Untopped Hollowcore Diaphragm with Openings

[DayRooster]

I have asked this question before without others capable of explaining much. So I’m going to crowd source because I keep seeing this detail on other engineers designs.

Disclaimer: this is a lateral stability question not a vertical gravity load question.

For an untopped diaphragm (see attachment), how are the local chord forces (from the lateral diaphragm) handled around a large opening? The reference load is the uniform load draw in red (-X direction). And the local force in question is in blue circled area. If I am not mistaken there would be tensile force pulling the planks apart at this location along the y-axis across the face of the opening. Also the hollowcore planks are grouted together. I understand the grout has shear capacity (forces parallel to the joints) even when cracked. But under this cracked condition the grout would have zero tensile capacity (forces pulling the planks apart perpendicular to the joints). Anyone done this before and have an detailed explanation for it?

Also I am searching for an explanation beyond “that’s how it’s always been done”. I have heard that enough times without any to support the answer.

 

[JLNJ]

Do you have a continuous chord across the open side of the opening? Maybe reinforcement in a bond beam?

If not, you are relying on relatively low forces and some tensile capacity normal to the joints. I’m sure these forces exist, but not at ultimate EQ stress levels.

 

[DayRooster]

JLNJ,

There is nothing across the open side of these openings for lateral resistance beyond the hollow core grout. The only thing that I see placed along the opening edge is a shelf angle that is vertically supports the split planks and is only supported via a bearing only clip angle each end. See video for a good example:

https://www.youtube.com/watch?v=Rx41heJm67E

. The only bond beams would be around the perimeter of the hollow core plank slab where the masonry wall or steel structure supports the hollow core planks. That is the only location where tension can be resisted to my knowledge. No clue how the local tension at the opening face is being addressed though...

As for your second point I am curious about it. I have never seen this type of design in seismic or high wind regions. In seismic regions, it seems that hollow core designs always place a topping slab on the units as a normal procedure. That being said, I know these grouted joints crack over time though. Would they still have some small tensile capacity even if the joints have cracked due to shrinkage and temperature forces over the years? I have always considered it to be zero force (conservatively) but I would be interested to hear an explanation in the other direction though.

Btw, thank you for talking the time out of your day to post too.

 

[bones206]

If I was concerned about the local forces at the header, I'd probably put some screw anchors into the plank along the length of the header, and through the support ears at each end. But with low enough forces, I imagine bearing friction might be helping the planks from pulling apart, along with the tensile capacity of the grout in the joints.

The header shown in the video has a rebar welded from end to end. I wonder if that helps the header act as a chord/collector. The grout plugs in each core would act as shear lugs, transferring the chord force into the grouted rebar as a tensile force. Although I think the actual purpose of the rebar is to be a spacer, so grout cover can be provided at the exposed plank ends for fire protection of the tendons.

 

[DayRooster]

Bones206,

I agree with what you are saying. But in all of the designs I have seen (by others) there are no screw anchors though. Also, to my knowledge, there is no anchors ICC-ESR approved to work in the very thin sections for hollow core planks. I should also add that typically the rebar is not present too. That was the best video I could find of the situation. It just so happened to have rebar for that product. Most often it is shelf angle without the rebar welded into it. See attached for the shelf angle without the rebar in it:

https://www.spancrete.com/application/files/7815/3013/0979/Hollowcore_Openings.pdf

.

Overall, I do agree that is COULD be some form of friction and/or tensile capacity of grout. I still have a concern if the grout cracks due to shrinkage and temperature though. As for numbers to back this up, I am still at a loss. We speak in generalities that the loads would be small in this situation. I agree they would smaller since we are in non-seismic regions without high winds. But what is the threshold for these types of structures? (question not directed at you, more of a global thought). I have seen it used for a variety of structure sizes. If the only methodology for validating it will work, is simply engineering judgement without even a back of the napkin set of numbers to justify it then I left bewildered.

 

[Brad805]

I would look for research by SK Gosh. He has made many presentations on un-topped hollow core diaphragms in seismic regions. I am not sure Piekko is overly keen on un-topped diaphragms.

I doubt your typical opening is coordinated with the standard HC dimensions, so I expect most openings involve some level of forming where you can detail rebar. The side rail shape do not lend to the use of anchors easily.

