|Home » Forums Ľ Cabinet and Millwork Installation Ľ Message||Login|
You are not logged in. Consider these WOODWEB Member advantages:
Stair Treads Cupping5/22/14
Please see pictures of stairs and cupping problems. Treads are rift cut white oak; 3/4" x 68" x 15" at widest point. Sealed front and back then finished with lacquer. Riser is glued/pocket screwed to tread. Wood treads are glued to metal with urethane wood floor adhesive. Treads were in for about 10 days before problems started. Treads show 14% humidity. No unacceptable deviation in room temperature (72 deg.), surface temperature (72 deg.), or humidity (48%) was observed over a 24 hr. period. Instruments placed above and below stairs with another in the main room. Any ideas on how to keep this from occurring would be greatly appreciated.
I would like to see the cross section shop drawings of the tread's construction, including the metal stair base. That would assist in my analysis.
Like Rick, I would like to see the cross sectional drawings, showing how you planned to accommodate for the inevitable wood movement. I would bet this will get worse before it gets better. I also would bet the fix is not an easy one.
Unlike panel product, solid wood will move, and nothing - not even steel, will prevent it. The mid-point landing appears to be all solid, and I would expect that to expand and contract about 1/2" in our interior climate.
Your treads and risers both gained a bit of moisture from the time they were fabricated. They expanded, and broke loose their bonds with other members - steel, wood, etc.
Is that a water feature below it?
See attached section view.
That is a water feature, but it does not currently have water in it. Although it had water in it when the treads were installed and for about 3 days after. The problems didn't start until about a week after that.
Click the link below to download the file included with this post.
We left 1/4" between back of tread and metal riser to accommodate wood movement. Linear movement was expected, cupping was not. The landings are 5" tongue and groove. Some shrinkage but still acceptable gaps.
The drawing is no help - what is needed is the end section showing steel, wood stringer and wood tread/risers. And particularly, how you planned to deal with movement.
What you have is classic cross grain construction. A rigid stringer or horse with no movement, and solid wood right next to it. The solid wood is expanding - both treads and risers, and has no place to go other than cupping.
Look at the stair in plan and you can see what you have made is a board that is about 30' wide. Look in elevation, and you have the same, guessing 20' in width.
If you used Rift/Qtrd hardwood at 8%, the shrinkulator predicts you will have 4-1/2 to 8-1/2 inches expansion on the treads, and about 2/3 that on the risers. Dived that out by the number of treads and you will get numbers that describe what you see.
14% is awfully high - there must be some reason it has gained so much moisture. I'd need to look up what the EMC is in your area. If the room is 48% RH, that is also high - 30% being considered the high side of normal. You also need to know what the wood was when you bought it and used it - moisture content, that is. The water feature below is not it.
There is 'linear' movement, it is just expressed as 'cupping' since that is the only place the wood can go. A 12" wide board that cups is still about 12" when it cups. You have a lot of expansion.
You also say that the landings are 5" boards and they show some shrinkage (!?!). This is the opposite of what would be expected, since the treads/risers expanded.
Your best starting point is to state the MC of your lumber as it was fabricated - ask your vendor if you did not meter it on arrival, then shop conditions during fabrication, then what the environment was on site, especially prior to finish if finished on site.
Did you finish the bottom of the treads before install. It might be that it is picking up moisture in the air and causing the cupping.
Treads were finished all sides. Measured MC of an unused/unfinished tread and it was 8% - just were it should be. This is in an unconditioned shop with RH of 50%. I don't understand how the MC in the installed treads increased. The cupping has to be caused by a "drier on top and wetter on bottom" condition.
Did "all" of the treads cup? Are the landings ok?
Landings are fine. All treads cupped.
#1 What are those white/metal grills that go around the upper perimeter?
#2 What is that area that is below the stair? It looks like a pool or fountain.
I'm just trying to get a better understanding for the area that it is install in.
Also, could you post the left or right X-section showing the tread & riser construction as well as the steel dimensions below, the first section you posted didn't help much.
A few key points:
First and foremost, we know that wood does not move (change size or shape) unless its moisture changes. (Of course, we can also bend wood and make it change shape too, but at first glance, that seems not likely...agree?) So, we know that there has been a moisture change.
So, either the top dried out or the bottom gained moisture (or perhaps a little of both). So, what would make the top drier than the bottom (assuming that the pieces were uniform MC when installed)? Answer: It is hard to see anything that would make the bottom wetter than the top in such an open installation, except maybe the bottom pieces near the water pond. But even then, with a good finish and the finish being on all sides, it just doesn't seem reasonable. (Perhaps the bottom is covered, so that the bottom side has a different environment than the top side...cannot tell from the photos. If so, this bottom enclosure needs to be well ventilated to the outside to assure uniform humidity, top and bottom.
