Thread: Eliminating movement in 2mx1m tuart dining table top

In short, it won't work. Even 2-pack polyurethane won't stop movement.

The way we do it at work is 6~8mm timber laid on 18mm MDF with epoxy with the mitred frame around it

Originally Posted by elanjacobs

In short, it won't work. Even 2-pack polyurethane won't stop movement.

The way we do it at work is 6~8mm timber laid on 18mm MDF with epoxy with the mitred frame around it

Harry, look up JohnG's comments in a post in the timber forum about wood movement from a saw millers perspective, how much shrinkage (swelling) / dimensional change occurs in a board depending upon the cut of the board.

We have discussed wood movement at length in segmented turnings with one theory being that if a piece is thin enough then all the pores in the wood can be filled with a sealer making it effectively "impossible" for the wood to take up or lose moisture. It works for small stabilized pen blanks & Cactus Juice type products but is not practical with larger work.

When you put a sealer on, it only fills the outer pores and does nothing to probably 98% of the mass of the wood. The finish coats form a film barrier, that is only as good as the integrity of the surface. The finish is not totally impervious to water vapours, so it will not stop changes in moisture content, only slow the rate of change in moisture content. Any environment that the table is placed in will over time experience changes in MC, perhaps a daily cycle (airconditioning cycles) or at lesst seasonally.

Wood workers have learned ways to accomodate wood movement. Tuart has 6% tangential and 3% radial shrinkage after reconditioning. Depending upon the cut of the boards, storage, drying & the original MC, current MC you could be facing 15 mm or more wood movement across the 1m width of the table.

Hmmm, thanks everyone. Looks like I'll have to settle for breadboard ends.

You can have a mitered solid frame around solid panels , as long as the wood is dry and as long as you do some good joinery at the miter. I do them all the time in kiln dried US Oak mainly. I went through a period when I first started doing them where I used biscuit joints in the miter . They failed , a slight increase in panel width seperated the joint. I then started biscuiting on the mitered frame sometimes using four biscuits in each miter . I would then flip the top and make a big butterfly / dovetail patch that was sunk in two thirds of the top thickness , the routing for this would take out two of the closest biscuits. This would also be skew pegged from the top side right through. Talk about going from one extreme to the other . I've never seen one move , after doing many tables like that . I don't know what would happen if one was left in the rain through a winter ?? I do wonder what would be the weak part and give . These tops had 30 mm thick timber for the edge frame , 19 to 22 thick for the panel and were roughly 2.4 to 3 meters long X 1 meter wide . The edge was usually narrower than the ends . If I was doing 100mm wide down the length then I'd do 150 at the ends, the join was not 45 degree. Talking with a few other makers who were building the same types of tables we would compare methods. All had discovered the hard way that biscuits alone don't work . One guy would create a through mortise in both sides then insert a cross grain tenon and peg from the top. Another would do the same but not all the way through at the visible edge. I know of someone who does more a more modern look with the butterfly / dovetail patch from the under side. I don't know if he pegs it as well , you wouldn't need to if its fitted the right way .
Rob

Originally Posted by harry wall

Instead of using breadboard ends covering the ends of edge joined boards (which permits movement across the grain) I am considering using a 100mm mitred border around the 4 sides of the top - just because I like the look of it. This, however, would mean a difference in grain direction at the ends - the mitred end piece covering cross grained ends of boards. Any change in humidity would then mean the mitred joins could split open. Of course that is only assuming the timber actually gets to absorb or lose the moisture.
So I'm wondering if sealing the entire table top with 2 coats of Wattyl Sanding Sealer and then applying two coats of the Wattyl Water Based Xtra Clear Estapol would prevent any moisture getting into or out of the tabletop. The timber would remain stable with humidity changes and so a mitred joined border should remain closed.
Can anyone confirm by direct experience whether this system would work?

Originally Posted by elanjacobs

In short, it won't work. Even 2-pack polyurethane won't stop movement.

The way we do it at work is 6~8mm timber laid on 18mm MDF with epoxy with the mitred frame around it

Harry

If you have access to a bandsaw can I encourage you to learn to use shop-sawn veneers.
At 6mm thick a veneer can be handled and edge jointed like a thicker board, but once glued to a MDF or particle board substrate you won't have the movement issues you would have with solid timber.
Not only will you use less of the expensive timber for the top, you can book match the top and you will develop new skills and open up further design possibilities.
The only "trick" with veneering is remembering to veneer the underside of the substrate as well as the top.

Harry I recently completed a second bench cupboard the top of which is edge jointed boards.

The first cupboard top was also done with edge jointed Boards. The big difference between the two is that
the first top is made entirely from recycled floorboards and I left the tongues and grooves in situ. I did my
version of breadboard ends whereby I routed across the ends to half their depth opver a width of65mm. my
end piece was routed in a similar manner along its length. The two outside joins were the cut up to the end
of the rabbet with a hand saw the next join was skipped but all the others were similarly cut with a handsaw.
I figured this left plenty of room for movement, The end piece was glued in place ( Sellys PVA ) and two holes
of 6mm were drilled in from the top and straight through the end piece and the long boards in each of the
second boards ( from each side). 6mm dowels were then glued and driven int the holes and flush cut when dry.
After 12 months no problems.

Now you may not like the idea of seeing dowel ends but if a bit of thought is given they can make an interesting feature.

