons exert. That upward force does not significantly diminish by applying a new external upward force on the beam. I would expect that it would add to the upward force exerted by the tendons. However, your point about the compressive strain in the concrete being too much is something I will think more about.


However, I am close to giving up on the concept of deepening the beam, not only because of some of the points raised in this forum, but also because because it is not wide enough to get all the bars and mechanical couplers in, unless I make it even deeper than 36".

 

[ajk1]

to rapt: one more thought before shutting down for the night: I don't follow your point about the compressive stress. Currently the beam is balanced for about 90% of dead load. So the compressive stress at the top midspan is essentially the axial P/A from the tendons. If all the tendons fail the compressive strss increases in the top fibre at midspan, but how is that any different than in any beam? I don't expect it to be excessive. I fact, it should be less with the deeper beam than the existing shallower beam.

 

[rapt]

ajk1,

We do not know the specific details. So we can only make general comments on what has to be considered.

For all we know there could be a large extra dead load on the slab and some permanent live load and the induced stress at service could be very significant. It could be a partially prestressed design with significant stress and strain profiles under the permanent loads.

 

[BridgeSmith]

"Why would epoxy creep be so much more significant in this application than in any adhesive connection?"

In this situation, a small amount of creep movement at the dowels allows a large amount of vertical deflection of the beam and redistribution of the stresses to other components of the connection (concrete friction, etc.). Engagement of mild reinforcement in the new section of the beam is delayed until the stress in the existing beam increases. Combined with post-tensioned strands or bars to 'take up the slack', it would likely work ok, but I would be cautious about a system with epoxied dowels and non-prestressed reinforcement.

 

[drbrainsol]

Inexperienced structural engineer here but I agree with the concerned people. The proposal to deepen the beams seems a bit...complicated. You would need to achieve a very strong shear connection between the bottom of the existing and the new concrete to ensure the stresses are transferred linearly. Otherwise you'll be in the situation described very well by HotRod10.

The only way I can see it actually working is if you jack the existing beams up so that they are completely flat (or even slightly more than that so you induce some tension at the top of the beams), then drill the dowels in at an angle (lot of dowels, alternative directions) to achieve some form of "locked" connection between the new and the existing. The existing beam would have to be kept jacked up until the new

Could you perhaps consider wrapping the assembly in a carbon fibre reinforced polymer membrane?

 

[bhiggins]

It looks like you are adding a ton of tension steel. Have you checked if your new member is still tension controlled? You might need to add compression steel to balance the additional tension reinforcement.

I also do not like this scheme, I would encase the beam with concrete on 3 sides so you have more surface area for bonding. If you encase the beam on the sides with concrete that gives you room to place compression bars, you can use full sized stirrups, and you don't have to worry about drilling into PT tendons.

 

[bones206]

Could you perhaps consider wrapping the assembly in a carbon fibre reinforced polymer membrane?

Have you looked into this? I think parking structure reinforcement is one of the more common uses for fiber reinforcement. Last time I checked, Simpson provides complimentary engineering services for it as well.

 

[Ingenuity]

Could you perhaps consider wrapping the assembly in a carbon fibre reinforced polymer membrane?

Based upon additional info from the OP from other related thread/s [thread507-446120], the beam section only has 2#5 bottom mild steel reinforcing bars, so it does not meet the code for min. bonded reinforcement. Adding carbon FRP (a near-perfect elastic-brittle material) won't increase the ductility, nor have the level of redundancy, by which the code's min. bonded reinforcement requirements are based.

Also, FRP for flexural strengthening may get you a max of 50% capacity increase (depending on existing rebar and loading etc), but in this instance the OP would need the FRP to provide equivalent capacity of 13 x 0.6" dia 270 ksi strands (that fail due to ongoing corrosion) - FRP is not going to get you there.

 

[bones206]

Good info Ingenuity, thanks.