Finishing and Wood Movement

[sawduster]

Are you ready for this? Conventional wisdom has it that one should apply equal amounts of finish to all sides of the wood to help minimize wood movement and warping. Makes sense, right?

Well, over the last few years I have read several articles in magazines by respected wood workers indicating that sealing the wood up tight has very little to do with warping and damage causing movement. They've looked at antique pieces of similar origin and geographics, some of which were fairly pristine while others had suffered warped tops and joint damage.

As was generally the case in the olden, olden days, finish had been applied in all of the cases studied only to the visible areas of the wood. The main difference they found with the pieces was that the finish on those which had suffered the damage had been allowed to degrade and that the damage was caused by direct contact with water, not water vapor migration. The direction of the warpage is indicative of water having entered the wood via the top of the visible finished face of the board rather than via the unfinished bottom of the board. The pieces on which the finish was maintained had remained in very good condition.

It has long been thought that warpage was caused by uneven migration of moisture from one surface to another, and that is likely the case for bare wood set down onto the cold concrete shop floor or laid out in the wet grass, but in both those cases we are talking about water, not water vapor in the air. In the case of the shop floor, actual condensation onto the wood. Every wood that I can think of is more porous at the end grain than on the face grain, and that is why you get checking and splitting at the ends of pieces you are drying if you don't do something to slow the moisture flow at the end grain, thus allowing that significantly uneven migration to occur.

I don't know that I have ever looked at the inside of even high quality case work that showed evidence of having been finished, and the most I have generally done is apply a spit coat to the non readily visible areas of pieces I have done, and have not really had any issues with warpage or movement damage, not-withstanding those couple of small pieces I sent to my folks in the desert, and I think that had more to do with the types of pieces they were. Keep in mind that the way I have bought lumber normally means it sits in my shop for sometimes months before it gets used, so it has plenty of time to acclimate, so maybe that has a lot to do with it.

I'm not saying you ought to change the way you do stuff, but that the jury still seems to be out on these issues and that conventional wisdom is not always gospel.

[Safetyboy]

I've seen those articles too - I haven't examined any high quality antiques myself, but the arguments have convinced me that finishing the inside of pieces is in most cases not necessary. So, I tend to spend my precious shop time on other areas.

Nothing I've built is older than 5 years though, so I don't really have any long-term evidence for the effectiveness of this approach.

[Stretch]

Just a few things that I think need to be mentioned.

1. When viewing antiques, most of the "sealers" that were used were very poor sealers even when pristine. Some were decent when dealing with water vapor, but none could really handle direct contact with water. So while applying BLO on one side and not the other may not cause warpage. Applying vinyl sealer or polyurethane on one side and not the other will (I know from experience) cause warpage. When dealing with modern sealers it's best to treat both sides although equal numbers of coats aren't necessary. You may put six coats of poly on the top of a table, but you really only need two on the bottom.

2. I think the majority of wood movement we see is caused by lumber that isn't fully dry compared to it's final environment. If your wood is green when you start, it doesn't matter what you do to handle movement, you're going to have some adverse effects if you're dealing with wide planks. If you start with wood that is acceptably dry and seal it with a modern finish, you're not likely to have movement issues.

[rrich]

Jerry,

A couple of interesting observations...

My son has a teak trestle table with long extended feet. In his apartment, there was carpet over a concrete slab floor. A ground floor apartment and 13 years sitting on the carpet over concrete slab. He has since moved to a house with hardwood floors over slab. The table drew moisture from the slab and the legs have curved upward. To fix the situation, he has raised the center of the trestle and placed weights on the legs. It is slow process but the legs are returning to their originally machined shapes. It will be another year or so before things return to normal. Obviously, the bottom of the legs were not finished. Once they flatten out, they will be oiled.

I had made a desk top, 27" wide by 40" long. The desk top was just glued up pieces of white oak. Looking at it now, the grain was sort of alternated but not consistently. I put the desk top in the back of my truck. A shell with a carpet kit. I took it to school to run it through the wide belt sander. It couldn't be done on Tuesday and I left the top in the truck until Thursday. To my shock, the top had warped (bow) about 2". The instructor suggested that I flip it over and leave it in the truck until Tuesday. The top was back to almost flat on Tuesday. After wide belt sanding the top was placed back in the truck but propped up so that both sides were in the air.

