cupped tabletop

[Wayne Minami]

I am new to woodworking. This is only my fifth project.
I am making a trestle dining table out of walnut.
The top is 72X40 and is made of four boards.
When I glued them together, the top was flat. Now I have a cupping that is 1/4" deep in the middle on one side and 1/8" on the other.
I put a thin coat of brush on oil based minwax polyurethane on the top and a day later flipped it over and put another coat on the underside. The next day I noticed the cupping (on the topside, the sides are higher than the middle).
The temperature in my workshop varies from the mid 50's to low 70's. Since it is winter here in Virginia, the humidity is low.
Can anyone tell me what happened and what I can do now?
I do not want to rip the boards apart, rejoint, replane, and reglue.
Is there any alternative?
Is is possible to clamp the top and gradually increase the pressure to flatten it?
Or would doing this just split the boards?
Thank you.

[Mike Barney Sr]

How are the rings of the 4 boards oriented? The boards need to be oriented pith side, bark side, pith side, and so forth. The rings will flatten out as the wood dries.

Also, when finishing a larger panel, coat the bottom, or unseen, side first. Then flip it over and place on pyramids or support points and coat the top. By finishing one side and letting it dry you allow moisture to enter or leave the wood almost exclusively from one side, risking cupping. Letting the wood acclimate to your shop for a few weeks before preliminary dimensioning and a few weeks after preliminary dimensioning will help.

I assume your table top is at final dimension as far as thickness. It may be possible to rip saw kerfs in the bottom of the panel half the thickness every inch or so allowing the top to flex without cracking. you can then attach cleats to the bottom using slotted holes in the cleats to allow for movement. Glue thin pices of wood in the ends of the kerf and sand flush to conceal repair.

[Wayne Minami]

Two which are glued together go in one direction, and the other two go in the opposite direction. They are two bookmatched pairs. I kept the middle pair together and separated the other pair.
Did I make a mistake?

[Wayne Minami]

How are the rings of the 4 boards oriented? The boards need to be oriented pith side, bark side, pith side, and so forth. The rings will flatten out as the wood dries.

Also, when finishing a larger panel, coat the bottom, or unseen, side first. Then flip it over and place on pyramids or support points and coat the top. By finishing one side and letting it dry you allow moisture to enter or leave the wood almost exclusively from one side, risking cupping. Letting the wood acclimate to your shop for a few weeks before preliminary dimensioning and a few weeks after preliminary dimensioning will help.

I assume your table top is at final dimension as far as thickness. It may be possible to rip saw kerfs in the bottom of the panel half the thickness every inch or so allowing the top to flex without cracking. you can then attach cleats to the bottom using slotted holes in the cleats to allow for movement. Glue thin pices of wood in the ends of the kerf and sand flush to conceal repair.
Thank you for the info.
Should I leave the top with the kerfs open except for the repaired ends? Should I try to straighten the top with clamps after cutting the kerfs?
Thank you.

[Tim Sproul]

Don't worry about the cupping. When you fasten the top to the base, it will flatten out.

[Wayne Minami]

Thank you, that would really solve my problem.

[Chris Fournier]

Did you make a mistake? Likely but I don't think that it has anything to do with your lamination layout necessarily.

A lot of folks subscribe to the up/down/up grain orientation that Mike has suggested but I personally do not as it doesn't allow you to lay out for grain and figure as the builder sees fit. It also gives you the "wavy gravy" look of a corduroy road which no amount of restraint is goint to flatten - especially with wide boards like yours. Before you start kerfing undersides etc. I think that you have to back up a bit and find out a few things.

What is the moisture content of your boards? If it isn't 4% to 6%, (8% at the most) at this time of year then you've rushed the job and the cupping was entirely predictable. MC catches all of us eager WWkrs so don't feel too badly but do get yourself a moisture meter and be prepared to spend $150 plus. Rarely mentioned on forums but this SHOULD be one of any WWkrs first tools as it will save you from a ton of work and lost materials if you work in solid wood.

Another thing to look at is your joinery and this will be tough to do now that you have so much cupping. Were your edges 90 degrees to the face? Did you compensate for any error by flipping side by side boards as you ran the edges through the saw (91 degrees plus 89 degrees equals 180 degrees)?

