Subscribe or Renew Membership Subscribe Renew

Cracking window panes

chazas's picture

Owner here.


Our windows are Weathershield true divided light casements.


Within weeks after they were installed, three exterior panes in sashes on the south face of the house cracked with no apparent cause – in the same odd pattern, like an inverted “U”.  Weathershield is replacing these sashes.  They sent 2 of the 3 replacements, though they got lost in our supplier’s warehouse for three months and when they did arrive they were the wrong color.  New ones are supposed to arrive next week.


Now, another similar crack has appeared in a different sash, also on the south face of the house.


Clearly, something is wrong here.  Multiple window panes should not just crack with no apparent cause.  I can’t imagine it’s anything other than a manufacturing defect.


We spent over $70,000 on these windows.  I do not want to be constantly having cracked windows on an extremely expensive house – even if Weathershield replaces them as they break, it would be a huge hassle.  The only foolproof solution I can think of is for them to replace every sash in the house now.  I can’t imagine this will go over too well with our contractor or with Weathershield, and the very thought of it at this stage of construction makes my head hurt.  Any thoughts or other suggestions?


(post #77417, reply #1 of 19)

Triple pane?


(post #77417, reply #2 of 19)

What kind of rough opening were the windows set in?

If the ROs were too small, I could see the shrinkage of the opening as the wood dries as a cause for the breakage.

Compression of the entire window frame could break the glass.

Can you remove the interior trim to check it out?

(post #77417, reply #3 of 19)

Double pane windows.  The rough openings were not in any way too small - the windows were shimmed into the appropriate places.  This appears to only be happening on the south-facing windows.

(post #77417, reply #4 of 19)

This appears to only be happening on the south-facing windows.

Sounds like the dreaded (to glassmakers) "thermal expansion."  What may be happening is that the window maker got a bad run of glass from their glazier (whether separate companies/divisions, whatever).  It does not take very much "out of spec" for the frames for the double-pane units to be too tight when made up at the factory.

The heat coming in fro mthe south heats the outside pane more than the inside one, just like it's supposed to, but there's not enough "give," and the stress has to go somewhere.  Goes into the frame & sash normally; abnormally, it goes back into the glass pane.

This can be a bit more dramatic when this happens on a 60x120" 300# unit a couple-five stories off the ground . . .

Occupational hazard of my occupation not being around (sorry Bubba)
I may not be able to help you Occupational hazard of my occupation not being around (sorry Bubba)

(post #77417, reply #6 of 19)

Compression of the frame would make the windows hard to open (if they are the type that can open). That's easily checked.

George Patterson, Patterson Handyman Service

George Patterson

(post #77417, reply #7 of 19)

....or the Argon, Kyrpton, Carbon dioxide, or whatever is leaking out and ambient air molecules are too large to get back in, leading to spontaneous breakage. Does the glass have a concave reflection when viewed, especially from the outside? Another way to tell is put a straight edge across the face of the glass.

(post #77417, reply #5 of 19)

Is there any sort of reflective surface within about 30 feet of the windows that could be focusing the sun on them?

This inverted-U crack, does it go from corner to corner, or start on the bottom or the sides? Inner pane or outer pane? You say these are true divided light -- what's the pane arrangement and which one is cracking?

Where do you live?

So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin

Of all the preposterous assumptions of humanity over humanity, nothing exceeds most of the criticisms made on the habits of the poor by the well-housed, well-warmed, and well-fed.  --Herman Melville

(post #77417, reply #8 of 19)

These are called stress crack in the industry. It is described as a spontaneous crack in the glass with no point of impact. The most common causes are glass out of spec from glass manufacturer (Cardinal Glass I believe), too small of a gap in between the glass and the frame it sits in, but the most common cause is the glass installed wrong side out. The Low E coating should be on glass surface #2. There are 4 glass surfaces in a double pane IG. They are counted from the outside in. So the outer surface is #1 the inner surface of the exterior glass is #2 the outer surface of the interior pane is #3 and the interior surface you can touch from the inside is #4. If the Low E is on surface #3, the UV rays from the sun will reflect out to surface #2 and stay in the airspace longer. This will heat up the space, expand the gas or air and then break the seal or the glass. You can test this with a flame. Hold a lighter near the glass about 1/2" away. You will see 4 reflections of the flame in the glass. One flame will be a different color, usually purple. Then you can see if the Low E is on the correct surface. I hope this helps.


"It is what it is."

(post #77417, reply #9 of 19)

The windows are not hard to open. I'm in the Washington DC area.

