Fire Rating & Precast Column Corbels 1

[KootK]

I've got a precast column corbel in a parking garage with a required fire rating of 3HRS. And I'm interested in the fire rating requirements for:

1) The beam bearing plates on the corbel below and, more so;

2) The primary corbel rebar welded to the bearing plates which, somewhat by definition, doesn't have a ton of cover.

I assume that there's some "out" for not having to fireproof the bearing plates but I'm unsure about the corbel reinforcing. Do I need to use 2" bearing plates just to get the rebar cover for 3HR? Does that even make sense given that any heat on the exposed bearing plate is likely to transmit that heat straight to the rebar anyhow?

Rationally, I feel that these element do not need to be fire rated the same way that primary members do because the odds of fire getting to and locally taxing these elements seems unlikely. That said, I'd like to see an exclusion in the codes rather than making this stuff up myself. Does anybody know of such an exclusion? Or can anyone confirm that I'm out to lunch in thinking that such an exclusion exists?

 

[GC_Hopi]

You can get a copy of the PCI, Design for Fire Resistance of Precast Prestressed Concrete. Based on 1988 version, they say no special fire protection is necessary for the type of bearing you described. I dont have a current reference though.
EDIT: IBC 704.2 does state that column connections to other structural members need to achieve the fire resistance rating. Also is it necessary to weld the primary reinforcement to the bearing plate?

 

[KootK]

Thanks so much GC_Hopi.

Also is it necessary to weld the primary reinforcement to the bearing plate?

That's how we're getting the bar anchorage at the moment.

 

[WARose]

I assume that there's some "out" for not having to fireproof the bearing plates...

I'd doubt it.

Do I need to use 2" bearing plates just to get the rebar cover for 3HR? Does that even make sense given that any heat on the exposed bearing plate is likely to transmit that heat straight to the rebar anyhow?

2" ain't gonna make it. Granted my thermo is a little rusty....but I did some calcs a few years back with steel as a insulator.....and I can promise you that (theoretically): that heat is going to get into the rebar (over 3 hrs) if attached to the plate.

I'd use a UL coating all around to protect it.

 

[GC_Hopi]

You might have to use the old ACI welded bar detail for the anchorage connection then that would give you the option to lower the bar and get some cover on it.

 

[KootK]

Thanks again gentlemen.

I'd doubt it.

For purely bearing, there definitely seems to be an out. The argument seeming to be that, for a purely bearing element, there's not catastrophic outcome if the bearing plate melts or whatever. However, used as corbel anchorage, this does seem to fit the definition of things that need to not melt under fire conditions.

and I can promise you that (theoretically): that heat is going to get into the rebar (over 3 hrs) if attached to the plate.

Agreed.

You might have to use the old ACI welded bar detail for the anchorage connection then that would give you the option to lower the bar and get some cover on it.

I may do just that, thank you for the suggestion.

As is often the case, part of what I'm up against here is precedent:

1) The detail shown above largely replicates those of local competitors who are using it, without UL coatings, in environments requiring fire rating.

2) A very analogous condition is shown below, straight from the PCI manual. To say that detail is not possible in an environment requiring fire rating seems to me to be a stretch given our collective faith in PCI to know their stuff. You don't seem many precast beams in environments that need no fire rating.

 

[GC_Hopi]

I will preface this with "I could be wrong" but that snip appears to be a precast beam with a bearing connection. There is some lateral component, Nu, but that is likely resolved thru friction at the bearing (I see no indication of a welded connection). So I see this as purely bearing and therefore the embed plates wouldn't need to achieve the fire rating.

 

[WARose]

I don't know what those (4) #5's are doing (I assume it's to develop the Nu force) or the (5) #6's......but if they are seeing load only under wind or seismic (something that you hope will be over by the time the fire starts).....I'd think that would be ok.

Now if they are stressed on a day to day basis under Dead or Live.....that's a different story.

 

[KootK]

..but that snip appears to be a precast beam with a bearing connection.

Oh, it is a beam bearing connection all right. However, I contend that the dapped beam end is pretty much a corbel turned upside down. In my opinion, whatever fire rating requirements are applicable to corbels would also be applicable to dapped beam end bearings.

