Hollowcore Roof w/ 3" Topping Slab - Sequence of grouting and topping

[bones206]

I'm working on my first job with hollowcore planks as the roof diaphragm. I'm working out connection details between the planks and collector/chord elements that tie into the top of the CMU shear walls. My details have loose hooked dowel bars that get set into the plank cores at the supported ends, or notches cut out every 4 feet along the sides of the plank.

My question for those with hollowcore experience is are these dowels typically grouted in at the same time the keyways are grouted (i.e., chord/collector elements are grout)? Or, is it more common to grout the keyways only, then come back and cast the chord/collector elements with the topping slab?

In this particular project, the CMU shear walls have concrete tie beams at the plank bearing level and there are also CMU parapets.

 

[KootK]

1) In my experience with this, which is extensive, the bars would get grouted in by the precast erection crew along with the keyway grouting.

2) What's driving the need for the topping in this situation? There are sometimes extenuating circumstances that necessitate the topping but, for most applications, I consider it an expensive waste. From a structural only perspective, the only solid reason to force a topping, in my opinion, is if collector issues at re-entrant corners etc are impractical without the topping. And, even then, there are things that can be done.

3) There's a rather awesome system, from the perspective of constructability, that can be used with parapets:

a) Build walls up to plank bearing elevation with dowels sticking out the top. Bond beam below as collectors/chords.

b) Grout bars into the plank keyways that stick out past the ends. Turn up the ends such that they will project up into the parapet.

b) Build the parapet.

4) I actually like intermittent steel clip angles for the sidelap connection but that's a tough sell to anybody that's not a structural engineer. Those "bash a hole & grout a hooky thing in" details are pretty much un-assessable as far as I can tell.

 

[bones206]

Thanks for the feedback KootK. I figured that grouting the hooky things would make the most sense from a constructibility standpoint, but I was concerned it would reduce the strength of the connections. On the other hand, the low slump of the grout probably reduces the likelihood of voids and poor consolidation inside the cores.

I've definitely struggled with the un-assessablity of the connections and have spent way too much time researching this topic. I did find the Spancrete Research Notes to be helpful and gave me some basis to rely on.

Point taken on the topping slab. It's probably overkill, but it is what it is. The structure I'm designing is part of an industrial facility with several similar structures being designed by multiple firms. They are all using topping slabs so we are just going with the flow. The design wind speed is close to 200 mph so it'll act as ballast if nothing else. The structure is also housing equipment that can go boom, so it'll add some extra degree of robustness.

I had thought about the detail you suggest with the dowel projecting up into the parapet, but I thought it might interfere with the parapet block if they don't space it correctly. Since the Ldh is the same either way, I figured a 180 degree hook around the chord steel might be a better way to go.

Another question on sequencing: is it more common to build the parapet atop the plank, then place the topping slab, or topping first then parapet on top?

Thanks again

 

[KootK]

I'd expect the parapet to come ahead of the topping. It gets the masons off the job sooner and the parapet can be used as the topping pour stop. It hard to say for certain though as the logistical landscape is a little different on every project.

Turning up the dowels definitely messes with the parapet block. Since the dowels go into the plank keyways, there's really no realistic way to game the spacing.

 

[mronlinetutor]

My first job is in Hollowcore concrete panels too. I just follow the book and did the calculation and the panels were fine. I was a junior engineer and I am not responsible for everything. I went out to the job site to help erection. What happens is that there are columns that go through the Hollowcore panels and under the columns, the Hollocore panels have to be solid underneath. Then there are support columns again. But the columns were not placed symmetrically on the construction plan. So the sense of direction of the Hollowcore panels has to be well aware otherwise, the solid section underneath the column is not there and the column would punch through the Hollowcore causing a collapse. The Hollowcore was placed in the wrong direction and the solid section was missed. But luckily the structural engineer checked the solid section by putting a nail through the slab and find it was missing and the problem was rectified.

disclaimer: all calculations and comments must be checked by senior engineers before they are taken to be acceptable.

 

[bones206]

Yup, the devil's in the details. Some of the other firms designing similar buildings on this project show the wall dowels field-bent and anchored into the topping slab. That detail didn't feel robust to me, so I opted to anchor into the planks instead. Still doesn't feel all that robust, but at least the connections eccentricity to the diaphragm is reduced.

 

[JLNJ]

We do a lot of schools and don't top the roof planks. The key is to keep the layout one-directional so that the camber is all working together. You don't want the 2" mid-span camber adjacent to a zero-camber plank bearing end.

At the bearing end you can bend the bars up into the parapet before the topping goes on. If you don't like jamming a bar in the keyway, you can saw at the core and get good purchase on a longer or bigger dowel. Since many planks are 48", the dowels will be on a modular masonry dimension. Even if they don't line up with the cores, at least they are misaligned with the same offset the same at every bar.

Admittedly, the plank buildings we do are mostly low-rise in low-seismic regular applications. If you have a hard-working non-rectangular diaphragm then all bets are off. Dragging forces perpendicular to the plank span requires some forethought, especially if untopped. I did a non-regular dorm where the dragging of forces in and out of the goofy diaphragm was a nightmare. The building won the architect a few awards and gave me nothing but lost nights of sleep.

The sidelap details are always iffy. Let the plank deflect and rely on some slotted clip connection to fasten the entire endwall onto the building and transmit the diaphragm shear? Can be calc'ed as KootK says, but these always seems a little tenuous to me. Plus you need guys who can bolt and weld on your field crew.