Second, we are told that the initial MC of the wood was 8% MC, which means that if the humidity were 42% RH, the wood would still be at 8% MC. We are told that the humidity is 48% RH, which would actually cause very little, if any moisture content change in the wood, if this was the RH for the past weeks. This RH value of 48% RH seems to be incorrect for a long time average because the original note indicates that the wood is now at 14% MC. This would mean that the wood was or has been exposed for a long time to about 80% RH. I must admit that this high RH seems unreasonably high.
The real moisture test is to use a pin meter with insulated needles and get the core MC. The core MC does not change very rapidly and so would give us a better idea of the past moisture history of the wood. (Getting this soon is better than waiting, as once the pieces have changed size, we know that the moisture content has already changed. We may have to rely on size changes and calculate the MC changes.
A pin meter also would allow us to measure MC gradients...surface, shell, core, etc.
Note that a change from 8% MC to 14% MC would result in about 2% swelling of the wood and such a size increase in width should be easily measured--around 1/4".
This swelling would affect the risers and if the risers are well fastened to the front of the tread, then it would lift the front of the tread, making it seem somewhat like cupping. I do believe that this is mostly what you are seeing...note that the joint between the TOP of the riser and the front of the tread is perfect...the riser has lifted the front. In fact, in one picture it looks like the riser is no longer vertical. Once the wood cups, it is very hard to return it to a flat condition, even if the moisture returns to a good level.
Do you have another tread? If so, can you pull out a defective one? Then it should be carefully measured for the amount of cup, size changes, MC and MC gradients, and so on. Of course, if you do the measurements, it would be self-serving and so someone might not believe you...have a third party do this.
If there is a chance that this will involve some legal activity, you need to consult with a lawyer right now to see what photos you need, what samples need to be obtained, what inspection by a highly qualified person needs to be done immediately, and so on.
Also, you need to find out if there is any safety hazard with the cupping and address that issue immediately.
One final possibility. If you take a piece of wood that you are holding fairly flat and wet the top surface, and if the top finish is not super waterproof, the water will wet the wood not he top of the piece and the wood will try to expand. However, it is held flat, so it cannot "crown" very much. In fact, the rest of the piece of wood will also try and hold the piece from crowning. The end result is that the wood on the top will be compressed so much because it would not expand, that the cells will duffer compression failure. This happens in a few hours. Now, when the top dries out from this wetting, it will try and shrink. Drying will be slow and shrinkage forces will develop over many days. These forces will exist so long that they will actually cup the wood. If you used a pinless moisture meter on the top of the wood, you would also see a high MC even though the average for the entire piece is much lower. This would also lift the front of the tread up and pull the riser with it, leaving a gap at the bottom of the riser...now this all seems to fit the pictures. So, were the tops of the treads wet mopped a few times with lots of water? Is there any evidence of this wetting (rusted fittings, water stains on the edges, etc.? How waterproof (to liquid water) is the coating presently? Maybe the coating is vapor proof but not liquid proof.
The treads are wood the risers are plywood(unless my eyes don't work anymore.)
Here's a crazy thought: The urethane floor adhesive is moisture cure. It sucked the moisture out of the back of the tread!!!
Adam, wrong direction!
Cupping is caused by differential. The pictures show shrinkage not expansion. The treads are laminated. They are not moving in a direction of growth rings. They are shrinking and the middle stays put while each edge moves to the middle.
Its not as simple as wet side vs dry side. There is a strong fastener(the glue) hold the middle down.
I am not sure about the lamination that Adam refers to, as they look like they are solid and edge glued, not laminated. Further, the grain looks to be flat sawn, so they would shrink or swell similar to solid wood...about 1% for a 3% MC change. Also, how could plywood be pocket screwed?
Note that moisture cure does release the moisture after the reaction has occurred. The moisture is not retained in the finish. So, moisture cure finishes would not create drying.
As I detailed, the pictures show that the bottom has expanded, or the top has shrunk, or a combination of these both. Can you please indicate how you can tell which of the three has occurred? I do not think we can tell. That is why it is critical to measure the size...size change will tell us for certain. But again, the question is why in an open construction would the top and bottom be different in MC...the humidity is the same top and bottom. If the RH was different, it would quickly adjust itself to become uniform by diffusion.
With the humidity being in the high 40% range and then the wood being at 14% MC indicates some wetting has occurred...wetting from an outside moisture source and not humidity.
I see where Adam is coming from about "plywood." The risers do look like ply. The risers appear straight. The treads look solid. It strikes me that the cup shown could be caused by either shrink or swell.
Ultimately it probably won't matter because any solution is going to be expensive and likely fall mostly on the fabricator. Bad design to start with. No heel clearances either. There is a reason that the tread nosing normally sticks out past the riser.
The finish is not moisture cure. I suspect the glue is moisture cure urethane. It absorbs moisture in order to cure(Bostik makes a lot of those for bonding wood to other stuff). I suspect it absorbed the remaining moisture from the tread and ply causing both to shrink.
