Seasonal Wood Movement Question

[ian maybury]

Here's one that may be hard to bottom out. Read another piece recently which set out the usual method for calculating likely wood movement when designing a piece (e.g. for top and bottom clearance with an inset drawer) - figure out your species and grain direction, sort out the dimensions that matter, look up the relevant movement number per inch/ % change in relative humidity, look up the numbers for max and min seasonal relative humidity in your area, and calculate the expected movement for your situation.

Which is straightforward enough. The complicating factor that the written piece quietly avoided mentioning (most don't) is the rate of take up of moisture by the wood. The number delivered by the calculation is in effect a likely worst case - it presumes that the wood actually takes up and loses moisture fast enough to reach equilibrium moisture content at the wettest and dryest times of the year.

The bit I've never seen hard data on is just how likely this is to happen in practice, especially when the moderating effect of finishes, well sealed end grain, shape (surface area/volume) and the like are taken into effect.

Perhaps it's a non issue - the wood takes up moisture fast enough that the basic calaculation is a decent predictor. My suspicion though is that the moisture content and hence movement may lag the theoritical maximums by quite a bit, and that the predicted required clearances may be overkill and unsightly.

Can anybody point to a source of real world data of this sort, and/or report experience of using it? Or alternatively a practical method they have evolved of dealing with the issue? The obvious method of sidestepping the much of the uncertainty/reliance on numbers is to keep test pieces typical of what you work with, and measure the movement over the year. The trouble with this is that it must take quite a while to generate a result, and the predicted movement might not be enough if you get a very wet year. Plus it may behave differently to a piece of a different shape that's finished all over..

[Kees Heiden]

I have an example. The panel of door from my fridge is about 1.20 meter wide. It's made of maple. Because of its width it's a good suspect to study this phenomen. When building the kithcen I watched this panel, because I needed to know how much it was shrinking. It easilly overshot the numbers from the "shrinkulator" by 50% or so, but it took several months. Even now, each winter the panel shrinks so much that it peeps out of its groove, but only reaches that level only when the spring is almost there. The panel is now part of a panelled door, it is finished with oil and wax. I don't have hard data, but I can say that the values from the shrinkulator are conservative and that it takes a lot of time to reach those values, but fast enough to reach them within one winter.

[Jim Matthews]

This is a very good question, particularly in coastal areas where heating
will further affect the relative humidity. My guess is that if you have
binding in moving parts, such as drawers - the construction
isn't allowing all parts to reach equilibrium.

The external parts expand, and the internal parts 'lag' as you intimate.

Local conventions in building may be the best indicator of how much movement
is to be expected. In the part of the US where I spent my salad days,
we would swing to near 90% relative humidity for much of July and August.

Without central air conditioning, this lead to many moving parts binding -
particularly large doors and flush drawers.

Overlays drawers and doors continued to function.

http://www.woodweb.com/knowledge_base_images/zp/equilibrium_moisture_content.pdf

[ian maybury]

It's a dead interesting and also very practical issue in many situations.

The numbers can get pretty scary when wide pieces are involved Kees - I unthinkingly and fortunately incorrectly estimated the potential movement across the full width of a large garage door a few years ago using the stock calculation and was shocked. There's a number quoted for the movement seen in a table top versus a breadboard end too in the thread linked below, and pic here: http://www.sawmillcreek.org/showthread.php?212304-Breadboard-end-extreme-movement&highlight= - it's far from insignificant. Luckily my garage door is a long grain frame and V boarded construction (the boards are free to move inside the frame), but it still moves enough to tighten in the Winter - by maybe 3/16in /5mm. (the vertical members at each side of the frame which are the source of the movement are each about 4in wide) This close to the max that the lock can accommodate.

A quick look at the Shrinkulator suggests that it too predicts the movement presuming the wood reaches the equilibrium moisture content for the seasonal extremes of average relative humidity - it again doesn't necessarily take account of how quickly/slowly the moisture is absorbed by the wood. Thats a great set of tables Jim, thank you. We don't get so large a swing in relative humidity here in Ireland as in parts of the US - it's just wet all of the time here.

A search since the initial post brought up this previous SC thread on the shrinkulator: http://www.sawmillcreek.org/showthread.php?213839-Has-anyone-checked-the-shrinkulator It adds a bit, in that there's a definite view (by Bob Lang no less) that the movement may be less than it predicts for the sort of reasons discussed (shape, screening, blocking effects of finishes etc) - but nothing definitive.

My (engineering oriented) tendency is to look for the means to objectively calculate the movement to be expected, but in practice it may not be a runner. If nothing else there can even be quite a lot of variation in different pieces of the same wood it seems. Yet skilled makers are clearly handling the issue of e.g. inset drawer fits fairly precisely. Perhaps its a matter applying rules of thumb. There's of course lots of situations where it has no visual impact - I guess in those cases it's easy to just provide for the worst case. There's others though where it's perhaps less straightforward because wide gaps etc will mess up appearances.

