Roof Joists - Thermal loads

[slickdeals]

Folks,
How are roof joists designed when the deck above is subject to thermal expansion?

Is there any published literature on it?

 

[msquared48]

How much relative expansion are you talking about?

Mike McCann
MMC Engineering

 

[slickdeals]

I have an exposed metal deck that is attached to the joists. I was wondering how thermal expansion/contraction of the deck is transferred to the joists.

 

[MiketheEngineer]

Assuming a 150 degree temperature change and joists at 2'-0'' on center - the change would only be .024''. Don't think I would worry about that??

 

[civilperson]

The deck wrinkles and the joists stay in place.

 

[msquared48]

As the strength of the metal deck spans across the joists which will inhibit "wrinkling" of the panels in that direction, and considering that the swing of the contraction/expansion is probably closer to 75 degrees rather than 150, I would not worry about it looking at the previous numbers. Parallel to the joists, the deck has room to move at the deck panel joints.

If these are wood joists, the screws will bend some where the panels attach to the joists. If they are metal, then maybe use slightly slotted or oversize holes and thru-bolts if you are concerned. Personally, I would not be.

Mike McCann
MMC Engineering

 

[MiketheEngineer]

And don't forget - if using similiar materials for both joists and deck - lengthwise - they will move the same - or closely the same.

Here in the midwest - 150 degree swing is not unheard of - esp if you include summer sun beating on a darkened roof. I have seen -17 up to 105 ambient - with attic temps going to almost 140 degrees!! Obviously - attic was not well ventilated.

 

[slickdeals]

Am I incorrect in assuming that the elongation of the joist is really from a heat transfer through the deck? If the deck sees no temperature change, the joist won't either. The joist is most likely in a conditioned space.

Temperature from the roof deck is transferred into the joists through its attachment.

I agree that the decks are weak perpendicular to their span. I am not really concerned about the expansion, rather the thermal force (axial stress) associated with the restraint to expansion at their supports.

 

[MiketheEngineer]

Yes - the joists will need some "wiggle" room if they are subject to high thermal changes. I have seen I-beams bust through the end of concrete walls. Installed too tight and the temp changed about 100 degrees and the beam was about 150' long.

 

[slickdeals]

Mike,
Are you then suggesting that I should have some sort of strees relief in the joist bearing?

This is for a project in the middle east and we are concerned about thermal loads.

If the support can deflect a little and take a little, then the joist should just arch up and down relative to the change in temperature, right?

 

[msquared48]

Middle east...OUCH! I hope you don't have to inspect it!

It depends on the framing scheme here in part. If this is the top of an exterior wall, more than likely the wall will move to relieve any thermal elongation or contraction. If you have hear walls or frames that tare transverse to the bearing wall, the wall movement will not be as great.

If you wanted to weld the ends of the joists to steel plates with slotted holes that are mounted to the wall, this should solve your concerns.

Really though, unless there is a lot of insulation in the roof structure, some of the heat from the steel deck will affect the top chord of the joist, probably not the bottom chord as much. Your elongation/contraction will primarily be with the top chord here. If the joists are top chord bearing, then use the connection described above. If they are bottom chord bearing, I wouldn't worry about it.

Mike McCann
MMC Engineering

 

[slickdeals]

Yes, OUCH !
I have a top chord bearing. The ends are bearing on a long span truss. The joists are perpendicular to these trusses.

I have a joist seat attached to these trusses and the joist top chord was to be welded, but I was a little concerned.

Any thoughts? I guess I will have to study a modified slotted hole connection at the joist bearing.

The joists are not spanning more than 35' though.

 

[ChipB]

Personally, I don't think I'd be too worried about. Maybe I'm overlooking something, but this is how I see it:

The thermal expansion forces would have to be transferred to the joist through the connection of the metal deck to the joist. The deck is spanning perpendicular to your joist. If you are using 3' deck, it's going to try to grow from the center connection, both ways, along the length of the joist, equally. So, if you drew a FBD of the forces stemming from the center of the deck, it appears to be placing the top chord of the joist in tension.
But.....
Adjacent to that section of deck, is another section of deck. If you draw the FBD of the forces from that section, the half of the deck directly adjacent to the first deck, is equal and opposite of the mirrored section, thus cancelling out. If you carried this all the way out to the end, you'd be left with only the tensile forces developed when you were looking at the first section, which shouldn't be that much. Plus, remember the top chord of your joist is in (usually) compression. Adding a little bit of tensile force would reduce the stress in the top chord.

 

[slickdeals]

ChipB,
I agree with your rational. However, what happens if the deck is attached during the higher temperature months. The deck will now shrink inducing more compression in the top chord. Right?

 

[ChipB]

Yes, it will induce slightly more compression. But the force will be small. Whatever one panel can induce.

I still wouldn't worry about it. The compressive force under your DL + LL combination will be high, and we all know, they rarely if ever see the Live load we design for.

 

[civilperson]

Work it backwards, what force or stress can change the length of the joist top chord equal to the amount of length of change from temperature change? This stress is additive to the DL and LL flexure stress and the total should only be of concern if it approachs the limiting stress.

 

[ChipB]

I do need to footnote my "rarely if ever see the live load we design for":
Industrial applications notoriously hang all sorts of stuff from the bar joists (cable trays, duct work, chilled water lines, etc.). Because of this, I try to use KCS joists and increase my live load in the utility aisles.