The Ecstacy and the Agony

[Dovetail]

Earlier this year I posted this: https://www.woodworkforums.com/f40/ta...binets-167229/

Three months after completing the task, I was about to head north for the winter and noticed the front of the top was not quite flush on the top rail and considered I had been too gentle on the attachement screws - so I pulled them up a bit to correct the perceived bow.

On returning some four months later I was devastated to find gross bowing/warping of the

in situ.jpgMeasurement.jpg

I had selected the boards for the top particularly because of the grain feature and although I am aware that one need not alternate end grain direction, that's just how the best fit turned out and furthermore to aid in aligning the boards before glue-up I inserted small dominoes along the edges.

The timber is Melville Island Bloodwood (Corymbia nesophila) and weighs 10.2 Kg. I think bloodwoods in general are prone to cracking/splitting and perhaps I erred in selecting a board of that width in the first place. The grain pattern is shown here:

Feature Grain.jpgNote the width of the central board.

However, to add to the agony, not only was there warping but this too:

Crack.jpgThis crack is in the end of the central board.

And this:

Split Joint.jpgSplit Joints.jpgSplit joints. The timber is quite waxy so its possible these were dry joints but that'll be a first for me. And if they were not, why did the glue (Titebond lll) not break the timber instead as it's supposed to do! Perhaps just bad jointing in the first place. It was all done by hand and I certainly didn't notice a poor fit before glue-up.

The top was fixed to the cabinet thus:

Buttons.jpgSlot Front Rail.jpgButtons on the sides and screws front and back - but how does one know whether or not the screw in the slot is in the correct location or not? Could be right at the front or right at the back and not allowing any movement at all!

So, what to do about this calamity? I've taken the top off and turned it upside down considering the warping was because of differential moisture absorption and have left it in the same location as before. I did 'seal' the bottom with tung oil but whereas I applied several coats to the top, I think I was mean on the bottom and applied one coat only - two at the most. The top was then finished with several coats of uBeaut wax.

Other points of note are that the rough sawn timber was stored in the open for many months in Muchilba (west of Mareeba) and then lived in my shed for quite a while before dressing and, if anything, the shed is on the moist side. The completed cabinet is, of course, inside the house which is certainly less humid and several degrees warmer than my lovely cool shed.

My plan is to leave it be for a while to see if the warp will correct as is with the bottom now exposed to ambient air/temperature. If it doesn't, I think I will try the 'Kitty Litter Method' outlined here:
https://www.woodworkforums.com/f9/rep...32/index2.html : Post #22

At the end of the day though, this tragedy is more than just warping. Even if the warp resolves, I could fill the blemishes with wax, I suppose, after applying 1000 coats of oil to the bottom, but would I be happy? I don't think so. Ultimately, I think I'm going to have to do something drastic particularly since I do not have enough timber to make a completely new top (470mm x 940mm) but probably have enough to conjure a couple of narrower boards of the full length. I'm thinking along these lines:

The Fix.jpgThree saw cuts along the length; remove the crack and saw along the failed joints. Then insert a new narrower board or two to make up the width. Or, alternatively, try to equalise the board width right across the top.

Does anybody have any words of wisdom? Who would do things differently?

Cheers
Brian

[AlexS]

I feel your pain!
How has the humidity been there? We're presently at 17% RH after a long stretch of hot, dry weather, and it's playing merry hell. Even some specially bought quarter sawn jarrah has cupped. I suspect that a big difference between the RH in Mareeba and Toowoomba is the cause of the problem. Do you have the aircon on - that dries the air out too.
Filling the gap could lead to problems, and you know you won't be happy. I think I'd go for resawing along the split and re-jointing. When you glue up, wipe the faces with acetone immediately before gluing. Use plenty of acetone, it dries almost instantaneously.
Some people put a slight concave curve on the faces before gluing up to allow for the ends drying out. I've never done that, and I wouldn't do it now as any drying has probably all been done.
I'd use buttons all round to attach it.
Hope this helps.

[Dovetail]

Thanks Alex,

Unfortunately, I have no means of measuring the RH but while I was away for the winter reports have indicated that it was a rather 'wet' winter in this neck of the woods. I'm surprised the changes noted in the top occured so quickly since the piece was inside the house and the house was shut up for that period of time so there wouldn't have been much moisture ingress from outside. Perhaps I'm wrong. Do you agree that this whole sorry affair is due to differential moisture absorption by the underside of the top? I presume this is the case although I was aware at the time that the central board was much wider than the others and the cupping that has taken place is in the direction that one would have expected from the appearance of the end grain which is definitely not quarter sawn. Perhaps using this broad board was too adventurous? And I kick myself now having not used acetone to prepare the glued surfaces. Will next time though.

