Read Part 1 here.
When predicting what a thick workbench top will do, I don’t consult the tables in the Forest Product Lab’s “Wood Handbook: Wood as an Engineering Material.” Instead, I prefer to think that building a thick benchtop is like moving into a haunted house.
Yes, the haunted benchtop is going to give you trouble at first. It’s going to scare the plop out of you when you stare into a gaping crack that seems to open up more each day. Your tools will behave erratically on the new benchtop, like they also have been possessed.
But if you are patient and observant, you will come to an understanding. Soon the ghost will be serving you tea.
Does that magic moment occur when it dries out and reaches “equilibrium moisture content?” No. I don’t think that thick benchtops follow the same rules of moisture loss that we expect from thin stock.
Benchmaker Richard Maguire recently told me about a slab-top workbench that he built from 50-year-old stock. After 50-plus years of drying, it still gave him fits.
“After all those years I imagined it was as dry as it was ever likely to get, and yet when I cut or drilled in to it it was apparent that there was too much moisture still in the middle,” Maguire wrote. “Thinking to the original French benches with these thick slab tops, I’ve begun to feel that these must have been very wet when built. Just based on the amount of time it takes for this thickness of timber to dry, I don’t imagine they had a good 50 years or more to leave it lying around.”
His experience lines up exactly with mine. When dealing with thick stock, getting the wood spirits out of the middle are next to impossible – and probably not necessary.
After observing how these benchtops behave, I have a theory for you to consider. I hope to ask the scientists at the Forest Product Labs about it someday. Here it is:
When you cut open these thick slabs, they dry from the outside-in and from the ends primarily. (That’s not the theory; that’s well-established). As the outsides dry, they move as they start to reach equilibrium moisture content. The middle stays pretty wet. Why? Don’t know. Perhaps something about being surrounded by so much dry wood keeps the moisture in.
But here’s the important part: I’ve found that the outside gets dry and hard – like the shell of a lobster – and is less susceptible to gross movements by further drying from the inside.
So waiting for a thick slab to reach equilibrium might never happen. And it might not be necessary.
A good example of this is the all-handtool Roubo bench I built for Popular Woodworking almost four years ago. The top had been in Ron Herman’s log yard for a long time. It was wetter than the French oak we used in Barnesville, Ga., this summer. And it was punky. But it was what I had to work with, so I built the bench with it.
The first year with the bench was rough. I flattened the top twice. It shrank around the legs, leaving the end grain proud. It didn’t, however, distort all that much. I thought of the top like a slat-bending frame you use for chairmaking. You put a wet slat in a bending frame and it stays that shape. You put a wet benchtop in a frame of legs and it is somewhat restrained from distortion by the legs.
It has been three years since I messed with the top of that cherry-top Roubo, and it is still functioning quite well. After my oak bench went wonkers on me, I decided to see how out the cherry bench was doing. Three passes with a jack plane brought it back to perfect.
So I wouldn’t be afraid of slab benchtops – even fairly fresh ones. Just remember my theory, which I’m sure is as accurate as my theory that a gnome or small dwarf in my stomach makes me sick each winter.
— Christopher Schwarz
So when can we expect “Make A Bench From a Tree”? Complete with The Schwarz hand drawn plans for an Alaskan Sawmill.
So when can we expect “Make A Workbench From a Tree”? Complete with The Schwarz hand drawn plans for an Alaskan Sawmill.
Chris, you said something about the holes being wet when you drilled into the benchtop. That gives me an idea: What would happen if you drilled holes all over the bottom of the bench? I’m thinking of maybe 1″ holes in a grid every 4 to 6″ apart. They should probably stop 2″ from the top surface. This would, in my thinking, allow the preponderance of H2O in the center of the slab to escape, in a reasonably balanced manner, so that the ghosts of the slab could be put to rest. I’d be interested if anyone would be willing to attempt this, but I myself will likely never have a slab benchtop on which to try it. I already have a rather nice antique English-style benchtop…
Just encase the entire benchtop in a nice 1″ coat of bar top epoxy. No more moisture loss.
Or: Simple make the ENTIRE BENCHTOP out of bartop. No wood.
Genius you are.
Or skip the bench all together and just go to the bar.
Which I shall now demonstrate by having dinner at Free State Brewery.
Rob,
Make sure you have no beer movement in the bar.
I think you know what I am saying.
Oooh you could even embed misbehaving tools in the middle of the 4″ of bartop, as a warning to others. Or Raney.
The bartenders might not like it when you try pounding a hold fast into their bar. “Can I get a Schlitz, please? Oh, and where’s the leg vise?
-Tim Aldrich
With this bench you’ve entered the realm of timber framing and there is no more reason to wait 100 years for a timber to dry before framing it (in fact that makes it harder) than there is to do the same for a bench.
But the thinking has to change. Mistaken assumptions likely being being made :
a) you have to do everything to get it dry. Timber framers simply worked with the understanding that wood was wet and will change shape over time and created joints that stay tight as it dried (cupped mortise faces, drawboring, shortened tenons) and created structural integrity to control the movement and maintain the intended shape.
b) it matters that it may still be wetter in the middle at some point – even 100 years down the road. I’m sure it makes not an iota of difference as long as structure and joints are properly made and accounted for as in (a) and the wood does not rot – but for this suitable species were used (like oak)
It is only recently that we have become obsessed with using dry timber and this was mainly because
a) it eliminated the need for some skills in the understanding of how wood moved so lower skilled (cheaper) labor could be substituted
b) lower quality wood could be used since drier wood becomes more dimensionally stable.
c) there was access to loads of cheap energy allowed the economies of (a) & (b)
Well, if you’d stop eating gnomes and dwarves they wouldn’t make you sick anymore!