It's not in the latter case an insignificant matter, and the choice and condition of the stock also matters a lot. e.g. Some quick calcs suggest that a 6in high drawer front in quarter sawn red oak (which is relatively stable) might move by a total of 0.066in (1.6mm) (inc 1/32 in safety margin) in say NE USA conditions where the EMC (equilibrium moisture content) might vary by 4% between Winter and Summer.

The total movement becomes 0.126in (3.2mm - again with 1/32in safety margin) if the wood instead is flatsawn white oak (the white variety moves by approaching half as much more for a given change in moisture content), and the movement could double again to more like 0.2in (6mm - again with the 1/32in safety margin) if you were unlucky enough to live somewhere on the east coast that got cold/dry winters and humid Summers - where the toal change in EMC might push up towards 8%.

The other factor is how dry your wood is when you build the piece. If it's so dry that it's close to the seasonal dryest (lowest EMC) in use, then the drawer has to be fitted with the largest/worst case clearance. If towards the wetter end of the range it could be fitted much more tightly - in the expectation that it's not going to tighten much more in the wet time, but may shrink quite a lot in the dry. (maybe an argument for using wood that's not too dry - less risk of scaring customers???)

The question remains as above - can we safely reduce from the maximum movement predicted by these calculators on the assumption that the finish etc will prevent the EMC of the wood changing through the full range predicted by the tables/seasonal range of relative humidities?????

[Kees Heiden]

In a modern house the heating is on and drying out the air all winter long. That's a long period and usually plenty for the wood to shrink a bunch.

[Jamie Buxton]

Atmospheric humidity is readily available data -- every weather station in the world records it. However, the humidity inside most houses is different. If you heat your home in winter, the inside air becomes drier than outside. And if you air condition your home in summer, the inside air becomes more humid than outside. And sometimes folks complicate the picture by running a humidifier in winter or a dehumidifier in summer. You might need to measure your home's air humidity over the course of a year. Or measure the EMC and/or width of a test panel over a year. If you did both -- measure the air humidity, and the wood EMC -- you'd be able to answer your original question about whether the wood completely tracks the air humidity changes.

Further thought... Whether the wood perfectly tracks the atmosphere or not depends on how thick your wood is. If it is piece of sliced veneer -- only 1/42" thick -- it will likely track quite sell. If it is a 4"-thick slab, it might not. So if you go the experimentation route, you might want to run one set of experiments on a door-panel thickness, and a different set on a tabletop slab.

[Brian Holcombe]

Outside of veneer, finish does not prevent any movement, it slows the transition but a full winter is still plenty of time for this to occur.

I have had one piece early on where this became an issue and that piece had been shellacked inside and out and glued on the edges. Plan for it, don't attempt to prevent it.

[ian maybury]

For sure it's not about trying to prevent movement Bryan, all this is against the background of correct design for movement. There are though places where from a visual point of view small differences in gaps can count for a lot (e.g. the drawer fronts described), and where getting it too tight (and there's lots of scope for this to happen - especially if a stock clearance is simply applied, and the moisture content of the wood being worked is not fairly consistent) can be disastrous too. So it's more about figuring out a strategy to predict likely movement - especially to gain some views on whether or not it's reasonable to presume that the wood will reach the moisture content equilibrium/EMC suggested by the relative humidity of the air.

The former is as you say Kees determined not just by the outside air conditions, but also by the further drop in relative humidity from that predicted by weather tables indoors (easily read off from a psychrometric chart) as a result of the air being heated within the house. The movement might actually be increased from that in the examples i listed above when heating is factored in - at least in a climate where Winter is the dry season.

For sure Jamie thickness is a factor, as may be enclosure. Your point brings to mind the old rule of thumb to the effect that most woods can be expected to reach equilibrium moisture content to a depth of about 1in/year, and that moisture take up tends to be much faster at ends where the grain has been cross cut and the pores are open. Which suggests as you say that a much thicker chunk of wood is unlikely to equilibriate over the six months or so from season to season, but that a relatively thin piece like a drawer front might get fairly close to doing so - working its way in mostly from the ends.

Wonder does the latter suggest that it's a good idea to make sure end grain is well sealed? Not to prevent moisture take up - more to avoid the area at the end of the piece of wood (which is also most prone to splitting) from running too far ahead of the body of the piece....

[Brian Holcombe]

My comment was in regard to what you were asking with regard to how much you can reduce the rate of expansion/contraction. Finish will slow the rate, but it will ultimately still find equilibrium. I would plan around the extremes, unless we're talking about huge panels the additional amount needed to accommodate the extreme is very little.