I see no indication of a welded connection

See below.

There is some lateral component, Nu, but that is likely resolved thru friction at the bearing

I don't know what those (4) #5's are doing (I assume it's to develop the Nu force) or the (5) #6's

They are the primary tie for the primary compression strut coming out of the beam (see below). They're primary function would be resisting gravity loads and they would be exactly analogous to the primary tension bars in a corbel.

So I see this as purely bearing and therefore the embed plates wouldn't need to achieve the fire rating.

That's just it. If this is "just bearing" the so is the corbel in my opinion.

 

[GC_Hopi]

They are the primary tie for the primary compression strut coming out of the beam (see below). They're primary function would be resisting gravity loads and they would be exactly analogous to the primary tension bars in a corbel.

Ahh...of course. Then I am not so sure if the detailing will work to achieve a fire rating. The anchorage plate of the tie is exposed to the fire and will weaken. Maybe if you do the calcs you can see what kind of rating you get. Similar to the corbel, to achieve a higher rating it might be best to move the tie to get adequate concrete cover or add a firestop sealant at the joint.

 

[WARose]

They are the primary tie for the primary compression strut coming out of the beam (see below). They're primary function would be resisting gravity loads and they would be exactly analogous to the primary tension bars in a corbel.

So the question then becomes: does it say specifically in the PCI manual that the beam detail you posted is good for 2-3 hrs? Your logic seems to be that because it's in their manual....it's got a good fire rating. Not sure I buy that. Furthermore, I've seen the people who do precast panels fireproof their exposed (steel) connections with spray on stuff.

 

[TehMightyEngineer]

KootK: I've always wondered how people got fire ratings for those dapped beams to corbel connections. Thankfully I've not had to design one of those for a building yet. Kind of annoying to hear that you're finding that it's not something people address and thus you have to step outside the "norm" to do what's "right".

I suppose the only thing I'd add is the exposed area of steel that can be heated directly is very low compared to the concrete thermal mass. It's been a couple of years since I ran fire rating calcs but I recall they were not only based on clear cover but also thermal mass. I believe there's an argument to be made that even if you heat the plates they will just act as heatsinks transferring the heat into the concrete. You even have nice long rods to carry the heat deep into the concrete, keeping the steel cooler than if it was just sitting there alone. Likely still a weak point but probably not a fire rating of 0 hours.

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

 

[GC_Hopi]

It does sound like fire rating of connections/joints are largely ignored or written off in the literature. Page 12 from PCI discusses the heat sink idea Link
I am surprised PCI has not done any testing on this topic. Does any one have a current copy of PCI, Design for Fire Resistance of Precast Prestressed Concrete? If so what guidance is given there.

 

[TLHS]

Can you justify the capacity of the corbel using an collapse prevention load case and just those lower ties that aren't welded to the bearing plate? I would look at the load combination in K.1.5 of S16. 1.0D+0.5L (assuming this isn't a storage area) might get you in the reasonable range.

 

[TLHS]

You could also justify moving the bearing point over in the extreme case, probably, if you have somewhere to take the resulting moment, like a slab.

 

[KootK]

Maybe if you do the calcs you can see what kind of rating you get.

It's been a couple of years since I ran fire rating calcs...

Hold. The. Phone! Is there an analytical method of determining connection fire resistance that is available to designers?? If so, do tell. I'm familiar with the rational method for beam, slabs, etc but I was not aware that there was a procedure available for connections.

Similar to the corbel, to achieve a higher rating it might be best to move the tie to get adequate concrete cover or add a firestop sealant at the joint.

The trouble with moving the tie and using different anchorage detailing is that the connections become congested and quite uneconomical. As you can imagine, I strive to keep my precaster client competitive with their local competition who are not doing such things.

Does any one have a current copy of PCI, Design for Fire Resistance of Precast Prestressed Concrete?

I do as of this morning. Or, at least, a stupid Adobe Digital Editions DRM version. There really isn't anything in there other than what was in the older version and that wall panel document that you linked. Verbatim on the connection front. If you'd like to see a particular section of the new document, let me know.

