Whenever I build a workbench, I try to work with stock that is as dry as possible. But I’ve never shied away from wood at 15 percent moisture content (MC). Thick timbers can be tough to dry, and all of the benches I’ve built with slightly damp wood have turned out fine.
Turns out, I think I might be a little too conservative on moisture in workbenches.
When we built the French oak workbenches in Barnesville, Ga., this summer, we were shocked at the moisture content of the timbers. Despite the fact that the trees had been felled for more than 13 years, the numbers on the moisture meters were alarming.
My benchtop was 30 percent MC. Other benchtops were 60 percent MC, which was off the charts for our moisture meters.
But we had only one week to build these benches, and we couldn’t wait another 10 years.
My workbench is still wet by furniture standards. Most areas of the top and legs register about 15 percent MC, and my meter reads only 3/4” deep. My suspicion is the center of my 6”-thick top is much wetter. I suspect this because my holdfasts rust immediately when left overnight in a hole in the benchtop.
Despite all this, the top has not moved significantly enough to warrant a reflattening. Today I planed a bunch of 1/2”- and 5/8”-thick panels and it was clear the top was still in spec for this high-tolerance work.
The only evidence of shrinkage or movement in the top is at one of the four joints where the legs pierce the benchtop. While three of these joints are as perfect as the day I finished the bench, the top has shrunk about 1/32” compared to the dovetail and tenon in the front right corner.
Oh, and there hasn’t been any additional checking, and the single existing check in the top hasn’t increased in length or width.
To be honest, I had experienced more wood movement on Douglas fir and cherry workbenches of the same design.
We will see how the bench fares as it makes the transition to equilibrium MC. But my suspicion is that the thick nature of the timbers and the joinery will help to make a bench that settles in gently.
— Christopher Schwarz
Are you planning on or doing anything about the existing checks, or are you going to let them run wild?
I am just keeping an eye on them. If they move I’ll secure them with dogbone fasteners (countertop levelers).
I spent some time with a local mill last week in Central MA and he was pretty clear that the MC of the lumber that he stores will always be = to or > than the ambient humidity level. Made sense to me. I haven’t been to GA in a while but if I recall, the air is wringing wet in the summer.
did you undersize the short stretchers between the workbench legs? otherwise as the top dries out and shrinks in width, the legs will become splayed front-to-back. in other words, they will be forced out of squareness to the top.
such is the current state of my roubo bench. the top was at 20% MC when i built it. i intentionally made the stretchers 1/8″ short at the time, but the top has shrunk more than 1/2″ since then, and the resulting lack of squareness is frequently an annoyance.
If I remember correctly your top was two quarter sawn pieces glued up, this would be a factor in the top staying flatter using high MC timber would it not? A more flat sawn top that was a bit wet would surely tend to move more in the drying process than yours has don’t you think?
Interesting discussion. We used to build post and beam with some very green timber but always used joints that would prevent twisting and allow for shrinkage from bottom up. A three story conventional frame house will drop 3/4 inch. Can you tell me why you use such heavy timber for your bench tops, it seems like a lighter section would do everything required. Is it just the old style?
I built a workbench with Chris at MASW in 2012. The moisture content was about 19% on the Douglas fir we used. I assembled it two weeks after I got it home and then flattened it with jack plane and jointer plane. I have not observed any checking or movement in the bench. It is rock solid. No glue used except when we glued up the top. The stretchers were connected to the legs by drawboring. The top was attached to the legs with wedged mortise and tenons. Heck of a bench and well worth the $$ and effort.
Hey Chris, I notice that your screw on the leg vice is positioned about half way down the vice, how is that working for you? I would think you would end up with the top only gripping the bottom of the work, I’m interested to hear what happens in real life.
Keep up the great work at LAP!
Jason,
I built this to the print from Plate 11 in “L’art du menuisier.” And that is where the screw is located.
The vise works fine. The interior face of the jaw is relieved just like Roubo specifies. And the bottom of the jaw *almost* scrapes the floor. Those details conspire to make a vise that works just fine. I’ll post more when I have some time to shoot some photos.
Are the ends of the top sealed?
Nope.