At what stage would you advise I start the process of re-sawing? Should I wait until the top has flattened first because if I don't, all my cuts will be at something of an angle and I'll be making more work for myself regarding the jointing. Once I've remade the top, do you think multiple coats of tung oil will be sufficient to seal the bottom, or should I use some other form of sealing? The top is currently lying face down in the warmest part of the house. Actually it's so warm (and dry) up here and will be for the next few days, I think I'll move it to a shady spot outdoors.

Using buttons all round was going to be tricky because the front and back rails are so wide and I was concerned the buttons would be located too far back from the front and too far forward from the back of the top. I thought fixed screws at the back (mine had small slots) and slotted fixing at the front (or vice versa) was a legitimate method of fixing tops?

Damn nuisance this, I've not had any other tops misbehave in this way. And the tops of the matching bedside tables are stable after a couple of years now in the same location. The mysteries of timber!

Wish I had done a woodworking course!

[artme]

Bugger!! Not the best thing to happen.

I made a small table from Bloodwood many years ago, and after two vastly different
seasons humidity wise it parted company on joins and in other places.

Lesson learned. Use Bloodwood to build bridges, not furniture.

[rustynail]

When it comes to furniture and furniture timbers we are often guilty of pushing our luck. Therefore, when things go wrong, it is best to learn from the mistake rather than feeling violated. It is acceptable to do both.
Thank you for the detailed description, it makes it much easier to draw conclusions.
IMHO there are several factors at play here:
1) Not a furniture specie.
2) Initial moisture content unknown.
3) Back sawn timber.
4) Wide boards.
5) Uneven sealer.
6) Possible over tightening of table clips.

Just because a particular timber doesn't rate as a furniture specie, isn't to say that it can't be used. It just means all care must be taken to address its inherent short comings.
What to do now. I would resaw as you suggested. Once the first cut has been done, the point you raise about
the cuts being out of square would no longer be relevant as the timber would loose the cup as soon as that first cut is completed. Particularly if that cut is down the centre. Return the piece to the saw for square up. The use of a fine saw here would be an advantage as the joints could be reglued straight off the saw which often gives a superior joint in difficult timber.
Run a moisture meter over it so you have some idea where it's at. I have never made a piece of furniture without knowing its emc.
Oils are a penetrating sealer and in so doing, when applied sparsely, leave a dry surface behind. An invitation for moisture absorption down the track. A wax sealer helps to prevent this, but I bet you only waxed the top surface
didn't you? Dont worry, we've all fallen for that one.
An old tradesman I had as an apprentice, used to call it Bloodywood. I think we both know why.
Hope all goes well.

[Dovetail]

IMHO there are several factors at play here:
1) Not a furniture specie.
2) Initial moisture content unknown.
3) Back sawn timber.
4) Wide boards.
5) Uneven sealer.
6) Possible over tightening of table clips.

Thanks rustynail,

Excellent info. Very happy to learn from my mistakes but, in this case, I'm really restricted by the amount of timber I have left - just checked and I may be able to squeeze out another board of a similar length. And you're right, I knew the BW wasn't an everyday furniture timber but I had 'experimented' with it on the bedside cabinets and it looked great. Had to match those, didn't I? Working with BW is not without it's challenges. Talk about cranky grain!

I don't have a moisture meter but I'll get one soonish. Any recommendations? And dammit, I hadn't waxed the underside but will when I've got the problem sorted. Tops are coming off the bedside tables tomorrow to get the treatment!

Incidentally, I left the top face down on the deck outside today to air dry the underside and the previous measurement from the straightedge to the top has reduced bt 2mm already!

[AlexS]

All the factors mentioned by Rusty are likely to have contributed. Does the top get direct sunlight?
You can never completely seal timber to prevent moisture change, but it's always worth using the same finishes on both sides.
It may be worth trying to remove the cupping - some people lay the wood concave side down on the grass, I've had some success putting a mist of water on the concave side and clamping it down flat for a couple of days, damp side down. No guarantees though, especially with a timber I haven't used.

