Thread: Timber Selection Problem

The shrinkage rate of different timbers is very much a variable. This will also be even further varied by the moisture content of the timber in question. The method of sawing (quarter cut or backsawn) can also play a part in the stability. Cedar has a reputation for being a stable timber while Tas oak can be variable due to the variances of species within the trade name and also the level of reconditioning after kiln drying.

I would like to see a picture of the table you describe, including how the top is attached to the base.
It could be that a small change to one or two construction details will fix the "problem"

A piccie would be good


Pete

Thanks guys,

Two pics attached. The first shows the cracking at the corners of the table. The second shows the centre of the table edge, where there is no cracking. As one moves from the centre of the table edge to the corner, the cracking gets worse. The apparent discolouration on the table edge of each shot is my reflection.

The top is screwed to a ledge around the frame. The holes through the ledge are oversize, so the top can move a little.

My coating supplier reckons part of the problem may be the acid cure lacquer I used. He reckons I should use a new two-pack finish he sells. That might help. However, the root cause looks to me to be timber movement that is different radially and tangentially.

Nice effect you have there John

normally people achieve that effect using veneers -- which can be as thick as 3mm
using veneer would normally eliminate the problem you've experienced allowing you to use what ever species you want for the show face

I an still a bit of a novice, but the more I think about it the more sure I am that the problem is timber movement.

I have never used veneer, but it seems you are right to suggest that using a veneer would have prevented the issue ... but I don't have the skills.

It seems that I must either make a new top or an entirely new table ... bugger!

Hi John,

That's a sad lesson to have learned the hard way as the table is an attractive piece of furniture. The problem isn't an attribute of quarter sawn Vic Ash its a function of the joinery you have applied.

Firstly the tangential and radial unit shrinkage rate rate of Eucalypt from the SE corner of Oz is quite high in itself whilst the longitudinal unit shrinkage rate is quite low. To complicate things even more the unit shrinkage on the face of quarter sawn timber is twice that on the face of of back sawn. These features ( and I'm assuming here that the timber was correctly dried to AS 2796) make it just about a mission impossible to enclose a solids table top with a solids tight joined and mitered frame without timber movement in response to ambient changes in EMC% causing difficulties.

One method of joinery that gives good results if you want that framed appearance is to set the frame about 5 mm off the table top per se on strategically placed brass or stainless dowels of about 12mm dia. with open miters of the same gap spacing. Make sure the gap is sufficient to permit cleaning. You could fix up this design iteration you have now by this method. I'd put a very gentle curve on the outside of the frame when I was about it.

You can look up the unit shrinkage rates of timbers in Bootle ' Timber in Australia' for reference. It's a great insight into why carelessly applied drying causes problems with timber floors.


Good Luck Old Pete

When a mitre shrinks, the shrinkage is always more noticeable as you get closer to the shorter side because the length of the timber is not changing but the angle is, so 45 degrees effectively opens up to >45 which causes the gap that you're seeing there.

In a mitred frame, you don't see the problem so much because the members are usually much narrower.

So in your table top, the mitres are pulling away from the corners. I don't think that timber selection is going to solve the problem, because as others have said, even a small dimensional change compounds over a wider board to be quite noticeable.

This is why certain construction techniques are avoided. Even if they work sometimes, you can't know what will happen in the future and so they are best avoided.

You can look up the unit shrinkage rates of timbers in Bootle ' Timber in Australia' for reference.

The shrinkage rates quoted in Bootle are from green to 12% moisture content. I don't think they really help much in working out how much it will move in service, unless there's a way to infer it, which I gather is what John was trying to do.

Originally Posted by silentC

The shrinkage rates quoted in Bootle are from green to 12% moisture content. I don't think they really help much in working out how much it will move in service, unless there's a way to infer it, which I gather is what John was trying to do.

I had a look in "Understanding Wood" to see if there was anything in there and there was a lot of dancing around the issue but nothing I could clearly identify.

What the OP is after are the moisture/dimension hysteresis characteristics of species. The moisture/dimension hysteresis is the boundary defined by the moisture absorption/desorption and expansion/contraction function. This appears to be not well understood or well known for specific species. It also gets pretty technical - eg http://web.byv.kth.se/bphys/reykjavik/pdf/art_032.pdf

Let's not get too complicated. Any empirical data for any particular species will immediately become irrelevant in the real world because that data is for their piece of wood, not yours. It may have been grown under different conditions, sawn differently back/rift/quarter have different grain pattern and it certainly will be unfinished whereas your piece presumably will be under some kind of finish- and different finishes will produce different transpiration rates - and form part of a construction where end grain - an area of the greatest moisture exchange - may be buried. You cannot hope to understand what might happen down the track by looking at numbers. You can only build by the motto that all wood moves and make allowances. Maybe your wood won't move, but if you build as though that is what you expect, then you've only got a 1 in 3 chance of the piece lasting. If you build to allow for movement then, shrink, expand or stay still, you've got it covered. What you can do is build yourself a wooden moisture meter (think a thin strip of wood cut from the end of a table top) with a pointer attached and over the period of a year you can plot the movement with the seasons. This will give you a guide to know whether, for example, at the time of year when you make some frame and panel doors where the wood is likely to be in the cycle as a guide to sizing the panels. Or if you like piston-fitted drawers, whether you need to leave more or less gap so they don't stick down the track. Something like this is essential in a workshop, because even if you do want to go by the numbers, you absolutely have to know where your piece of wood is on that day.

Originally Posted by John Samuel

I an still a bit of a novice, but the more I think about it the more sure I am that the problem is timber movement.

I have never used veneer, but it seems you are right to suggest that using a veneer would have prevented the issue ... but I don't have the skills.

It seems that I must either make a new top or an entirely new table ... bugger!

let's address your apparent lack of skills first

we're talking band sawn veneers around 3mm thick
handling them is a lot like working with solid timber. having sawn the veneer you can pass them through a jointer or drum sander to get the non show side smooth
you can joint and edge them like you would solid timber,
you can arrange them into the diamond pattern you used for the solid timber and get the mitres to line up using essentially the same processes as for solid timber

The only real difference to building the top with solid wood is the underside of the veneers all need to be in the same plane when you come to glue it to a substrate -- this is easiest if the veneers are all the same thickness -- another reason for passing them through a thickness sander
Your veneered top will need be overside to allow for final trimming but so would a solid wood one

I'd then pull all the joints really tight using clear packaging tape and then glue the veneers onto something like 16mm particle board using cauls to apply pressure away from the edge
The underside of the top will need to be veneered with wood of a similar thickness at the same time

you now have made a flat panel veneered on the top and bottom with timber which after final sanding will be in the order of 2mm or a bit less thick.

cut the top to size and attach the outside frame as you did with the solid wood top.


It might just be me, but I think you've demonstrated all these skills in the table you've already built

Give it a go

To all of you who have posted, thank you very much. I have learned a lot today. Being self-taught can be a lonely business, but now I don't feel so lonely. I should have come to this forum some time ago.

IAN ... I definitely will rebuild the top ... might rebuild the entire table.

A couple of years ago I built an entertainment unit with doors that use identical construction (for my daughter). Of course, the doors are much smaller than the table. Two of the doors are OK. One has warped. I will cut off the borders, veneer the diamond pattern panel onto MDF and refit a new border.

I have been under the impression that one needed to veneer both sides of ply, but that it was necessary to veneer only one side of MDF. Ian's post indicates I may have been labouring under a misunderstanding.

Again, thanks fellas. There is a mountain of good information on this page.