I would be happy to go over with you the nuances of the diffusion equation and how it explains why a thick piece of wood can stay so wet for so long. In the meantime, take a look at Figure 12-2 in the FPL Wood Handbook.
Steve,
I am serious about this: You should be writing about this stuff. Not as a commenter, but as a full-on blogger.
Second!
It’s not so hard.
You’re right, I should. And like my work colleagues say, I should blog about my investigations into blending functional and object-oriented programming techniques. And like my birding friends say, I should blog about birds and identification and taxonomy (I actually did write a monthly newsletter article about birding many years ago, in the pre-Internet days; it lasted about a year).
But I probably won’t do any of those things, because I find writing to be excruciatingly, viscerally painful. You know how some authors are tormented by demons, and write in order to exorcise those demons? For me, it’s the other way around: it’s the act of writing that creates the demons that torment me.
Steve –
Confused by your reference to figure 12-2 of the FPL Wood Handbook on several levels:
1) No publication at FPL has the simple name “Wood Handbook”. Closest I find has full name “Wood Handbook: Wood as an Engineering Material.”
2) Within said handbook (http://originwww.fpl.fs.fed.us/documnts/fplgtr/fpl_gtr190.pdf) chapter 12 is titled “Mechanical Properties of Wood-Based Composite Materials”.
3) Within said chapter 12, I find no figure 12-2.
So I’m assuming either I have found the wrong handbook or there is a version miss-match between what I have found (link above) and what you are referencing. If I do have the right publication, just a different edition, perhaps chapter 13 or chapter 4?
Please clarify and provide a link to the correct publication if possible. Thanks.
It’s 13-2 in that edition.
Thanks! Chapter 13 seemed a bit more on-target.
I am still trying to wrap my head around the fact that Chris managed to write a post without insulting some race, ethnicity, religion, or other protected class.
Rest assured that gnomes and dwarves everywhere are offended.
Best imagery on your blog so far. I kept imagining your apprentice Casper, the friendly ghost, sawing some sticks for you to work.
Out here on the wet coast the carvers that create columns for buildings will cut a chainsaw kerf the full length of the log to the pith to relieve stresses and let moisture out. It’s hidden on the back of the carving. Seems to reduce checking. Perhaps a kerf or 2 on the bottom of the bench slab would be of value.
As a lad my family visited ‘back home’ in Appalachia many times. I had no specific interest in woodworking at the time, but I was always interested in how resourceful these kinfolk of mine were. A couple times I saw work surfaces made from half a log. They look very much like Roman low workbenches, but on a Texas scale, i.e. 20+ inches across and at least 7 feet long. Reading this post brought the memory back. The reason i’m commenting is that they did indeed have a V notch cut into the bottom still round portion of the log. They could have been one log split and then abutted against each other, or half a huge tree with a relief cut; too many years have passed to find out for sure now.
You said you flattened the oak bench in 45 minutes. I spent almost 2 hours the other day trying to flatten my hem-fir roubo and it’s still not perfect. Either I’m being too fussy or I’m missing something rather obvious. When you say you made 3 passes to get the cherry bench back to perfect, did you go solely straight across the grain or did you switch to opposing diagonal strokes? Did you chamfer the far outside corner to prevent spelching? Did you switch and plane from the other side after each set of passes? Any clarity you can provide would be greatly appreciated. Thanks.
And if you use a French slab, it is rumoured that the ghost of monsieur Roubo himself shall walk about your shop at night and sharpen your chisels.
Wood moves for ever. as log as the general climate changes, so does wood. People in england and arizona don’t have the problems we have with wood. They are always humid or always dry where we are at zero humidity or 100%. In the winter i keep my expensive ultimatum braces in my humidor to keep the heads from cracking. After 150 years of life all it takes is one winter in N.Y. for the head to split wide open. So here’s what you do. Build a humidor big enough to…
Would it be worth sealing the end grain and dog holes to slow the drying make it more even across the bench?
A coat of sealer in the dog holes would also raise the grain to give good grip and stop the bench dogs going rusty when you leave them in the hole too long.
Who is to say that Roubo didn’t do this but thought it so obvious, it didn’t need mentioning?
If I recall from my wood technology classes in school if the slab includes mostly heartwood from the center of the tree it will never really dry. This is due to the pits closing and not allowing moisture to pass between fibers. In sapwood these are still open and allow more even drying. I think you just have to plan for movement in your joinery and do some regular flattening and all will be ok. My split top roubo which is laminated from 8/4 hard maple has moved a fair bit since building it last year and needs to be flattened so I don’t think this problem is limited to slabs. In my opinion this is a great excuse to own a number 7 plane for bench maintenance.
Makes sense if you think of what “equilibrium with its environment” really means. For the outermost layer, its environment is the air and there’s a lot of thermodynamics going on there to wick moisture from the surface. But underneath that layer, the next just needs to find equilibrium with the outermost layer. And it doesn’t have all that pesky surface area to deal with, so it can be lazy and never quite reach equilibrium. And then the layer below that just has to deal with the lazy layer outside it (and it can be just as lazy). And so on, to the point where a sufficiently-thick top will have a liquid core.