header clearance under sliding door?

[LonnieP]

I've got an issue with a client (architect) who is trying to lay the blame of a sticking residential sliding glass wall on me alone.

Beam spans 50'-0" and has a floor mounted (top guided) sliding glass wall system under it. The beam has deflected enough that the 1/2" the glass installer allowed has been used up and the glass is sticking in the tracks around midspan. The glass installer set his upper track before all the dead load was in place and has since (last two years) used up all the adjustment space he had. The installer was the one who decided 1/2" was enough space and he didn't check it with anyone else.

I figured the glass installer would allow about 4" for adjustment, like they do with top hung folding partitions. Looks like i was wrong. Question, is there an industry standard for upper track clearance for bottom supported sliding glass walls or is this something that should be discussed between the installer, G.C. and the SE. In retrospect, I'll put a note on the drawings that will start that discussion on future projects.

 

[ron9876]

I'm not aware of a standard. Is the beam deflection within acceptable levels?

 

[MiketheEngineer]

Well, a pickle.

What was the originally calculated deflection. I would have used that and added some for "give". Did you know it was going to carry glass sliding walls??

1/360 seems appropriate - but that might not be enough under these circumstances. I use 1/600 for masonry, tile, etc.

Any simple way to "jack up" the beam and stiffen a bit??

 

[msquared48]

I sense there was no communication between you and the installer on this? That, if so, is the sticky wicket.

You should never, but never, "assume" these things. Verify, verify, verify, and in writing.

Mike McCann
MMC Engineering

 

[haynewp]

Check the project specs and the manufacturers installation instructions and see if there is anything about installing after dead load.

 

[rowingengineer]

If your beam is within acceptable deflection limits as per mike for the usa, than you need not do anything more than state this to the client.

Just because the installer isn't smart enough to know the beam will deflect has nothing to do with engineers. The builder should be in charge of this and if experienced (as required by most building regs) should have been able to handle this without issue. Also the architect should have been experienced and included this in his window spec, I would just pass the buck back.

I don't communicate with installers, I see nothing sticky about us designing to reasonable limits and everyone following suit.

ANY FOOL CAN DESIGN A STRUCTURE. IT TAKES AN ENGINEER TO DESIGN A CONNECTION.”

 

[Brad805]

Do your general notes stipulate your design criteria? We cannot be expected to talk to every trade that shows up on site that is somehow related to our structural components. Nobody wants to pay those kind of fees.

I am not surprised you did not talk to the glass installer. Unless this is high rise, your job was done long before that trade showed up to site.

I can't imagine the glass installer is a full time commercial guy. Any commercial trade we deal with would know that a 50'-0" piece of iron will deflect more than 1/2".

Did the architect ever ask you how much the beam might deflect? What does his detail show? Were you provided his detail? Who approved the shop drawing for the glazing? It is his job to specify the glass requirements, not yours.

Brad

 

[msquared48]

Brain freeze here - "installer" did not register. Fridosos I guess.

I normally design to a 1/2" total deflection limitation, depending on the span ratio, for headers over windows and doors. That lends to some larger beams than normal, but keeps me out of situations like this.

Communication, if you know the supplier, helps. But, I agree, that proio communication with the installer is not the norm. However, as you suggest, a note to the installer as to what limitations you set in your design would be helpful.

Mike McCann
MMC Engineering

 

[JLNJ]

It seems to me that 1/2" on a 50 foot span is not reasonable. The glass guy needs to allow for more.

 

[rowingengineer]

1" wouldn't have me reaching for the phone.

ANY FOOL CAN DESIGN A STRUCTURE. IT TAKES AN ENGINEER TO DESIGN A CONNECTION.”

 

[boo1]

Nanawall installation instructions:

http://media.nanawall.com/files/SL25_OwnersManual.pdf

"The structural integrity of the header is critical for proper operation. Vertical deflection of the header under full live and dead loads should be the lesser of L/720th of the span and 1/4”. Structural support for lateral loads (both wind load and in the stacking area when the panels are stacked open) must also be provided.