 

[DayRooster]

Brad805,

I will look into this research and see if anything is relevant. Again it should be noted that this is for non-seismic regions. Also, I am personally not keen on this detail either. Just trying to do my research to understand why it is being done before I make any firm decisions with my engineering judgement.

Also, you are correct, from what I can see the openings are not coordinated with the standard HC dimensions. It should be noted that some are even larger than the typical 4'-0" wide planks. Also, I imagine you could place rebar at these locations. Just haven't seen anyone do it. Where would you suggest rebar to handle these forces?

As for the side rails and hollow-core, I agree they do not suit anchors well.

 

[DayRooster]

Brad805,

I found four articles from SK Gosh on un-topped hollowcore diaphragms. All were well written articles but none of them mentioned openings in the diaphragm for un-topped roof planks. All of the research was based on un-topped diaphragms without openings. Seems to be a big gap here since roof openings are quite common feature for roofs.

That being said, I would love to hear SK Gosh's opinion on this subject since he seems to be very knowledgeable. Anyone have a connection to plant the seed?

 

[Brad805]

I would reach out SK Ghosh and see if you get lucky. Another option is to ask PCI for their opinion. If they cannot answer I suspect they will provide a contact that may have more knowledge on the topic. I recently reached out to PCI about another topic and was pleasantly surprised how helpful that engineer was.

Un-topped HC diaphragms seem to be evolving. I myself do not do much work with HC's, but some of the products we work with are used as diaphragms. The last webinar I attended hoping to gain a better understanding focused largely on the ASCE load requirements and offered little on many of the nitty gritty details. I don't quite follow this trend for commercial applications. I have yet to see a HC without pre-camber and I can only imagine how much levelling compound one might need for a project if you have a large space that needs a tile floor.

 

[DayRooster]

Brad805,

I sent a message to PCI hoping for a detailed response. I'll keep you posted. Also, I am looking into how I could get a message to SK Gosh too. I agree that it is worth a shot.

I appreciate you responding to the message though. If anything I hoping to get this topic more exposure. I have heard "this is how it's always done" too much without explanation. Only one person in the past has expressed similar concern (KootK). I feel that the only way to evolve on this topic is to force the question "but why". To me it seems like a gap and potential concern within hollowcore design. I could be wrong and hope that I am wrong. Hence the reasons for doing my research and crowd sourcing again. Finally, I would not be surprised if there was some low low low limit where these large openings fly with untopped hollowcore planks. Just trying to nail it down since I am grasping at thin air with calculations on this topic.

 

[molibden]

Why don't you do a FEM analysis of the floor and see how it behaves under in plane loads? It is same question with CLT floors or any other rigid prefab elements construction. You need to tie them together.

 

[HTURKAK]

Mr. DayRooster (Structural)(OP),

The picture that you have posted does not give any idea for the SFRS , so do not have any idea how the hollowcore diaphragm is shaken.
I screened the previous responds and i think you have the documents below ;

-

https://nvlpubs.nist.gov/nistpubs/gcr/2016/NIST.GCR.16-917-42.pdf

( Seismic Design of Cast-in-Place Concrete Diaphragms, Chords,
and Collectors A Guide for Practicing Engineers )

-

https://nvlpubs.nist.gov/nistpubs/gcr/2017/NIST.GCR.17-917-47.pdf

( Seismic Design of Precast Concrete Diaphragms A Guide for Practicing Engineers )

I will suggest you to look the following books

- SEISMIC DESIGN OF REINFORCED AND PRECAST CONCRETE BUILDINGS ( By Robert ENGLEGIRK )
- Precast Concrete Structures ( By K. ELLIOT ) This book is useful to see how to develop ST Model and concept for pc Diaphragms ).

You could develop a ST model and design the diaphragm connections in accordance with the PCI Design Handbook,
Chapter 3.

 

[MIStructE_IRE]

My gut feeling says this is nonsense. If you want a reliable diaphragm, pour a structural topping in my view.

 

[DayRooster]

Molibden - I might try to model it in FEM and release the tension capabilities, worth a shot to see what happens. Again I feel the grout can transfer shear only after it cracks (no tension). Also from what I have seen the only spot where the hollow core planks are tied together is at the exterior walls. Otherwise they are typically only held together with low strength grout, with nothing extra at the openings with regards to lateral forces.