I tend to lap the precast over the sidewall and have them (gently and carefully I'm sure) "bash a hole" to make a positive connection. We use drypack grout under the plank after the topping (if at an intermediate floor) goes on to make up for the camber and then let the plank's small live load deflection warp it between the last joint and the wall.

 

[bones206]

I noted on my detail that the side notches shall be made by the precaster, hoping for some better quality control than bashing holes on site. But is that a bad idea? I suppose it would be better to make the notches after plank installation, so they aren't being handled by a crane with reduced strength. Maybe it would still be ok, but I don't want to place any onerous requirements on the precaster that aren't typical.

My structure is a very simple, one-story rectangular box divided into 3 sections by full-length shearwalls. But... it's my first rodeo with hollowcore and the wind loads are significant, so I'm spending some extra effort on the connection details. I feel like the information out there from the hollowcore industry is pretty opaque and would like to see these connections codified in some manner, so I can have more confidence calculating capacities.

 

[KootK]

Let the plank deflect and rely on some slotted clip connection to fasten the entire endwall onto the building and transmit the diaphragm shear? Can be calc'ed as KootK says, but these always seems a little tenuous to me. Plus you need guys who can bolt and weld on your field crew.

No slots and dry pack between plank and wall. Grout the last plank core and use something like a Titen HD into both plank and wall to avoid welding. I'm even cool with the shallow drop in anchors so long as there's no vertical demand.

I tend to lap the precast over the sidewall and have them (gently and carefully I'm sure) "bash a hole" to make a positive connection.

That's the iffy connection that gets precast erectors winking at each other at the company BBQ. 1/3 to 2/3 of those just fall apart during install depending on erection crew competence / whether or not it's Friday. The other wrinkle to this connection is that spancrete is very specific about the dimension of the lap and the placement of the dowel within it. And for good reason. At the dims specified, there is some hope of a competent connection. Fiddle with the dimensions much though, and you're crossing the first or second web with the dowel and a breakout is nearly guaranteed. And almost nobody sticks to the Spancrete dimensions religiously.

I noted on my detail that the side notches shall be made by the precaster, hoping for some better quality control than bashing holes on site. But is that a bad idea?

I wouldn't do it. I doubt that the plant crew would do any better of a job than the erection crew.

I feel like the information out there from the hollowcore industry is pretty opaque and would like to see these connections codified in some manner, so I can have more confidence calculating capacities.

I hear it. I've searched English speaking planet earth pretty thoroughly for some decent design guidance on these connection and there's little out there to be found from what I can see. That, in contrast to the 20 pages of typical connection detials that you can get from any precaster. Lot's of clever details... hardly any that can be quantified.

From time to time here, you'll see some well meaning engineer try to put capacity to CIP wall / CIP slab joint. You know, 8" wall + 8" slab. Turns out even that's kind of hard and you have to play some games with the shear friction requirements. Now go and eat up 3.5" of that lapping joint with plugged hollowcore, and now what do you have? It only makes things worse, right?

 

[bones206]

Ever try delegating the connection design to the precasters (with calcs)?

 

[KootK]

1) Most of the time, when I'm working with hollowcore, I'm actually doing it as the precaster. And I dread it when the EOR delegates these connections to me for just these reasons.

2) Technically, I believe that precast to base structure connections are the responsibility of the EOR. Of course, that doesn't make the connection design any easier for whoever winds up doing it.

3) Most jobs go out with the stock "dowels in keyways at 4' oc" and that's the end of it. Using the Spancrete information, one finds that many basement walls topped with plank ought to collapse inwards as a result of insufficient connection strength. And that, without even considering combined diaphragm demands. Good thing soil pressure is a government hoax.

4) Most precasters will want to swap out the Spancrete dowels drilled through the plank for dowels installed in the grouted keyway joints. The keyway dowels are much more economical to install. Unfortunately, to my knowledge, no publicly available testing has actually been done on that joint typology (I don't know that any has been done at all).

 

[bones206]

I've actually changed up one of my details based on this discussion. Decided to put the bars in the keyways on the supported ends rather than grouting them into the core. Partly because of constructibility and partly because of this Spancrete research. My logic: if testing has shown that #4 bars grouted into the keyways will reliably yield in bending, then I can have some confidence that the keyway can develop the yield strength required for shear friction at my wall connections. I feel better about that than hoping the grouted core doesn't yank out the end of the plank.

 

[KootK]

So to do shear friction up right here, on a vertical plane, it would require a small diameter hooked bar with a minimum 6" extension into the wall beyond the plank ends, right? Since the planks need 3" bearing, that means you'd need 12" block to get the job done. Do you have 12" block or are you making the play that a small diameter bar is developed upon being wrapped around a perpendicular bar? I'm not challenging your detail so much as just curious what you're up to. Folks interested in the finer points of this as few and far between.

 

[bones206]

I do have 12” block. But I’m also detailing the 180 degree hooks to wrap around the chord steel as cheap insurance. Not taking any credit for the hook-around though. For similar reasons I made the chord bar a single beefy #6 rather than two smaller bars. It’s less congested and *feels* like it gives the hooks better purchase. That’s all gravy though... hopefully.

This job is in Florida, and concrete tie beams and tie columns are commonly used in masonry construction there. One nugget of wisdom I’ve learned is that people sometimes make the tie beams wider than the walls in order to get that elusive Ldh. Seems like a huge waste of concrete, but it’s one way to meet code if you are using smaller block.