THE ROUGH OPENING SHOULD BE LEVEL, PLUMB AND SQUARE AT ALL POINTS. THERE SHOULD BE NO UNEVENNESS OR BOWING. MAKE SURE THAT THE HEADER IS NOT TILTED OR TWISTED. THERE SHOULD BE NO BUMPS ON THE SURFACE WHERE THE SILL WILL SIT. THE SIDES SHOULD BE IN THE SAME VERTICAL PLANE AND NOT OFFSET OF EACH OTHER. A TRANSIT AND OTHER SIMILAR PRECISE MEASURING EQUIPMENT SHOULD BE USED

For better performance .... all dead loads such as upper levels, roof, etc. be constructed before a unit is installed."

 

[rowingengineer]

Boo1,
This is great, did the architect inform the engineer of this? is this for all systems of similar build? what is full deal and live loads? Should an engineer have to change his design to suit this specific door, or should the door work to the engineers design?

In my opinion this is a cop out statement by the door manufacturers. If they want a door to run 50' on a header than they need to have more adjustment in there system.

ANY FOOL CAN DESIGN A STRUCTURE. IT TAKES AN ENGINEER TO DESIGN A CONNECTION.”

 

[emmgjld]

Wouldn't this be similar to installing partition walls under roof trusses? Deflection must be accounted for by the trades or the General Contractor (who is usually in charge). The Structural is responsible to keeping deflections at or below the normal for the particular element.
I have had this discussion when extra interior loads suddenly (magically?) appeared and the footing/pad settlement was excessive.

 

[LonnieP]

Well - - emmgjld put his finger on it. We had a typical non-bearing wall to roof framing detail that shows room for 1/2" vertical movement. If that is where the installer got the 1/2" from then my goose is cooked. This kind of detail seems to infer I've managed to limit live load deflections to 1/2", doesn't it? life lesson: watch out for typical details!

As for the beam in question it easily met all applicable deflection criteria including L/360 for plaster. Just not the 1/2" for the remaining 1/3 of the dead load that was added after the upper track was set.

The Nanawall requirement of 1/4" or L/720 gives the installer ammunition for his argument. Seems the best course for the next project is a note on the drawings to have the installer talk to the design team before setting the upper track. Communication is usually good to have and would have stopped this problem before it got started.

 

[MiketheEngineer]

L/720 - not even sure the mill can roll a 50' beam that flat??

 

[haynewp]

That should be overridden in the contract documents. It is not practical.

 

[MiketheEngineer]

1/720 is less than 7/8'' in 50'!!

I don't know many carpenters/steel workers/even engineers who could measure that...

Maybe with a good laser .....

 

[boo1]

What was your contract requirement?
Does the beam element meet these requirements? (What is the measured defection?)
Don't the architect, GC, steel supplier, steel erector and glazing installer all share in the liability??

Since glazer installed the glass wall out of sequence (before all loads were applied), I would think that the glazer and GCs have primary responsibility.

 

[LonnieP]

boo1

There was no special contract requirement other than the building code. We didn't even get the architectural details untill just before backcheck. The beam easily met all code defelction criteria. I think GC and installer should share in remedial costs. Not so sure about the Arch. and other trades. The out-of-sequence issue is the most troublesome. We'll see what happens.

 

[a2mfk]

The beam spans 50ft, and there is ONE sliding glass door under it? That is a big ass door. But seriously, everyone else understood the original post but me? Can you describe the structure and the door a little more? One standard 5ft sliding glass door? Lots of doors for lots of residences or this is a house, condo, high rise, etc?

How does anyone know how much the beam deflected?

If its one door, the amount of overall beam deflection is not very important, its the amount of change in the sliding glass door opening, which should be insignificant on a standard door.

I need more info but this does not sound like your problem unless the beam has a major deflection issue.

I look at residential doors and windows all of the time and people try to blame structural movements on their poor operation. Rarely is it the structure...