Opening in Hollow-Core Slabs

[EZBuilding]

I am working on a project where the hollow core specialty engineer has proposed the following sequencing for providing shaft openings in a hollow core slab system.

1. Erect Hollow Core Slab
2. Grout Hollow Core Slab Keyways
3. Sawcut openings in the hollow core slabs

I have reviewed the 1979 PCI Journal "Load Distribution Test on Precast Hollow Core Slabs with Openings", which provides some empirical evidence that a grouted keyway can provide for vertical load transfer between adjacent hollow core slabs. Nonetheless, I am looking for some further justification for the transferring of loads from one panel to the other.

I have a quick sketch of the condition shown at the link below:

[URL unfurl="true"]https://res.cloudinary.com/engineering-com/image/upload/v1544384007/tips/Hollow_Core_Slab_Openings_Examples_gnidss.pdf[/url]

I can visualize a load path in which the hatched sections of the hollow core slab spans from the contiguous sections of hollow core slab A to the hollow core slab B. Vierendell Truss action between the top and bottom concrete flanges and the concrete webs. The HCS B would be designed for it's own tributary width plus half of the tributary of the hatched section and account for the sections of it's own slab which is not continuous. This then requires the grouted keyway to transfer a reaction from HCS A to B.

Is anyone aware of any publications which provides some justification for calculating the capacity of the grouted keyway to transfer these loads?
Is the transfer mechanism solely bond strength of the grout to the hollow core slab? Or is a keyway type of mechanism considered?
Any general thoughts, comments or concerns? Would be interested to hear the opinions of some of our resident precasters on this topic.

 

[cooperDBM]

You can find a Hollowcore Manual on the PCI site (pci.org) that discusses openings and load transfer. When you have a large opening across a slab you will often see steel transfer headers across the opening that bears on adjacent slabs, particularly in shear critical areas.

 

[EZBuilding]

CooperDBM, I have reviewed the Hollowcore Manual, and found that they have considerations for the design of the hollow core slabs to account for the additional load imposed on them through openings nearby. I did not find any discussions on the capacity of the keyway joint to transfer vertical shear loads. Do you have any insight on that portion of the topic in particular?

 

[TehMightyEngineer]

keyway type of mechanism considered?

Yes, grouted shear keys act as a keyway mechanism. There's a decent amount of a shear transfer over those joints.

Generally I don't see any problem with their proposed steps but would agree with cooper that with large openings you need a steel transfer header; but that could definitely be installed after cutting the planks. PCI hollow core slab design manual would be my go-to reference here and it has references for shear key capacity. Edit: I see that you couldn't find the reference. I believe it's in section 4.6.1 but my hollow core manual is older so it might have moved.

I got to run; but I'm sure KootK will be here in a minute with a bunch of other good references as well. He's definitely got more experience with hollow core than I do.

Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA, FL) Structural Engineer (IL)
American Concrete Industries

https://americanconcrete.com/

 

[KootK]

I'm not one to resist a good summons. The grout keys are to be considered shear key mechanisms that do not transfer moment in the absence of fancy stuff like transverse rebar in the topping etc. I do this kind of thing on a regular basis and would love to have more definitive references. I've scoured North America and much of Europe and the South Pacific for such information. If it's out there, it'll take a better man than me to find it.

The sketch below shows how I would likely approach the design as the precaster. That said, if you drove down to the precaster's office right now and demanded to see the result of similar checking, my very strong hunch is that you'd be waiting in the lobby for a good long while. A weird thing happens to plank designers. You design enough of it, and you gradually start to forget that it's not quite a two way slab. Some kind of mass delusion of convenience it seems.

I've never seen a header in this situation but would like to introduce that technology myself. Even though there is no erection necessity, it would simplify some things nicely. I would expect to get substantial feed back on it from my fab fabricator friends though. It would be complicated by the hanger ears probably not landing over webs.

 

[GC_Hopi]

If they cut across the entire width of the plank I would expect them to use a header, below is from PCI hollow core manual.

 

[EZBuilding]

Thank you all for joining the conversations, I'll get to your invoices as soon as I can!

I am familiar with utilizing the inset steel headers at large openings, but I guess part of the question is what constitutes a "large" opening versus an opening that could be resolved through the keyway action. I also think that adding the steel header after cutting the openings is only partially benefitial, the self weight of the hollow-core slab will be engaged through the two-way action which has been discussed. Although the self weight is only about half of the design load, it's likely the majority of the actual demand seen by the system.

TME I have that same reference but was not able to find any information on the grouted keyways having any vertical shear capacity. There are some discussions regarding horizontal shear capacity for diaphragm design - something I may have a separate eng-tips post on another time. I will look through it again tomorrow at the office to make sure I did not miss it.

KootK, I agree with your load distribution diagram. Without the grouted keyway providing a vertical shear transfer, the cantilever moment and torsion on the remaining section would likely become a significant deisgn challenge. It does appear that the HCS industry has some level of "that's how we have always done it" mentality. Nothing against empirical evidence, but my head rest easier at night after I have had the chance to run some numbers.