The windows actually are not low E (the house is passive solar, we didn't want to block anything), so the wrong-side coating issue shouldn't be a problem.

The cracks are primarily in the third floor windows - one on the ground floor, though it's (barely) conceivable that that one is attributable to the alarm sub. The windows are side-by-side casements, four panes to each sash. The cracks are primarily in the bottom pane, though the latest one is one pane up from the bottom. The cracks are in the outer pane. A pic of the window arrangement is below.

Sounds as if the most likely issue is the size of the glass used in the manufacturing process. Maybe the issue is occuring mostly on the third floor because of the difference in temperature between the shaded and nonshaded portions of the windows? It didn't happen during our (abbreviated) winter - the one that broke this week was the first problem since last fall, which also would seem to support the expansion theory.

We are trying to get a tech from Weathershield out to the site but I don't want to just get dismissed out of hand. Thanks for all the advice - it should help. Any further advice continues to be appreciated.

Edited 4/25/2007 9:00 am ET by chazas

(post #77417, reply #10 of 19)

Photo Sharing and Video Hosting at Photobucket

(post #77417, reply #11 of 19)

What's on the roofs of those bump-outs? Shingles, metal, rubber? Are there heating panels there?

So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin

Of all the preposterous assumptions of humanity over humanity, nothing exceeds most of the criticisms made on the habits of the poor by the well-housed, well-warmed, and well-fed.  --Herman Melville

(post #77417, reply #12 of 19)

Galvalume standing seam roof - no solar panels there. 

(post #77417, reply #13 of 19)

That may be at least partly the problem.

So convenient a thing it is to be a reasonable Creature, since it enables one to find or make a Reason for everything one has a mind to do. --Benjamin Franklin

Of all the preposterous assumptions of humanity over humanity, nothing exceeds most of the criticisms made on the habits of the poor by the well-housed, well-warmed, and well-fed.  --Herman Melville

(post #77417, reply #16 of 19)

Another possible cause (though D.C. is hardly high altitude) could be collapsed glass. Andersen offers a "High Altitude Breather Tube" option on their windows for those typically headed for Colorado. This allows the window to "fart" when it needs to blow off excess pressure. I wonder if the heat isn't changing the pressure inside and doing something similar.

I can't believe in this day and age that anyone would put in windows that aren't LowE. There are different types for northern climates and those for southern, making it a no brainer.<$10 a window saves about 25% on energy costs.

(post #77417, reply #17 of 19)



"It is what it is."

(post #77417, reply #19 of 19)



You are very welcome.  I strongly suspect that you have a thermal stress issue with these units – no matter where located – and I also strongly suspect that changing to tempered or heat-strengthened glass should solve your problem. 


As I mentioned previously, annealed glass (with an unblemished edge) is “good” up to about 3000psi or so.  Tempered glass is required to be 10,000psi or better by code.


Heat-strengthened glass (although tempered is by definition “heat-strengthened”) is  between tempered and annealed in strength characteristics - being typically somewhere between 6000 and 8000psi.  Heat-strengthened glass is not a safety product (unlike tempered which is a safety glazing), but it is about double the strength of regular annealed and may be a possible option for your situation as well, if you are so inclined.


Also, I understand about not wanting to retrofit to LowE.  When I wrote the reply I was at the end of “one-of-those-days” and it does sound like I was apparently carrying some of that into the reply.  I still would always recommend LowE in every possible situation, but the comment about your situation simply wasn’t needed in my reply. 


Wango1 and ChicagoMike,


In this particular case I would suggest that collapsing glass is not a contributor to the breakage – although it certainly can be in other situations.


From an altitude consideration, as you pointed out Wango1, there just isn’t much of change between central Wisconsin (where the windows were manufactured) to Washington DC – certainly not enough to matter concerning breakage of the glass. 


Although heat inside the IG can be a problem in certain situations; in this situation the inside airspace of the IG simply cannot heat up enough to be an issue.  While this does sound somewhat counter to what I stated earlier concerning thermal expansion causing the cracks; in essence, it is virtually impossible for the interior of the IG unit to get hot enough to cause the glass to “bow” from interior pressure to the point of cracking because the heat would dissipate (conduction) through the glass and be radiated outward before it could reach the required temperature/pressure. 


What you suggest can certainly happen, but not in this specific instance. 


Thermal expansion of the glass is the likely cause, in this situation - uneven thermal expansion of the lite (meaning part of the lite is trying to grow while part of the lite is attempting to remain static), can definitely cause cracks in windows – just as described.


Chazas, I would suggest that when you look at the crack, from where it starts at the edge of the sash, the crack looks to be exactly perpendicular (90°) to the edge of the sash.


