Help / Advice please on edging wood movement

[Phuture]

Hi everyone, I’m hoping someone here can help me to understand exactly where I went wrong with this project so that I don’t make the same mistake again.

I’ve used some offcuts of birch ply to make a herringbone tabletop and then brought some Iroko to edge it. The Iroko is cut so that edge grain is flush to the top and the face grain is facing out from the table (seemed like a good idea at the time). The Iroko had mitred corners, was secured to the top with dominos and glue and then the completed top was finished with two coats of Oli Natura hard wax oil.
Happy with it, I moved it into the house. All was good but after about 4 days, I decided that the surface finish could be better so took it back to the garage to give it a refinish. It took about 30 seconds to get into the garage but as soon as I put it down, I noticed 4 large splits in the ply, two of the mitres had blown out quite badly and the edges were a very long way from being straight. It was definitely not like this just before leaving the house as I’d just given it one last inspection before deciding to refinish it.

My questions are:
Was I just being foolish to think that any wood movement in the Iroko wouldn’t be a problem as it was only constrained along one side?
Was the orientation of the Iroko just an inherently bad idea?
During the glue up, I used a fair bit of clamping force to close the mitres up. When coupled with the wood movement, was this just bad practice and should I have just been more accurate when cutting the mitres?
Any other advice on how I can edge the ply with hardwood and not run into this problem again?

Thanks
Dan

 

[Spectric]

During the glue up, I used a fair bit of clamping force to close the mitres up.

I would say that the mitres should close without any force and the clamp is there to keep it in place whilst the glue dries.

 

[Doug71]

It looks like the ply panel has dried out and shrunk?

 

[Sgian Dubh]

Hi everyone, I’m hoping someone here can help me to understand exactly where I went wrong with this project so that I don’t make the same mistake again. Dan

Your problem isn't the lipping; the problem is the main field of the table top that the edging (suspiciously beech like in appearance) encases. Because you have cut small bits of plywood that you have then joined together plywood top/bottom face to plywood top/bottom face, your main field is essentially all cross grain in its thickness, i.e., left to right and end to end. Since making the table the main field of your table top has shrunk as it's dried out causing both the splits in the main field and the gaps in the edging/lipping, and gaps evident at the opened mitres and between the inside face of the edging and the table top's main field. It's also the cause of the edging not being straight: I strongly the suspect the edging that's still stuck firmly to the main field is concave rather than convex.

There isn't a good solution for applying a fairly thick edging to this sort of structure you've manufactured, except to perhaps to not apply an edging at all, or if you 'must' apply an edging, you might just about get away with a thin edging of perhaps 3 - 6 mm thickness, but I'm not even sure that would be satisfactory and remain gap free. Slainte.

 

[recipio]

It looks like the ply has shrunk and separated from the lipping. ? How long did the ply have to dry out after glueing ? I always find this technique very challenging - the mitres have to be perfect and if out by a fraction of a mm they will refuse to stay together. And yes, I think you have beech instead of Iroko !

 

[Phuture]

Thanks everyone for the feedback. They say every day is a school day and I'm definitely learning today. I hadn’t even given any thought to the possibility of the ply shrinking as I’ve always been told it's supposed to be dimensionally stable.

The ply pattern was glued up for around a month before applying the lipping but it had only been kept in the garage so probably hadn’t dried out at all. And the lipping has indeed got a concave profile to it where its still firmly stuck so I think you’ve definitely nailed the problem there.

As for it being Beech rather than Iroko, again, you're probably both completely right. It was sold to me as Iroko and being fairly new to hardwoods, I just blindly believed what I was told.

Thanks again for the help. I think I'll have another crack at this when I've built up enough scrap ply but next time I'll get it into the house for drying before lipping to see if I can attack the issue that way instead. It may well end up the same but it will give me chance to practice my mitred corners again if nothing else (only this time with less clamping force).

Dan

 

[profchris]

Your glue probably dried, but your garage is almost certainly more humid than indoors. Plywood is only (fairly) dimensionally stable along and across its face - it gets thicker or thinner with humidity. See this thread Humidity and wood movement - the basics

If you look at your top and work out which direction it moves most, you can see that it will want to tear itself apart anyway! You might get away with it if you formed the edges of the top into a tongue, and set them into a groove in the lipping with space for expansion.

Sgian Dubh would know better than most if that might work.

 

[Jones]

Who sold you beech and called it Iroko? They're completely different .

 

[Phuture]

Thanks for the link to that thread, very useful information. I suspect I'll probably be getting myself a copy of Cut & Dried soon as well.
The wood was from the off cuts pile at my local timber place but in the defence of the young lad that served me, he did say he 'thinks' it was Iroko. I'll know better to check with some of the other staff there next time before blindly believing him.

 

[murphy]

Yes, I say it's Beech too, nothing like Iroko
https://www.slhardwoods.co.uk/products/solid-timber/par/iroko-planed-all-round-board

 

[Sgian Dubh]

If you look at your top and work out which direction it moves most, you can see that it will want to tear itself apart anyway! You might get away with it if you formed the edges of the top into a tongue, and set them into a groove in the lipping with space for expansion.