Makes sense. I wonder what the timber framers say on the subject? So are you saying I should now feel procrastination guilt for delaying my Roubo (to satisfy the yr/in rule) from those 4″ 2yr old slabs of red oak drying out back?
I think he is lucky so far. Wood remains stable when the MC is above the fiber saturation point which is approximately 30%. Moisture readings above the fsp represent liquid stored in the cell cavities and loss of this stored water does not affect stability of the wood. Moisture readings between oven dry and fiber saturation represent water stored in the cell walls, and any change in MC in that range causes the wood to shrink or swell. The faster the change occurs, the more dramatic the movement. I suppose if the moisture change is a slow enough, the drying stresses can dissipate without any major checking, etc. Its hard to believe the wood was at 60% after 13 years, but storage location, protection, etc can affect the drying. I stored some Doug fir 2×6’s in my garage for a year with 1″ stickers, and the MC was down to 20%. It started out green and was probably above 60%.
For maximum stability, I still believe it is best to do the milling and assembly as close to the final MC as possible.
Well I am not alone in this experiment. We have a dozen more benches built using the same wood and same joinery. We will see how it plays out over time.
It is interesting to me that Roubo says nothing of the seasoning of the wood for a workbench – though he discusses it many times in reference to cabinet work.
I have to wonder about our forefathers and the moisture content in their benches and work stock. Without the benefit of our modern controlled interiors, air conditioning, central heating, and sealed windows, how did the mc of their wood stack up? Plus, our woods today are new growth trees that are grown rapidly, and I bet tend to check and warp much more than the old, slow growth trees grandpa was used to working with. Has anyone seen any written accounts of these issues?
Chris, not sure why type of MC meter you have, but for comparison my bench (one of the French Oak benches) measures at 12-15% with a handheld Lignomat unit. I have had the hold fast in a bench top dog hole for the past couple weeks, and have never had any rust develop on it. I recall the holdfasts on Bo’s bench getting rusty after just a couple hours.
Not sure why the MC indicators appear to be lower. Perhaps the plank where my top came from was stored in a position that had slightly better air circulation over the past 10 years.
I have only had the top flattened for a couple weeks, so not enough time to see any significant movement. Also, my bench is in a garage workshop, so more subject to the outside humidity levels – perhaps acclimating to equilibrium MC at a more gradual rate than a bench in a shop with air conditioning.
Will be interesting to see how the other’s benches are affected by MC as they come together.
Thanks,
Brad
Interesting topic. As an amateur woodworker and someone who has been milling some of my own lumber for a few years now I can say that moisture movement in wood can at times be a mystery. It seems certain species are very temperamental while others are much more forgiving. Like for instance, it seems red oak is very finicky and likes to be at least 10% mc before it decides to settle down and behave. On the other hand I made a key rack for our foyer out of a scrap piece of walnut that read 15% mc and it has never warped one bit. I air dry my rough sawn boards in a homemade drying hut (a thick clear tarp draped over an archway of metal cattle panels). I live in central Illinois and find that dry Illinois winters provide the best time to dry green lumber. I’ve dried 2-inch boards to 12% mc within 6 to 9 months, depending on the time of year.
Amazing that the benchtops you mentioned were 30-60% !
Do you have a band mill by chance? I have a hand full of 6X8 white cedar barn beams I need milled and I’m not far from you. 🙂
They proved to be too heavy for my 14″ band saw.
I don’t have one yet. I’ve been looking at Lumber Smith, or something similar that I can haul in my pickup. I’ve been milling with my Granberg small log chainsaw mill. Not as efficient as a band saw mill but a lot cheaper, especially for a hobbyist like myself who does occasional milling. With my more recent stint at being laid off I’m considering more lumber milling Heading out in a few minutes to cut down some birch trees for a relative. I will have four 12″ diameter logs to slab up soon.
I also have a slab bench top for my Roubo – 3″ thick white ash from three 8″ wide boards – and I have no idea the MC of the timbers I started with, only that they had been air drying for an unknown number years at the mill on the Texas gulf coast. High humidity environment to say the least. As I recall the boards felt a little damp in the center when I cut into them down and bored the holdfast holes.