If you can, it's best to wait until the wood's flat until you resaw it. Also, perhaps resawing several times to make the boards narrower. Is it possible to alternate them by flipping end for end so they tend to form a wavy surface rather than cup in one direction.

[Dovetail]

I might not have made myself clear on a previous post. Yesterday I had the cabinet top (panel) outside on the shady deck with the top surface (concave side) down and the bottom (convex side) facing up so that it could air dry given the very warm weather here yesterday. By the end of the day there was a reduction in the warp by 3mm which is most heartening. Out it goes again today...

I have two lengths of undressed timber that are only just long enough to be used to make up the deficit in the panel resulting from the resaw procedure. I plan to use a board of reasonable width rather than just wide enough to fill the gaps left by the saw kerf. Unfortunately I dont have a thin blade saw blade. I am concerned that I may find defects, especially checks, at the ends of the undressed boards that will preclude my using them because if I do, I'll simply be replacing one check(s) with another.

Alternative fixes include:
1. Join two pieces end to end using scarf or half lap joints to make up the length. Not really an option IMO.
2. Use shorter, 'clean' lengths of timber to create breadboard ends. This could solve several problems like preventing the top from misbehaving again; hides any potential smaller checks in the ends; resolves the issue of using a length of timber that wouldn't normally be my 'first choice'. Is this a legitimate fix for a cabinet top? Would it look very strange. Who cares if it resolves the problem long term?

[rustynail]

I would suggest that if the top has responded so quickly to a sun kick, it would appear your problem is uneven sealer application. Repeat the process until flat, resaw as required, refinish BOTH SIDES and wax BOTH SIDES.

[rustynail]

I have a Tramex non invasive moisture meter. It has given many years of good service. Advantage here is there are no spikes to drive in, so the meter can be run over the surface of a finished piece. Does require a depth of material of at least 25mm for an accurate reading. This can be overcome on thinner material by placing a second piece to make up the thickness.

[jimbur]

I don't know if it will make you feel better but I've seen men with decades of experience driven to distraction by supposedly stable pieces of timber. Doubtless Chippendale had his moments too.

[Dovetail]

Thanks everybody for your help (and support).

Because I'd rather not go through this again and even if I go down the re-sawing path and board replacement and follow the finishing advice as given, there is no guarantee that this timber species may not play up again,
I've decided to fix the job by applying breadboard ends. I've not done these before so I anticipate a nice challenge. This will resolve the blemishes evident at this stage because I'll cut the ends off thus removing the offending bits altogether. Thorough finishing of both sides will follow!

Still waiting for the top to flatten. Much cooler day here today. Early days, I know.

[rustynail]

Breadboard ends come with their own set of problems. It is important that an unstable specie has some ability to flex its muscles, otherwise more serious cracking is going to be likely.
The fact that you used this timber to match previously made items is going to be negated when breadboard ends are used on only one of the pieces. It will look like you ran out of wood.
What about a couple of cleates screwed to the underside of the top with slotted holes for expansion and contraction? If they are fitted across the grain and kept within the carcase frame, nobody will be any the wiser.

[Dovetail]

Thanks rustynail,
I did consider this option and would have gone that way except that by using cleats, I would definitely have to resaw and replace boards. The point is that the lengths of spare timber I have are only just long enough or marred by the appearance of sapwood and knowing now just how prone this species is to checking, I would fear ending up exactly where I am now in the not too distant future. So, the supposition is correct. I don't have enough timber! Busy dressing what I do have, so the answer will be a clearer soon.

I don't think the difference between this and the other pieces of furniture will be an issue as they are really 'the same, but different', if you get my drift?

If I do use breadboards, I may use cleats as well! Belts and braces!

[rustynail]

I have returned to your pictures and am of the opinion that the need to resaw may not be necessary. As the joints have already cracked, they should now come apart quite easily. Should this be the case, only a shot with a plane would be required to clean up the edges, a minimal decrease in width. The centre board appears to be the main offender. If a saw cut would be too much loss of width, then a wait with a weight until the piece returns to flat may be the go. Once flat, reglue the three pieces, wack the cleats on underneath, repolish and wax both sides and bung it back on the carcass, before something else goes wrong.
I have trouble with the idea of breadboard ends as I have always found the joints to be prone to cracking. This cracking can be quite inconspicuous on a larger piece, like a refractory table, but on a small piece, become very obvious. Add to this an active timber and the whole thing becomes rather risky.