Look at the gaps they are shrinkage not expansion.
The risers are ply. The treads are rift sawn strips edge glued(not laminated..my bad). The point being that those treads will not follow typical growth ring shrinkage rules.
You have a piece of stable wood & ply that is being dryed from one side and is fixed in the middle. Will it is cup up or down?
For arguments sake: let's pretend its a particle board tread. Screwed to a piece of steel only in the middle. If you wet it from the top. How would it respond?(forget about the risers).
Tim , At some point in the day is there direct UV or the sun hitting the treads ?
The pur adhevies do indeed (even hot melts) use moisture as a catalyst for curing. However, the moisture is released after the cure. Further, the mount of moisture used is small compared to the total moisture in the wood. Further, with edge glued pieces, the moisture would be the same top and bottom.
Also, note that the original posting says the wood MC is 14%...that is 6% wetter than manufactured, assuming the moisture reading is correct and not influenced by the metal framework.
Consider a "normal" oak floor that is put down with the pieces being tight against each other, edge to edge in the winter. When the humidity increases in the summer, the pieces try to expand, but cannot (except if the entire floor is floating). So, the wood gets a bit compressed. We can also see the edge to edge joints swell a bit upwards from this compression, giving the piece a look of being cupped. Now, in the next winter season with low humidities in a house, the pieces dry, shrink and there will be a gap between the pieces, even though the pieces are back to their original MC. Also, the cupped appearance persists. If the oak floor is wet-mopped or gets other sources of liquid water, this effect is even larger. This is a well known effect with wood floors and I do believe we are seeing that here as well. That is, the dry wood gets wetted, tries to swell, cannot swell, compresses, and then with refrying, warps. As mentioned, the wetting can be wet mopping of the top, in which case the problem is with the maintenance and not the initial installation...hence my suggestion that an expert look at the floor asap to see what moisture gradients exist, and other features that would show the pieces were wetted. Sunlight may have indeed accelerated the drying of the wet top surface, but sunlight alone would not create so much cupping.
Regarding the extension of the tread over the riser, it is my understanding that this is no longer considered a good practice because it has been shown that a few people (especially more elderly people who might not have full flexibility of younger people, so they drag their toe; also some handicapped) will catch their toe on this extension and trip. So, the flush design is now preferred. With this design where the one side of the stair treads are narrower, the lack of an overhang would be really important.
Lets go for the obvious, is the night time cleaners using a wet mop and leaving a lot of water on the treads?
Sorry guys. I have been off the internet for a couple days. Thanks for the response.
I strongly urge you to open the plywood bottom and look for any signs of water...probably from wet mopping. This would indicate that the bottom has expanded duet the high RH in the bottom compartment, which would indicate that the issue os a combination of bad design and poor maintenance and is not your problem, financially.
The moisture level of 14% MC also needs to be confirmed, as I mentioned, with a pin meter. If you could show that the bottom is higher in MC than the top (with a pin meter), you would indeed have proof positive that this is not going to cost you.
Iím sorry but solid wood over steel plate is akin to a waffle on a griddle. This is a big ďNo NoĒ and a major, architectural, design flaw. Even with full temperature/climate control, the slightest variations will effect wood and steel at extremely different rates. Plate Steel for example is a near perfect (instant) temperature conductor, while wood is the perfect insulator. This time difference creates constant temperature transference's between the two dissimilar materials.
With the transfer of heat comes the transfer of moisture,.. moisture drawn from the air and ending up between the wood and steel. This will rust the steel and all the metal fasteners and may even cause serious corrosion problems from interactions of wood acids and adhesives.
Even kiln-dried wood is wet, at about 14-18% and easily changeable. Daily variations in barometric pressure alone affect everything under heaven, both inside and out. The inevitable, slight temperature variation of an exposed wood face for example, and itís underlying surface (lying flat on the steel or concrete) is a recipe for disaster.
The answer was simpleÖ
And no, itís not wrong wood, bad glue, improper sealers, inadequate fasteners or anything like that at all (rift, white oak is the best solid-wood tread-stock to be had). Itís also to late for added sealants and coatings.
Face it,.. what you have now is not going to work or rather, itís going to continue to work loose. Iím sorry but itís only going to get worse.
I'm an old stair builder but what I've said here is not uncommon knowledge by any means.
Way late to this, but gosh.. That sent chills up my spine. What a colossally bad day that would be.
Architect has admitted design flaws. We are now exploring solutions. I am thinking about 3/4" engineered flooring.
Carpet will stick to steel.
An engineered floor is somewhat more stable than the sod you used, but if there is a moisture difference, they will warp just as you have already seen. The solution is not to change products but is to address the moisture issue...for example, do not wet mop and do not close in the back side. Note that the chance of warping increases with thinner material--if the oak treads had originally been two inches thick, you would not have posted here most likely.