The second top was made from a couple of shorter lengths of the same flooring plus some hardwood beam material that
I, with great difficulty I might say, were milled to size by myself. Unfortunately these moved a little and the glue up was a
right royal PITA!! Now the two floor boards were relieved of the tongues and grooves so as to get a good surface to match
the board edges. I have done the ends in the same manner as the first.

I expect this top to suffer some movement because the original hardwood beam was from backsawn timber. Time will tell!!

Originally Posted by auscab

You can have a mitered solid frame around solid panels , as long as the wood is dry and as long as you do some good joinery at the miter.

Thanks for the different examples of joins Auscab. A couple of questions:
(1) how has the timber been finished / sealed /coated ?
(2) to what extent do you think the coating has eliminated moisture take up in the panel ?
(3) do you think it is the strength of the mitre join that held the panel to its original size or did the end pieces just happen to expand at the same rate as the panel?

Originally Posted by harry wall

Thanks for the different examples of joins Auscab. A couple of questions:
(1) how has the timber been finished / sealed /coated ?

Polished on top with a seal coat of estapol sanding sealer , thinned. Then shellac.
sealed underneath with a water colour with some pva mixed in to make it a water based sealer.

Originally Posted by harry wall

(2) to what extent do you think the coating has eliminated moisture take up in the panel ?

It helps , specially if you do the same on the underside as what is put on the top. We seal the ends of timber when drying it to prevent rapid moisture loss, it works for that , so it does have an effect. I don't think anyone would rely on it 100% for eliminating the risk though.

Originally Posted by harry wall

(3) do you think it is the strength of the mitre join that held the panel to its original size or did the end pieces just happen to expand at the same rate as the panel?

I think if the wood is dry and the piece is kept indoors , its not going to have to deal with massive moisture gain, and the joins in the mitre can handle the pressure of that . Just a glue join with dowels or biscuits is not good enough. Given time and periods of neglect , the antique tops Ive seen have moved and cracked .

Rob

Also take in to account where in the room the table will be. If one end is going to get direct sunlight for several hours a day or is near a fireplace or heating vent the movement risk goes waaaay up on that end.

We've had normal plank tables (35-45mm kiln dried US oak) crack in those environments before, I don't know how a mitred frame would hold up

Originally Posted by auscab

Originally Posted by harry wall

(3) do you think it is the strength of the mitre join that held the panel to its original size or did the end pieces just happen to expand at the same rate as the panel?

I think if the wood is dry and the piece is kept indoors , its not going to have to deal with massive moisture gain, and the joins in the mitre can handle the pressure of that . Just a glue join with dowels or biscuits is not good enough. Given time and periods of neglect , the antique tops Ive seen have moved and cracked .

Rob

Originally Posted by elanjacobs

Also take in to account where in the room the table will be. If one end is going to get direct sunlight for several hours a day or is near a fireplace or heating vent the movement risk goes waaaay up on that end.

We've had normal plank tables (35-45mm kiln dried US oak) crack in those environments before, I don't know how a mitred frame would hold up

Hi Harry

some good advice there from Elan, however, I believe your choice comes down to
(1) a mitered frame with strong mechanical joints at the corners -- Rob's dovetail keys
(2) inserting a piece of all-thread at each end to constrain any expansion of the flat boards -- tension in the all-thread will stop the boards expanding, and the ends are where the greatest risk of expansion is
(3) combining (1) or (2) with quarter sawn boards from a species with low radial shrinkage and expansion -- but this might be very bland grain wise
(4) using shop sawn veneers glued to an MDF or particle board substrate -- which is then has the miter frame attached.

my personal preference would be (4) as it gives you the widest choice in species and grain display.

I do the sawn veneer type as well , it's how I like to do parquetry top tables . I saw the solid down to around 4mm thick and lay / press it onto marine ply , with a two pack glue , techniglue . Ply , is the quality / more expensive way to go if your laying onto manufactured board . And it will still be around as long as the solid example but with no shifting .
Rob

Yes, some great information there so thanks to all who contributed their experience.
Since I have already built the panel, but not yet the mitred border, my choices are a bit more limited. The tuart panel is 20mm thick and not yet sanded flat.
Yesterday I was thinking I could keep a mitred join but strengthen it by half lapping it (dunno if that is a correct term but the idea is to get much more glue area by lapping the joint rather than just use the end grain to glue). That means that one end of the border will have exposed end grain visible. Actually I suppose the mitred join could also be half lapped and also hidden but I haven't thought that out yet. I suppose there is a specific name for that as well.
You can see I'm not a furniture builder can't you !

Or how about this..... use a standard mitre join (not lapped) but countersink and plug a couple of 75mm screws through the corners of the join to keep the joins together. It would not be as strong as your method 2 threaded rod Ian but it may suffice given it is an indoor table and I'll be trying to seal the timber as much as possible with those Wattyl products I mentioned. That would also be simpler than me trying to cut a perfect butterfly type insert/locking piece. Has anyone tried this method of using screws across the join into the end grain?

Harry
given where you are with the build, and your knowledge of furniture construction, I think your best option may be to attach the long edge pieces of the frame to your existing Tuart panel (these are glued on?, biscuited? or ??), seal the ends of the Tuart boards where they will be covered by the end pieces of the frame, attach the end pieces using a glued half-lap mitre joint -- BTW how were you proposing to align the ends with the Tuart boards -- sand the panel flat and apply your preferred finish.

You may get lucky and find that any movement is not enough to destroy your table.


If in a couple of years the table does blow apart, come back and we'll help you fix it.