Bill, I agree 100%, this is a great thread with loads of very good information in it.

I have very much enjoyed reading other responses and participating in this one!







(post #77417, reply #14 of 19)



Glass cracks in order to relieve stress.  Thermal expansion in glass is no problem so long as the glass has (a) room to expand and (b) expands evenly across the surface of the lite. 


Thermal expansion in glass can be a bad idea and can result in stress cracks (a) if the glass is restricted in its expansion and (b) the glass is not expanding evenly across the entire surface.


I am leaning the same way that I think DanH is leaning – it sounds like you are getting uneven thermal stress on the glass possibly caused by reflections from the metal roofs of the two bump-outs combined with shading from the roof line.  


A single crack, without “branching” indicates a low-stress break, or that less than 1500psi was needed to propagate (but not necessarily cause) the crack.   In general, a simple low stress crack requires a flaw in the glass in order to grow.  This type of flaw is often a chip or some other defect in the edge of the glass – a defect that would not be visible when the glass is in the sash.


As a basic rule, ordinary (not heat-strengthened) glass should be able to withstand stress to about 3000 psi without failure.  Temperature-induced stress in glass is generally computed at 50 psi for each degree temperature variation within the lite – so that a difference of only 30 degrees between the warmer and the cooler portions of the glass would be needed in order to produce sufficient thermal stress in order to propagate a thermal stress crack.  Again, with the caveat, that it is highly likely that the glass would have to have a (relatively) significant edge flaw in order to crack at the lower stress level.


Heat reflected from the roof panels onto the glass, unevenly, could very easily create temperature variations well in excess of the 30 degrees difference needed to grow a crack in the lite.  I would strongly suggest that the reflected heat could (and very well might) cause you problems with your window glass in the future. 


The inverted U indicates that the edge of the glass was likely cooler than the center of the lite and that the stress crack was following a line of “least resistance” parallel with the edge of the sash but along the line of heat variation.


One fix might be to use heat-strengthened glass which will have a much greater resistance to heat-related stress than will ordinary annealed.


If you choose that option then I would also strongly suggest that you include a LowE coating in the IGU since clear (uncoated) glass is always less energy efficient than is LowE coated glass.  You are going to lose much more heat than you are going to gain in winter thru that massive expanse of glass, and in the summer your heat gain thru the uncoated glass is going to prove to be a significant expense as you run your air conditioner way beyond what would be needed if the glass was coated. 


Massive expanse of uncoated glass was the passive solar “holy grail” back in the seventies, but in today’s world it is well understood by most energy experts that the correct use of LowE coatings on windows is far and away a much better solution – even in solar heat gain conditions.





Edited 4/25/2007 10:03 pm ET by Oberon

(post #77417, reply #15 of 19)

Thanks for your well-considered thoughts.

The galvalume on the porch roofs below may be an issue - but it's hard to imagine, as the cracks have occurred mostly in the windows between the porches, where there does not appear to be a significant amount of reflected light.  We also had one crack on the first floor.

My partner (the architect) had already suggested replacing the top floor sashes with tempered glass - glad to see the agreement.

A digression in response to yours:

It's a bit too late to switch to low e windows now that they're all up!  Aesthetics prevent having different window glass in different parts of the house.  Anyway, we have 2x6 construction and R-20 Icynene in the wallls, R-48 cellulose in the attic, and very few windows on the east and west exposures.  The sunlight infiltration is very deep into the house in the winter (where it was very comfortable on sunny days, even without the attic insulation in) and drastically reduced the closer we get to the summer solstice, just like it's supposed to be.  With 10 kw of photovoltaic and high-efficiency heat pumps, I think we'll be good to go.  You can't have everything...

(post #77417, reply #18 of 19)


   You've gotten some great information from some people who seem to know their stuff!!  It's been very interesting!!

   I'd like to pose a concept dealing with the center windows cracking:  With the slanted shiny roof angles from the pop-outs, could it be that these angles are reflecting the sun on the windows as it travels across the sky?

   Having built and learned from an early 80's solar batch water heating project, the rear reflectorhalves were curved in such a way to focus the sun on the side and back of a 40 gallon tank from almost sun up to almost sun down.  Your roof angles seem to reflect  the sun on the far window in the morning, both at high noon, but as the sun angle changes, the heat is reflected to the opposite window for the remainder of the day.

   Maybe covering the outside of windows with some temporary attached aluminum foil would highlight the roof reflected sunlight and help you see any sunlight concentrations.


Edited 4/27/2007 12:23 am ET by BilljustBill