HTURKAK - I have designed global diaphragms per PCI’s handbook. My problem doesn’t occur with the global forces. My problem is with the local forces around large openings when the openings intersect the grout joints. The PCI Handbook doesn’t provide any guidance around the method to calculate local lateral forces around these openings. I will check out your other references to see if they discuss methods to handle untopped local diaphragm forces around openings though…

MIStruct_IRE - I’m with you on this one. In my area (Midwest USA) there are a lot of untopped hollowcore diaphragms being used. I usually only let them fly if the openings are very small. But when the openings get large (based on my judgement) then I place an additional topping slab as the diaphragm. Even with this judgement I still get questioned as to why I would do it this way when others have allowed untopped hollowcore diaphragm with these large openings. Hence the reason for the crowd sourcing and research. I’ve hit my threshold on this subject and want solid answers…even if it turns out I was being too conservative…

 

[MIStructE_IRE]

I hate that one, “everyone else lets us do it this way”. I’ve come across that one plenty of times and more often than not its because (some of) “everyone else” doesn’t understand it enough and have been lucky that its worked!

Unfortunately I don’t have any specific guidance on this particular topic, but it feels wrong to me.

 

[DayRooster]

Brad805 - I just received a response back from PCI. I have removed the responders name for privacy. Below is the response...

"PCI’s role is to provide guidance to precast producers and specialty engineers regarding precast design and fabrication, we rarely take a firm stance on any design decision where engineering judgment is available. For an untopped diaphragm, the chord forces around an opening greater than the plank width could be accommodated through additional reinforcement or additional structural steel. There is presumably support for the bearing of the plank along the opening, which could be combined to provide the chord force. There are also instances where the top flange is locally removed to install and develop reinforcement along the boundary. There are options that the hollow core manufacturer had likely experienced before. The attached NIST document on precast concrete diaphragms may be useful to you."

So it sounds like PCI is backing off this topic. Also, the NIST document provided " NIST Seismic Design of Precast Concrete Diaphragms - A Guide for Practicing Engineers" does not address openings either. Also, from my experience precast suppliers typically defer the lateral diaphragm design to the EOR and only take responsibility for the vertical load selection. I have only seen once where a precast supplier designed a diaphragm; but they requested to use double-tees with welded attachments and top flange reinforcement in the double tees (maybe a reason they did not select un-topped hollow-core planks).

Also, as for removing hollow-core top flanges and installing reinforcement, I am familiar with this practice but I am not seeing many other people elect to do this on un-topped hollow-core diaphragms. I am only seeing the shelf angle supplied that is not anchored to anything. In addition with all the different opening sizes it could easily become problematic to resolve the local tension force into the global system too.

Finally, I have been playing around with FEM modeling some examples. I have been releasing elements and making grout links shear and compression only. Without any chord reinforcement it almost appears as if the structure is developing a hinge at this points and its turning the diaphragm into two separate three sided diaphragms instead of one four sided diaphragm. Still reviewing the results but it might be the break through I am looking for to justify some of these designs. It is a very complex design scenario though. I still hope for something more practical to come about that is easier to justify with calculations. That also gains acceptance amongst the community for being "standard" design protocol for this situations. I was hoping for PCI to help more but I may need to reach out to some local suppliers to see if any of them are willing to discuss this topic with me. The search continues...

 

[struct_eeyore]

If I wanted to be a cowboy, I'd extend the flat leg of that support angle all the way across the planks, and tapcon it generously all the way along. If I wanted to be cruel to the field crew, I'd have them drill horizontal holes perpendicular to the span, pass a few threaded rods thru, and secure them at each end making sure to grout the last couple of cores (and whatever elese) so that the tension doesnt punch thru the sidewall. But I'm neither of those, so I would just complain to the architect that we need an extra 2 inches of topping, or we cant make this work. Also, I'm not sure what your spans are, but I'm leary supporing more than one plank width on an angle like that.

 

[DayRooster]

Structee - I completely agree. It’s just the push back I get that “other engineers don’t do it”. I am drilling down into that mindset to determine if there is any merit to it. I currently require topping slabs unless the openings are very small (little pipe vents only). Even then I situate them to not interfere with joints.

Btw, I liked that little journey you took me on with your post. Made me smile at the end.