Failure mode of window fasteners?

[structuresguy]

We are looking at some windows in an existing building which have been questioned as to adequate connection of the window frame to the structure. The frame is extruded aluminum, with a 1x wood board around the aluminum. The frame is screwed into light gage metal studs. The screws pass thru the aluminum frame, thru the wood 1x, and into the metal studs.

It seems to us that this puts a large eccentricity on the screws, approximately 2+". Basically we feel the wood does nothing but act as a spacer. However, it may keep the screws from bending due to the confining effect of teh wood.

Anyone have any insight into the failure mode of this condition. We can look at the shear capacity of the screws to determine if the windows have enough fasteners for the wind loads. However, with this eccentricity, some reduction of capacity must occur.

Thanks.

 

[ishvaaag]

As long the aluminum frame, wood frame and supporting stud remain tied together and without relative displacement, basically you have a screw in shear with the (purportedly small) concurrent tension that is putting the 3 parts together, i.e., there is no bending on the screw nor wind cycle widening (from bending action) of the hole of fixation to the stud.

So as you state this it becomes a matter of inspection. A number of windows comensurate to the total number of the same need be inspected, or at least any that shows the more reasonable appearance of having failed somewhat. Inspection will say if the elements stay together, are separate, if the contact with the screws and studs remains tight or the holes were predrilled to excessive diameters etc.

Don't forget to consider if the observed failures may be related to other causes, such weak studs, frames, or windows.

 

[bkeeler15]

Structuresguy,
I agree with your intuition that there is more than just shear loading on the screws, and think of this condition each time I review connection calculations associated with window systems. However, I have not been able to quantify the condition any further than you describe above.

Does anyone know of any technical data/literature/design aids for the condition structuresguy describes? I would think that the storefront/curtainwall industry would have some design data for this connection, but have been unable to locate any. I cannot believe that structural engineers design these connections based on nothing more than "Well, it usually works ok."

 

[ishvaaag]

Certainly the stone facing industry has and provides both literature and software -many times free- to sustains things at arm length and so on, so how much easily to the lesser distances pf a window frame setup. Think for example of the firm HALFEN. But of course these things come as proprietary systems that provide or understand something of enough strength is reciving the loads.

I think the biggest problem here is the dynamical load, since wind loosens the bolts of traffic signals in the roads and also many other things in the reach of its strength. The problem here may be that the resisting part, the stud is weak (to permit be joined) and maybe the screw distroys a bit the hole, so no reliable positioning as joined can be counted on. Even if calculating the screws with the stud somewhat separated is no problem, in all V and M, maybe tension, the problem of dynamically uprooting the screw is more related to the damage inflicted by the screw when going unto the stud than with this calculation.

If the screws are properly sized, maybe the solution may be reinforcing the stud or making these things bolts or something better engineered.

 

[bkeeler15]

I typically see the note "assume wood blocking prevents bending of the screw" on window connection design calcs. Can anyone out there say that they have designed window connections using this assumption? If so, please reply so as to enlighten me on the validity of this assumption. I'm still of the opinion that more than just shear is being carried by the screws.

That being the case, I agree that your existing windows pose a problem. Redesigning connections which have quantifiable load paths is the easy part. Proving 'adequacy' of the window connections may in fact involve proving that this "industry standard"(?) assumption is invalid.

 

[JCWilson]

structuresguy,

I think you are just scratching the surface of an entire industry of "standards" and assumptions that trouble many an engineer. I've been working in the industry and providing calcs for the very thing you are critiquing for over a year now (after working for years with heavy steel projects) and it has been tough dealing with these issues.

I'm sure you've heard these before:
"Well its worked okay for the past ten years."
"It never failed before."
"How can we do it that way if all of our competitors do it this way?"

That's what I deal with every day. It has been a real excercise in engineering judgement and sometimes a leap of faith, but I find myself sometimes ignoring or bs'ing my way through such calcs, just to appease the critics. I've been searching and searching for examples/testing/theories to back up some of the questions like you have, but have come up quite empty. I recently went through an exhaustive search about bending on structural bolts, but never found much that I could help me model them accurately.

If you truly believe there is a problem with these screws, then you have to push for it. But be prepared to hear a lot of what I hear every day. That's were you have to use your judgement to determine if there is any merit to those claims. I have found this to unfortunately be a VERY gray area of engineering with a lack of a lot of valid computations or research to back it up.

And getting back to your original question, I generally concur with Ishvaaag. Conventional structural analysis would never allow you to rely on wood blocking as acting in a structural sense in conjunction with steel. But light screws and window attachments and related assemblies are far from conventional. I have had to re-define failure in my mind because a simple bending check on a screw (or bolt) like this will not provide a result that will work.

I hesitate by saying this, but what is the worst that could happen if a screw holding in a window does "yield" in bending per conventional analysis? Probably nothing. I think it would fail (completely) only if it bent over and then failed in tension (provided the screws were not of an overly brittle material.) And that ultimate failure condition is going to be way above failure defined by normal bending checks.

Anyway, Good luck with your review! I hope someone else has had better luck with these problems and can help both of us.

And, please, go easy on "our" calcs.

 

[structuresguy]

Wow, well I guess I am not the only one concerned about these types of grey areas. It comforts me to hear about other engineers with the same problems. Like you JCWilson, I hear the "well, thats the way we've been doin it...." line all the time, sometimes even from the other engineers I work with. And that always drives me up the wall. If you can't give valid reasons for doing something, how do you know you are doing it right?

But like you said, in this case, what's the worst that could happen? Well, when its a 15 story condo with a couple hundred units or so, and all their windows are leaking or jambed, and the owners are all after the builder, I guess from the builder's perspective, the "worst" could be financially pretty bad.

If anyone does have some rational way for treating these window connections, please let us all in on it. Of course, you could keep it all to yourself and outdo the lot of us. But that wouldn't be very neighbourly, now would it.

Structuresguy

 

[jheidt2543]

Can you say Boston Hancock Center?