The top has been finished the same on both sides and has remained flat.

The point of both of these examples is that wood, IN A NORMAL ENVIRONMENT, needs to be finished on both sides. (The carpet over slab is not what I would call "normal".) I don't think that if the top has ten coats of polyurethane, the bottom must also have ten coats. I think that the bottom needs a few coats just to retard the transfer of moisture. An oil finish may be different.

[Leo Voisine]

I have read those articles - and I have seen furniture that has finish only on the places that show. I have also built pieces with finish only on one side.

I have yet to see any warping from normal use. Sure contact with water will cause havoc - but normal usage, no.

There are lots of people that swear that finishing only one side is problematic - and they say they have seen it - but I don't believe it.

I am of the "show me" group. If finish on only one side is a problem - show me. So far - I have seen nothing - whether in real time or online.

It's an old wives tale.

So it is also with alternating the growth rings on a panel glue up.

[rrich]

Leo,

You're probably correct. However, finishing the inside is good practice to see what needs to be done on the outside.

AND

Finishing the inside only takes a couple of days extra and it is cheap insurance.

[art3427]

About 3 months ago I started a kitchen table built from SYP that were originally rafters in my fathers barn for about 5 years. The top is 1.25" X 35" X 60". The individual boards range from 5 - 5.25 inchews wide. I edged glued the boards with TB3 and biscuits after carefully straightening the edges. That single piece of unfinished SYP has been laying on the top of the assembled legs and stretchers for 3 months. It has also been in an open carport for the entire time.

I took a few minutes to check its' condition last week and it is still perfectly flat and straight. This is after many down pour rains and foggy, humid days with the temperature going from the high 30s to low 90s in a 24 hour period.

art

[Stretch]

May 16, 2010 13:53:05 GMT -6 Leo Voisine said:

There are lots of people that swear that finishing only one side is problematic - and they say they have seen it - but I don't believe it.

As I stated above, I know from experience that sealing one side of a panel with a modern sealer and not sealing the other can cause the panel to warp. It's not something I imagined and quite frankly I don't understand why it's so unbelievable. I agree that it's over stated and that most will be okay most of the time, but I have seen it happen on a piece I built and I have seen pieces that were warped because the finish on top failed when the finish on bottom remained intact.

The main piece in question was a table I built out of plain sawn red oak when I lived in New Mexico. The panels that were in the table top were glue ups that were about 36" wide. I let the wood acclimate in my shop for around three months before I build the project. Our house was not air conditioned so the environmental conditions were the same as the shop during the time from when the table was built to when the warping started. I finished the top with minwax stain and polyurethane but only used the stain on the bottom. Within about two months the panels had cupped due to the expansion of the unfinished bottom. I had alternated the growth rings but that didn't matter, the whole panel warped the same direction. This was in New Mexico where the humidity was very low and I still got movement due to water vapor absorption. It was enough to convince me that at least on wide panels it's important to finish both sides when you're using a finish that is a very good barrier to water vapor.

[Ruffnek]

A table top is much, much different from casework where the sides are probably going to be frame and panel. I think one would be hard pressed to find a commercially made piece of casework that had the inside finished.

On a tabletop, I think it is very good practice to finish both sides. The underside is easily accessed and spraying a couple of coats of your favorite finish under there is not difficult at all.

I have personally not experienced a piece warping from finishing only one side but I did get a surprise when I built our Mesquite dining table.

Mesquite is one of the most stable woods you will ever work with and I know a guy in San Marcus who works exclusively with Mesquite. He will glue a breadboard end solidly onto a table and never worry about the top splitting...it's that stable.

Well, I had glued up the tabletop the day before I had to leave for two weeks offshore. I thought I would be smart and to ensure the tabletop stayed flat, I clamped it down to the melamine assembly table with Quickgrip clamps.

Two weeks later, when I got home, the Mesquite top had curled enough on the ends to pull the clamps loose. With one side open to the air and the other captured against the Melamine, it started taking on the shape of a potato chip.

I ended up ripping the boards apart at the gluelines, re-jointing and re-gluing the top.

I learned two lessons from that: 1. Keep both sides of a wide panel open to the air while the waterbased glue is drying. 2. Let wood acclimate fully to your shop before working it.