Looking at the table's main field, Chris, the shrinkage is actually rather interesting. The reason for that is the herringbone pattern as viewed from above. I can't see from the images available if it's a multiple herringbone pattern, or if it's just one with a single centre line joint. In any case the expansion and contraction direction of one set of laminated plywood blocks on one side of the join line is going to be at 90º to the expansion and contraction of the set on the other side of the join line, so there's built-in stress there, and the result of that stress can be seen as splits between veneer layers in the main field of the panel. That pattern of movement and failure would almost certainly be repeated for each one of the herringbone arrangement if there are repeats: as I said, I can't tell from the available images if that's the case, but I suspect not.

I'm not sure that your idea of tongued edging glued into a groove on the edge of the main field would work any better than the original which, as we know, didn't work. It might but I suspect, as I said before, that any edging would need to be thin (3 - 6 mm), with or without a tongue. Thin stuff is relatively flexible compared to thicker stuff and would more likely to bend with movement of the edges of the large panel. Even so, I'd half expect failure of some sort, even it was only for mitres to show alignment problems at the corners, e.g., open up, or the outer points of the mitres to become misaligned as the main panel (field) expands or contracts in its different directions (90º to each other) in response to MC changes caused by changes in RH.

I'm not sure I've provided a solution to Phuture with the above. Personally, if I had to make a table like this for a client I think I'd avoid altogether adding a lipping or edging, and I'd warn them that there's a strong possibility of splitting as already seen in the example seen in the original post. In other words, I wouldn't be willing to provide the customer any sort of guarantee, except to say it's highly likely to fail, and don't come back to me and complain about that, ha, ha. Slainte.

 

[Phuture]

Sgian, thanks for the detailed explanation, it’s definitely given me food for thought and a much better understanding of what’s happening here. As for providing a solution, I’ll try it with a thin 5mm lipping but if that fails as badly as this effort has, I’ll drop the lipping altogether as you suggest and start experimenting with a round over or chamfer at the edges.

One idea that I was originally going to try with this table was bonding it down to a 9mm MDF before applying the lipping. From what you’ve said, it sounds like that wouldn’t have made any major difference anyway.

I’ve attached a better photo of the whole table top and labelled where the ply has split. Interestingly, of the 4 sections of ply, only three have split. Would I be correct in assuming that if I was to make the herringbone section from solid strips of hardwood instead of ply, I would run into a similar issue with the shrinkage pulling in different directions and thus distorting the lipping and opening the mitres?

 

[Sgian Dubh]

One idea that I was originally going to try with this table was bonding it down to a 9mm MDF before applying the lipping. From what you’ve said, it sounds like that wouldn’t have made any major difference anyway.

Doing that probably wouldn't make much difference to the splitting problem, but it would most likely create an additional one, i.e., it would induce cupping of the MDF with the concavity on the upper face.

Would I be correct in assuming that if I was to make the herringbone section from solid strips of hardwood instead of ply, I would run into a similar issue with the shrinkage pulling in different directions and thus distorting the lipping and opening the mitres?

Solid wood parquet flooring frequently includes that herringbone pattern, and to cope with the inevitable cross-grain movement the blocks aren't packed together tightly - without going into too much technical detail they are essentially laid on to an adhesive base with all the blocks simply butted up together and with the whole assemblage stopping short of each wall by 12 - 15 mm to allow for expansion and contraction. In old parquet floors it's common to see that small gaps have developed between each block.

Herringbone type pattern panels (a table top for example) with real wood grain as the display are most commonly approached as a veneering job over a base of some sort, e.g., plywood, MDF, etc and, historically, veneer applied to a solid wood panel best made of something relatively stable, although I've seen quite a number of antiques in which the last strategy hasn't worked particularly well over the centuries, although other examples have survived very well. Slainte.

 

[Kayen]

Nothing more to add, but just wanted to say what a beautiful job you had made of it and with a great-looking finish. I'm glad you're ready to take on the challenge again and I hope you re-post it

 

[Iestynd]

Nothing more to add, but just wanted to say what a beautiful job you had made of it and with a great-looking finish. I'm glad you're ready to take on the challenge again and I hope you re-post it

What he said - nice job

 

[profchris]

If you try again, my best suggestion is to glue up in low humidity. Then the top will only expand, not shrink, unless you take it to Arizona. Expansion is less likely to open up the joins or crack the wood.

In theory each panel should move at much the same rate, though there are no guarantees that all the pieces will behave the same.

Anything fixed to it (including a sub frame) will resist movement unless fixed in a way which allows movement (screws in slots, etc). And the movement will probably cause distortion if it's being resisted.

It's a great look, so I hope you find a way to succeed.

I'd be working on the basis of 1% to 2% maximum expansion, so if this is 1m square that's 10-20 mm. That's a lot for any lipping to cope with!

 

[Hornbeam]

The problem is down to the the differential expansion/contraction along and across the grain. Plywood is very stable in 2 dimensions because of the strong adhesives, which mean the long grain laminations, restrain the movement of the cross grain laminations. When you look at the pattern you have cut all of the grain in all of the laminations is cross grain and so has the highest level of movement compared with the beech edging strips.
The only way around this would be to lay the plywood herringbone pattern as very thin strips no more than 3mm thick onto a solid plywood base, doing the same thing on the other side to prevent warping. This is the same sort of technique to laying oyster veneers.