How did you store the wood in your shop throughout the processing? Especially in your shop where the temp and consequently the relative humidity is swinging through 20 degrees (part of your problem), you need to ensure that both faces of any board or panel are exposed to the same sort of conditions - stickered and protected from drafts or direct heat/sun. This can be tougher than it sounds I know. In the end you really should have your shop environment controlled for humidity a bit more than you do at present. Temperature is the way to do this in the winter, dehumidifier in the summer.

Right now I have to recommend the following. You should rip the panel apart and get the boards down to the MC I mentioned, then machine them flat one more time. You can likely reduce the cupping a bit by orienting the convex face up and leave it exposed for a bit, it should dry out faster and the boards should flatten somewhat. Rejoint the edges as I mentioned and re-glue the panel. Your top will be just fine and only you will know about the lesson you learned. Kerfing the bottom surface will help you force the panel into submission but this is the wrong direction to take. A trestle table is pretty light and any forces in the top may be transmitted to the trestle and cause it to rack and twist. Also, you'll likely have to really tighten down the panel to flatten it, this will restrict its seasonal movement and it will tear itself apart.

There may be other repair suggestions forthcoming but what you are up against is a fundamental woodworking lesson that all of us learned when we were starting out. Some of us who rush to meet deadlines, expectations or are just too eager to see the final result "learn" this lesson over and over again - myself included.

[Wayne Minami]

Wow.
Thanks for the reply.
I will get the moisturemeter.
As far as taking the whole thing apart, I have to mull that one over. I know it is the best way to handle it, but the thought of that wll take some time to sink in.

[Chris Fournier]

Well I know exactly where you're coming from Wayne! The thing to remeber is that the table will be representative of your efforts for the rest of its existence. There are a few early projects that I'd like to have back if you know what I mean.

It's not that much work - really.

A long time ago I had two guitars that I was building for clients. I had installed a new to me type of truss rod which was buried deep in both guitars. I had made these rods myself and it involved brazing and L.H. and R.H. threads so that the truss rod would act in two directions.

I had not tested these truss rods prior to installing them. Classic mistake.

Just as I was getting ready to apply finish to these two instruments (and I could see the final $$$ payment) I thought that I would check out how well the truss rods worked. As I carefully tightened the first one the resistance grew, no change in the neck and finally as I carried on the truss rod snapped at my brazed joint. I felt sick. Could it happen to the second rod too? Of course it could, and it did. Now I was sure that I was gonna barf.

Normally I'd have grabbed the "offending" object and smashed the "stupid" POS to bits. This time I had the restarint required to tidy up, turn out the lights and leave the shop for a week as it turned out. I then came back and removed two glued on finished fretboards and figured out what went wrong. I ordered two new truss rods and got on with the project. The set back cost me two weeks but taught me some great lessons and the guitars came out as close to perfection as I will likely ever get.

Every time my customers bring the guitars back for a tune up or repair I relive that awful experience and oddly enough it makes me laugh and I'm greatful.

[Howard Acheson]

One thing to keep in mind is that a panel (or even a board for that matter) must be stored in a manner that allows air to get equally to both sides. Makes no difference whether the panel is finished or not. Store panels on edge up off the floor or by inserting stickers between the panel and the surface it is setting on. Never lay a flat panel directly on top of another flat surface.

[Tim Sproul]

I will disagree with Chris.

Your base should be proportional to your top. The members of the base supporting your top should be sized appropriately for the size and thickness of the top.

Take some scrap and mill some to the dimensions of the base. Screw your top to these. How flat is your top? How hard was it to drive the screws?

As with all components, there are tolerances. Of course, there are differences in what each of us considers acceptable tolerance numbers. And then there are tolerances that result in a functional difference. In the extreme, I can guarantee that you could take a cupped sheet of 1/4 inch plywood (say 1/4 inch cup like your tabletop) and flatten it with a frame made from 2x4's. You could also take a flat sheet of 1/4 inch plywood and attach to same frame. You couldn't tell which sheet was warped or not once attached to the frame.

My argument is that you are well within functional tolerances. Fasten it or remake your top flat. Once attached to your base, there is no difference. Of course, we are always aiming for perfection but we shouldn't let our pursuit of perfection prevent us from getting work done. Next table top, finish both sides at the same time.

[Chris Fournier]

I will disagree with Chris.