The hollow core slab keyway geometry on my project has a very small "indented section" which would allow for a calculation similar to a typical 2x4 keyway + dowel calculation in conventional concrete design. Ignoring that type of calculation, and utilizing the slab geometry above, I would share the hatched section's load between the HCS A and HCS B. This would require an ultimate distributed load of 166 pounds per foot or about 2.5 psi "vertical bond shear stress".

This value feels reasonable. Nonetheless, I can't help but imagine a reinforced concrete slab with a pour strip in which the contractor failed to provide any reinforcement across. I would be hard pressed to convince myself to provide a similar grouted application without a more positive connection. I expect that magnitudes of loads would be smaller in that consideration.

I am expecting further information from the specialty engineer. If I receive any new information that could be of service to this throught process, I will come back and share.

 

[Ingenuity]

Spancrete Manufacturers Association have a bunch of free research/testing info on various hollow-core topics: Link

I am not sure if it has a direct topic that covers your specific problem, but a good resource, none-the-less.

 

[TehMightyEngineer]

I found the attached paper a while back; been a bit since I read it. I recall it not having enough data for my needs at the time; and it's regarding dry shear key joints, not grouted joints.

Still, some interesting information.

Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA, FL) Structural Engineer (IL)
American Concrete Industries

https://americanconcrete.com/

 

[TehMightyEngineer]

Oh, I forgot I had this as I only downloaded it a few weeks ago. I haven't read it other than a quick skim, and it's about precast bridge deck shear keys (not hollow core keys), but it has many great references to various research that I suspect will have some goodies regarding strength of shear keys. It's on my "to-read" list so if/when I get a chance to go through it I'll report back if you haven't already done so.

Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA, FL) Structural Engineer (IL)
American Concrete Industries

https://americanconcrete.com/

 

[KootK]

I guess part of the question is what constitutes a "large" opening versus an opening that could be resolved through the keyway action.

In my opinion, as long as you're only counting on the keyway action for shear, there is no practical situation where that shear demand could not be resolved via keyway action. Per your own calculations, you've got gobs of capacity for anything resembling a uniform load. An interesting example that I've thought about often is the distribution of large point loads from the edge of one plank to the next. As you probably no, design of these systems usually makes heavy use of load sharing between planks. It's technically response sharing but let's not dive into the weeds on that. This is one of the few situations where I've been concerned about localized shear transfer from a plank to it's neighbor. This condition has been tested and used in practice for loading of this sort as high as 20kip, based on my personal experience. No issues that I'm aware of. An intact side web does a decent job of distributing concentrated loads out to a few feet worth of keyway I believe.

The keys to understanding when steel headers are necessary are these:

1) Is it needed for erection of the planks and;

2) Given that keyways don't generally transfer moment, is it required for the equilibrium of any planks in the finished state?

the cantilever moment and torsion on the remaining section would likely become a significant deisgn challenge.

If either of these things is true, plan for a header. If not, expect your precaster to push pretty hard for not having a header.

No, not in my experience to date. I apologize if my intent was unclear. I'd meant to imply that these are things that I think are prudent to check, not that they are things are are likely to be critical in an application such as yours.

I would be hard pressed to convince myself to provide a similar grouted application without a more positive connection. I expect that magnitudes of loads would be smaller in that consideration.

1) I agree with the sentiment. I think that most designers would feel this way and there was definitely a time when I would have too.

2) There are some aspects of the analogous precast situation that would surely make you feel somewhat better. Firstly, cross plank shrinkage has already taken place for the most part. There's little tendency for the joint to pull apart. Secondly, you've something of a guarantee that you've located the joint at a zero moment location. Not so in CIP. Thirdly, you have effectively perfect quality control on the construction of the joint. Fourthly, the precast system may be specified to utilize grout that is low shrinkage.

3) I came to precast after extensive experience with CIP. In retrospect, I wish that I'd started with precast because, in my opinion, precast design is a great way to become a ballsy, confident CIP concrete designer. Exposure in the precast world to alternate methods of dealing with things, and gobs of empirical evidence of system effectiveness, has been a game changer for me.

@TME: nice contributions with the keyway papers. Your collection on that topic must surely be approaching the world's best by now. The dam stuff might be marginally relevant here too. Different scale of course but, still, and unreinforced key after a fashion.

 

[TehMightyEngineer]

Your collection on that topic must surely be approaching the world's best by now.

I had a couple of years where I tried to do everything with dry shear keys that weren't match cast or grouted. I spent a decent amount of time trying to calculate their strength and eventually concluded that there wasn't enough research to justify my approach without empirical data. I couldn't sell my boss on testing these kinds of joints and decided it was overall a bad idea. I now generally use the more typical dry ship-lap joints or grouted shear keys.

The dam stuff might be marginally relevant here too. Different scale of course but, still, and unreinforced key after a fashion.

Yeah, the scale difference in the reason I didn't post it. Usually the dam shear key research is based around the assumption of huge compression pressures on the keys combined with many rows of shear keys providing redundancy. That said; if OP wants to dive into those there is additional research on shear key capacity for dam structures.

I wish that I'd started with precast

It's a good thought, but then you would be like me where "two-way slabs" and "post-tension" are effectively foreign terms. I'd say a healthy blend of both is really the ideal. You're spot on when you say it gets you to push your personal envelope more.

Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA, FL) Structural Engineer (IL)
American Concrete Industries

https://americanconcrete.com/