Construction Joint design

[BengalEngineer]

I am designing a construction joint on a base slab. I have found a detail that resembles the picture attached. I see the vertical bars will be taking part of the shear. The question is are the horizontal bar groups accounting for shear friction?

Can i completely rely on the vertical bars for shear or are these horizontal groups needed?

I will be finding the shear using the strip method. Do I distribute this shear load across the whole shear plane or do I apply this shear load per foot across the shear plane? I would think applying this load per foot across the shear plane would be overly conservative. But at the same time i do not know the exact location of this shear.

 

[hokie66]

How thick is this slab? Keys in slabs are usually poor practice, as the outside parts tend to break off, leaving no shear resistance. I don't see any bars or dowels crossing the construction joint plane, and thus don't understand how you consider that the bars shown would assist in developing shear across the joint.

 

[BengalEngineer]

That slab is going to be pretty thick, 8-10 feet.

 

[civeng80]

Slab 8 to 10 feet thick (or is that 8 to 10 inches thick ?). Thats one mighty thick slab.

May I enquire what this slab is for? It obviously cannot be for any typical building project !

Can the thickness be reduced by any means?

 

[hokie66]

If the slab is 8 to 10 ft thick, then I assume it is a mat foundation slab, either supported on the ground or on piles. Either way, construction joints in such a slab should be located where shear is minimal, such as near centre span or centred over supports. That way, you just need to continue all flexural reinforcement through the joint, make sure the joint is rough and clean, and cast the second section. Relying on keys, dowels, or shear friction in such a massive structural element is not wise in my opinion.

 

[kslee1000]

First to define the name of joints (per my personal dictionary

1. Construction Jt - a construction cold joint, continuity is to be maintained by cont. bars, or dowels.
2. Control Jt - a weakened surface, or through, joint that is mainly used for shrinkage/crack control.
3. Expansion Jt - as name suggests, it should be constructed to allow free movement of the concrete segaments. Depends on the application, the joint can be grouted at later stage, or filled by flexible joint material. The shear capacity could be achieved by concrete shear key(s), or sliding dowels (one end fixed, one end free with injection of suitable grout material). Continuous reinforcing should be avoided.

Your case, from the thickness of slab and the joint detail you have selected, suggests a expansion joint is required. From past experience, the shear key works with proper dimension and geometry to prevent vertical movement due to differential settlement of the adjacent segments and shear. If there is concern of relative rotation, add dowels as mentioned above. For your slab, I would recommend using more than one shear key (2 looks good). However, I don't like the reinforcing detail shown on the attached sketch (look sloppy), closely spaced small, trapezoidal, bars bend to shape with adequate development length and horiz. reinforcing is recommended (to protect key from undesirable cracks).

 

[kslee1000]

Correction:

"...If there is concern of relative rotation/longitudinal offset due to sliding, add dowels as mentioned above."

 

[BengalEngineer]

kslee1000

Do you know where i can get any literature about the design of shear keys and their reinforcement?

 

[kslee1000]

BengalEngineer:

I don't think there is explicit discuss on design of shear keys, and it is unclear how the shear stress behaves accross the keyway. However, you may understand better by reading ACI requirements on design of corbel. Though the geometries would be different, the concept is the same, especially true for smaller keys. Also, you will need to familize with design of shear friction reinforcement.

I anticipate you will need deeper (vert. dim), accross length (horiz. dim) keys, for such case, the bearing force usually will be low enough for concrete to handle, however, I will simply let rebars to carry the entire load through concept of shear friction to simplify the design process. I might try, at a later time, to upload examples of shear key reinf. I used to use.

 

[BengalEngineer]

O, Thank you for all your help.

 

[kslee1000]

BengalEngineer:

Here is a lift key detail similar to the one you had in mind. Please note only the additional bars around the keyway are shown for illustration purpose. Use the smallest bar size that would work, for ease of fabrication.

 

[BengalEngineer]

Thank you kslee for all your information. It has helped me greatly.

There is still one thing that bothers me....I am still iffy on the shear calculation. Using the strip method you find a resultant shear force at a certain point on the cut. Do you design for that shear force per foot along the cut. Or do you distribute that shear force along the cut?

 

[BengalEngineer]

?

 

[hokie66]

I don't understand the reason for your question, and don't concur with some of the other opinions offered, but refer you to my posting of 7 March.

 

[kslee1000]

BengalEngineer:

Sorry to leave you in dark for so long. I too don't quite follow your question, but hope the explanation below helps.

The shear key in discussion is to provide lock mechanism in two mass bodies. The load on the key usually is a uniform line load in nature, so does the reaction. You can visualize it as precast pipes with tougue and groove linking the adjacent pieces, it prevent uneven vertical movement, but still allow small amount of lateral free movement (mainly shrinkage & expansion). This type of joint is more prevalent for at grade, below grade, or under water civil works, it is rarely seen in buildings (for which keys are usually having bars accross the joint to maintain continuity).

Back to your question on load distribution, as mentioned above, the load on this type of key is usually uniform in nature, however, you may have localized concentrate loads, if it is the case, yes, you can distribute over a finite width (a region bounded by 35-45 degrees to each side of the load). If two loads are closely placed and the concrete mass is quite thick (deep), there would have stress overlap, then just add the stresses up and check an unit strip for shear friction.

If the above does not cover your situation, you may need to upload a sketch, or describe your question further (what is the application, type od structure, geometry, how it is loaded...etc).

 

[OzEng80]

kslee - I take it the load across the joint could be either up or down? A corbel joint would be much simplier if this is not the case.

 

[hokie66]

The OP said his "base slab" is 8 to 10 ft thick. Hard to imagine that you wouldn't know which side of the joint is supported, and if that is the case, probably the simplest way of providing positive shear resistance is to slope the whole depth of the joint. Just make sure the formsetters know which way it slopes.

 

[kslee1000]

OzEng80:

This key is provided along the length of the concrete slab/mat to provide inter-lock between structure units, for which free lateral movement is essential, but vertical movement is not desirable. I am aware of "corbel", but don't have clue on the term "corbel joint", so I couldn't comment on its viability. But, note that a trapezoidal key is always preferred over the square/rectangle key for the reasons that the latter is more more difficult to vibrate around the right angle corners, which would result in poor quality; and the latter is prone to crack due to stress concentration around the sharp corners.

Yes, a shear key is intended to receive paired loads in a up-down manner (one unit up, one unit down).

 

[hokie66]

kslee,

As far as I can see, the OP never said this was a movement joint. And a trapezoidal key is not always preferred, in fact they don't serve the purpose intended. When a joint opens horizontally, it stops transmitting vertical load unless the supported part slides down, thus creating a step.

 

[OzEng80]

Apologies kslee – I meant to direct my comment to the OP.

I would avoid using keyed/inclined joints for a high load/movement joints as described by hokie. It is also worth noting that the bearing stresses within the key increase as the joint opens (reduced bearing area).

Corbelled joints have the advantage of having a horizontal seat on which proprietary bearing strips (that permit movements and rotations) can be installed that ensure movements are not restrained and direct the load away from the edge of the concrete (prevents damage). In this instance only a very short corbel would be required (huge depth and shear capacity) so compaction issues should be able to be resolved (I imagine there would be a pour break somewhere anyway?).

Bengal – make sure that you design the joint as fully ‘open’ for whatever system you adopt.