Embedments in concrete 1

[ADCanada]

Hi To All

I have a client that is asking to run his round HVAC in the flat concrete slab... the issue is that the pipes are 150mm in diameter and the flat slab thickness is 230mm ... I am trying to find a provision in the Canadian code or a paper... I found in the ACI318 clause 6.3.5.1 that the embedments shall not be larger than 1/3 of the thickness of the slab... but how is that calculated? is there a paper on that?

 

[JedClampett]

I doubt there's any kind of big study on that. It seems to make sense. You would just be building in a major weak plane and the concrete would crack along the whole routing, if not fail it.

 

[hokie66]

A ridiculous suggestion, and as the design engineer, you don't need a "paper" to reject it.

 

[Ron]

The provision for control joints in a slab is that they are between 20 and 25 percent of the thickness of the slab. Your pipe constitutes a cross-section reduction of 65 percent....guaranteed to crack and guaranteed that the HVAC piping would take the direct load of anything placed on the slab.

As hokie66 noted....a ridiculous suggestion. Just reject it.

 

[ADCanada]

Thank you hokie66 and Ron for the quick reply...I totally agree with you both that the request is ridiculous but the client has some background in engineering and a simply rejection will not be enough... however; the HVAC came back with a new diameter and not it is 100mm .... so basically the scenario is as follow

1. the slab thickness is 230mm one third of that is approx. 76.66mm therefore we are here 100mm (dia of HVAC) minus 76.66 equals to 23.4mm ... that is 23.4mm more than the ratio stipulated in ACI318.
2. Based on item # 1 I chased the equivalent stress block for rectangular section to find the neutral axis and I found it at 23mm from the top of slab.

Now does the neutral axis get affected (in other words shifts) if there is an opening?

 

[TXStructural]

ACI's limit is a simple diameter-to-thickness ratio (1:3), so 100 mm : 230 mm is 1:2.3, or not even close. HOWEVER, keep that in context, as the limit does NOT apply to engineered solutions. If you compute the stresses and evaluate them for engineering concerns, as the EOR, you are permitted to embed objects or voids of any size.

This all depends upon where in the slab the ducts will be run. It is common practice is some slab systems to void certain areas of a slab to remove dead weight, but we would never consider this at supports or areas of high shear without computations. Is this a one-way or two-way slab? Will ducts run randomly throughout the slab, or can they be confined in "good regions"? As far as allowing random voids of this size, in general, I'd be skeptical about crack performance, but doubt you would fail a reinforced slab in any meaningful way because of a duct here or there.

 

[ornerynorsk]

"the request is ridiculous but the client has some background in engineering"

If that were so then the client should even better understand the problem of what he/she is asking you to do. "Pulling rank" as having some engineering background tends to belie the true ability of such individuals.

It is better to have enough ideas for some of them to be wrong, than to be always right by having no ideas at all.

 

[msquared48]

Well, if the conduit is 6" thick and the slab is 9" thick, you could deepen the slab in the area of the conduit to 15" or so, similar to a localized strip footing, and add in extra longitudinal and transverse steel top and bottom for cracking - but it will still crack there.

Mike McCann
MMC Engineering

 

[dik]

Mike... you could have added to send a letter to the client to advise him that there would be a likely chance for cracking... I've had to thicken slabs for conduits on numerous occasions... another concern is the interruption in 'smooth' slab undersurface for SOG construction.

Dik

 

[hokie66]

Mike and Dik,

We aren't talking about a SOG here.

 

[msquared48]

OK. Then I agree that the idea is weird, but with all sorts of money, I could make it work!

Mike McCann
MMC Engineering

 

[ADCanada]

TXStructural
The phrase “as the limit does NOT apply to engineered solutions” this is an excellent phrase… to answer your question, the slab is a two way flat slab… the duct as I mentioned earlier is now 100mm after a good discussion with the client… now I took your advise and started with what you called "good regions"…. see below my side of the solution

ornerynorsk
Well we all had people that try to push the limits of engineering… but as I mentioned before to many clients a simple answer of “no” or “it can not be done” is not enough… you need to give them proofs or in other words you need to support your claim in an engineering way… and at the end of the day we are liable to what we say or what we claim and responsible to what we design.

The solution in my opinion…
Now… my opinion is as follow … I went back to the basic principles of the equivalent stress block for rectangular section and found that the neutral axis is at 23mm below the top of slab (reinforcement is accounted for in calculating the equivalent stress block) , and correct me if I am wrong here, the neutral axis will not be affected by the circular whole in the section then the stresses are above the duct… therefore by avoiding the high shear area and concentrating in areas that has reasonably low moments.. (coincidentally it is at 1/3 of the span between columns) the slab should be fine … again correct me if I am wrong here…
Finally for the chosen locations I checked the shear capacity and the factored shear at that location…and I made sure that there will won't be any concentration of other sleeves from other services.

 

[TXStructural]

Sounds like a solid approach. And as long as you are reinforcing top and bottom of the slab, a conservative approach is to simply design the reinforcement as a moment-resisting couple (for strength) and move along. Since you are in the middle third of the span, the only other concern I can think of would be moment capacity of the column strips perpendicular to the duct, and it sounds like you probably have that handled in the design method.