Your base should be proportional to your top. The members of the base supporting your top should be sized appropriately for the size and thickness of the top.

Take some scrap and mill some to the dimensions of the base. Screw your top to these. How flat is your top? How hard was it to drive the screws?

As with all components, there are tolerances. Of course, there are differences in what each of us considers acceptable tolerance numbers. And then there are tolerances that result in a functional difference. In the extreme, I can guarantee that you could take a cupped sheet of 1/4 inch plywood (say 1/4 inch cup like your tabletop) and flatten it with a frame made from 2x4's. You could also take a flat sheet of 1/4 inch plywood and attach to same frame. You couldn't tell which sheet was warped or not once attached to the frame.

My argument is that you are well within functional tolerances. Fasten it or remake your top flat. Once attached to your base, there is no difference. Of course, we are always aiming for perfection but we shouldn't let our pursuit of perfection prevent us from getting work done. Next table top, finish both sides at the same time.

How can you compare the resistance offered by a solid wood panel likely 3/4" plus thick to 1/4" plywood? You have even said it yourself, the base should support the top, it isn't there to take defects out of the top. Pre-stressed concrete, Ok, pre-stressed tressle table, No way. I agree that there are tolerances but I guess that I wouldn't accept these and you would. Fair enough I suppose.

[Wayne Minami]

I love a lively debate.
In the last 2 hrs I have sanded off the poly from both sides and the edges.
I put the top on supports running longitudinally in the middle and put wgt on the 4 corners to see if it will flatten out.
If is does, I will luck out. If it does not, I have to consider either Tim, Barney, or Chris's solutions.
Being a newbie, I have to try and find out these things.
The top is 1" thick by 72 by 40, so it is substantial. My trestle has two stretchers, so is also pretty strong. I had planned to use walnut buttons to attach the top. I do not know if they will be strong enough to uncup a tabletop of that size.
Thanks for all the advice.
Wayne

[Chris Fournier]

I love a lively debate.
In the last 2 hrs I have sanded off the poly from both sides and the edges.
I put the top on supports running longitudinally in the middle and put wgt on the 4 corners to see if it will flatten out.
If is does, I will luck out. If it does not, I have to consider either Tim, Barney, or Chris's solutions.
Being a newbie, I have to try and find out these things.
The top is 1" thick by 72 by 40, so it is substantial. My trestle has two stretchers, so is also pretty strong. I had planned to use walnut buttons to attach the top. I do not know if they will be strong enough to uncup a tabletop of that size.
Thanks for all the advice.
Wayne

There certainly are plenty of opinions on any given topic.

Another suggestion is that you take a look at Bruce Hoadley's book Understanding Wood. It's very helpful to us woodworkers.

By treating the top as you have I'm assuming that you hope to force the panel flat, this requires that the fibres of the wood to slip or tear away from each other and take up a new arrangement, this is somewhat unrealistic unless you are using heat, ususally wet heat like steam. Even with steam you have to apply an awful lot of force to change the wood fibres natural state.

[Wayne Minami]

Chris
I suppose what I am trying will not work. But, I thought I had to give it a try.
The top warped in no time at all, so, I figured it may straighten with a little persuasion.
I will check it tomorrow, and assuming it does not straighten, I will recut the top, or try some other solution.
I will hit Amazon for the book.
Thanks.
Wayne

[Mike Barney Sr]

Two which are glued together go in one direction, and the other two go in the opposite direction. They are two bookmatched pairs. I kept the middle pair together and separated the other pair.
Did I make a mistake?

I think I know what happened. You say you bookmatched the boards. Can I assume you resawed 2 thicker boards into 4 boards? If so, then you exposed the inside of the thick boards which are higher in moisture content. These sides are all the top surface of your table and when drying caused the cup. When resawing you need to let the boards stabilize for a few weeks before dimensioning.

[Wayne Minami]

Mike
I bought 4 boards from a lumber dealer in Pa. They are 2 sets of bookmatched pairs.
I do not know when they were resawn. If it was done recently, your explanation certainly would be the cause. You are right, all of the inside surfaces are on the top (finished) surface.
I had them plane and joint them S3S.
I have gotten so much helpful advice. I am grateful for the assistance.
At this point, I am mulling my alternatives.
You live and learn.
Thanks.
Wayne