Can you justify the capacity of the corbel using an collapse prevention load case and just those lower ties that aren't welded to the bearing plate?

I can use a load combination similar to the one you mentioned but making a go of that using the lower ties is unlikely. The loads are still very high and those bars are pretty small and at a reduced lever arm. And it's not like I'll be having #8 ties etc.

You could also justify moving the bearing point over in the extreme case, probably, if you have somewhere to take the resulting moment, like a slab.

You've suggested this as a way to reduce the stress on the tension tie, right? I get that but the whole strategy then depends on there being a "rational method" of determining the fire resistance of a connection. I've got my fingers crossed that there is and that somebody can direct me to that. This is one of those cases where I'm seriously hankering to be poorly informed.

So the question then becomes: does it say specifically in the PCI manual that the beam detail you posted is good for 2-3 hrs?

It does not say that. However, the overwhelming majority of these connections be installed without UL coating and will occur in structures requiring some fire rating. This leads me to believe that either a) folks are not worrying about fire rating these connections or b) there exists some rational method (calculation). for determining fire resistance.

Your logic seems to be that because it's in their manual....it's got a good fire rating. Not sure I buy that.

My logic is as stated previously:

As is often the case, part of what I'm up against here is precedent

When I look around me I see:

1) Other experienced precast engineers using these connection details, without applied fire proofing, in situations requiring fire ratings.

2) PCI illustrating these same connections for members which will almost always require fire ratings. And doing so without mention of applied fire proofing.

In my mind, when it's my dissenting opinion against that of the rest of the precast universe, it's logical for me to assume that they are probably right and that I am probably wrong. Note that this thread exists only because I have specifically questioned the validity of what appears to be normal practice.

 

[KootK]

This detail from the PCI manual also becomes an interesting case. As above, if you melt the angle, the welds, or the welded dowel, the beam falls down under gravity loads. Yet this detail, without applied fire proofing, is ubiquitous in situations requiring fire ratings.

 

[WARose]

When I look around me I see:

1) Other experienced precast engineers using these connection details, without applied fire proofing, in situations requiring fire ratings.

2) PCI illustrating these same connections for members which will almost always require fire ratings. And doing so without mention of applied fire proofing.

In my mind, when it's my dissenting opinion against that of the rest of the precast universe, it's logical for me to assume that they are probably right and that I am probably wrong.

I think the only answer is a thermo calc. The PCI reference posted earlier is just too generic to satisfy me. (How may bars attached to the plate? Exposure of plate? Etc.) That PCI uses this detail is unpersuasive. (For starters, how do they know how many hours the user will need?) What is used in practice isn't persuasive to me either. (When I first came into this business, common practice was still to use the Nelson stud manual for anchors....that's history.)

I may dig out the thermo book tonight and see what I can dig up. I don't know that the surrounding concrete is such the great life saver. Seems to me like the steel would send a lot of the heat on rather than losing a great deal. (Think of a wire with plastic insulation.) But I gotta lot of rust in this area. Will advise.

 

[TehMightyEngineer]

Is there an analytical method of determining connection fire resistance that is available to designers?

ACI 216.1 has analytical methods and is what I've used in the past. Granted, it doesn't really cover what we're trying to do here but it's got lots of useful analytical equations for heat penetration into concrete.

As above, if you melt the angle, the welds, or the welded dowel, the beam falls down under gravity loads. Yet this detail, without applied fire proofing, is ubiquitous in situations requiring fire ratings.

You're probably not meaning "melt" literally but I just want to make sure. As I'm sure you are aware, fires rarely get hot enough to melt steel [insert tired joke about "jet fuel can't melt steel beams"). What you get is a loss of tensile capacity in the metal but steel will still have very high compressive capacity even at high heat. Thus a bearing connection like this isn't that bad at high temp. What I'd worry about it the loss of capacity of the horizontal tie welded to the bearing plate/angle, wouldn't take much heat for that weld to have half it's strength.

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

 

[hokie66]

Too much faith in PCI, I reckon. And in the precast industry in general.