Anyway I’ve had the bench more than four years now and have only had to flatten it twice since the initial build. The top has shrunk maybe 1/16″ and some of the legs are now a little proud of the edges but I’ve experienced zero movement issues as it’s dried and none of the little surface checks have spread either. Don’t know if it’s due to the Roubo bench design or just something about thick hardwood slabs but I’ve been pleasantly surprised at the lack of movement issues.
Hey Chris,
This is a great posting, thanks for sharing it…
I am always fascinated about the subject of working with “green wood” (read wet) and the alternative to dry. It still surprises me, how surprised folks are that you don’t need “dry” wood to be successful with you build. Your experience thus far sounds just about right for working with “heavy stock.”
As a Timberwright and general maker of “all things wood,” I work almost exclusively with wet wood in its varying permutations, as did most of our forbearers. Be it a timber frame, log structure, furniture or good old massive workbench, if you “read” the wood well, understand it’s nature, and follow the old ways of building, wet wood will perform as well, (sometimes better) than dry…I have seen way more issues with “kiln dried wood” than I have wet. Current case in point is a “jeontong malu” 전통 마루 (traditional floor) that is pushing a small frame into distortion as the “too dry wood” takes on moisture. It even got to acclimate for several months that I still felt was not enough, but the sub contractor insisted on “dry wood,” when the traditional system would have only use green or air dried locally on such a floor.
Look forward to further post on this topic and about the progress of the bench as it ages.
Regards,
jay
Chris –
I noticed the mallet looks to be made from wood similar to the bench. If it is, what is the MC differential by side? Are you planning to post on building the malllet?
The mallet is live oak. Very different than the oak in the bench.
Comparable oak in my shop is at 5.5 to 6 percent MC. Checked it yesterday.
As long as the top is similar MC, it should shrink more or less evenly. You’d run into a lot bigger trouble if you’re constructing a top from boards with dissimilar MC.
The key difference that I’m not seeing mentioned in the post or the comments is how the lumber was dried. One can only assume here but based on the known provenance I can assume it was air dried. Air drying is much easier on the wood and not hardening of the fibers results. When lumber is kiln dried is it essential baked. As a metaphor take a soft and fluffy flour tortilla and put it directly on the oven rack. Within minutes you have a crispy and not at all malleable chip. Try to bend it and it will crack. Try to rehydrate it and it won’t take on moisture very easily or predictably. Now back to the wood in question. Air dried wood will pick up and absorb moisture from the air more readily because the cell walls are not hardened and still somewhat spongy. At the same time this spongy texture means it will flex more without breaking, hence the absence of checks. Now specifically lets look at the Oak. As we all know, White Oak is a ring porous wood. It has big pores too, all nicely ordered in neat lines or rings. This is essentially open space that the denser late wood can expand and contract into as the humidity changes. More space means less deformation of the board. Add to this the overall spongy nature and you have a recipe for a very forgiving and stable board.
In other words, viva la air dried wood! (says the guy who sells kiln dried lumber)
Nicely done Shannon. I’m learning!
I could not agree more with what you wrote Shannon…
One of the issues “green wood, air dried, and kiln dried,” we see this all the time with “modern” timber frames, and the “younger companies” promoting “kiln dried” timbers as being a “good thing,” …It is not, it is very bad, causing “case hardening” of the big timbers, and of course severe checking. Air dried or green, is the only wood to use…but that is just me…
Great topic, it is my belief that these heavy benches were made with Green slabs, and the joinery was such that it allowed everything to settle in nicely over time. In fact, it would not have surprised me IF the Slabs were put on Trestles or Saw Horses to use them while the undercarriage was built. Not much has changed from back then till now. We start working with what we have at hand and graduate up making improvements as time and materials become available. The 60% MC sounds like an error reading to me which can occur if the probes go into an odd density part of the wood or just happens to have hit a moisture pocket. If the wood was sawn 13 years ago and never allowed to have allot of fresh dry air pushed through it, Some of the stacked planks could still almost be green. Especially in Georgia. In closing, I suspect Joinery 500 years ago was mostly frame and panel and everything was draw bored together and panels floated to allow allot of Green Woodworking and allowing for seasonal movement.