*
After a long discussion regarding the possible pitfalls of a walnut slab entry door, my clients and their architect want one. We discussed problems, such as wood movement, cupping, cost, and weight. Then, a few moments ago, I was awarded the commission to add this to others of this Silicon Valley remodel’s artistic projects. Before we all travel up to Peter Laing’s sawmill in Windsor, California, and especially before I take tools to this expensive slab (at least $1,500 for the single piece), I hope one of you made a similar door and has some advice for me.
This’ll be a 2½-in-thick, 3’0” by 7’0” door, with a squared, but live edge on the lock side, and a straight, plumb edge on the hinge side. Although the architect hoped for a one piece door, I believe breadboard ends are appropriate to arrest cupping. I’d put a 6 to 8-in-high rail on the top, and a 9 to 12-in-high rail on the bottom. These rails would be joined to the slab via sliding dovetail joints, glued at the hinge side, to allow the slab to move while keeping a straight edge against the jamb’s weather stripping.
Wood movement, I believe, is the single most important problem to overcome – others, like weight, I can deal with. These questions come to mind:
1. Although we want a figured slab, what movement problems can we expect with full width slabs from the butt of a walnut tree?
2. Should I look for a piece cut from a particular part of the log, i.e. close to but without the center?
3. Laing’s walnut slabs are air dried eight years (I’ll take my Delmhorst along) – is this a problem? I’ve never made a door of anything but kiln dried lumber.
4. Should I expect more wood movement at one end of the slab than the other?
5. Aesthetically, sapwood is o.k., but will it present movement or finishing problems? I usually avoid it.
6. What questions am I forgetting?
I once made a door with 289 parts – a 4-part frame, ¾-inch plywood panel substrate, and 142 pieces of ebony and white oak epoxied on each substrate face. That project seemed simple in comparison to this one.
Can anyone help, please.
Gary W.
Replies
*
Gary,
What did you mean about the architect wanting a "one-piece" door? Is he crazy, or just uninformed?
*Sounds like a fun project.If I were you, I go ahead full speed on the expertise and liability of anyone execpt yourself. If the architect wants it and the client wants it, then by God I want it too. That's my attitude always. But, I always, and I say again, ALWAYS, tell them that this is something that I have never done before and can therefore express no warrenty at all. But I tell them I will be glad to do whatever they need. And that I am as excited as they are about a possibly favorable outcome. It sounds like you've found the wood, all that's left is hanging the door and hanging the liability on the client or architect.Good luck,Ed. Williams
*I'd focus my efforts on realizing proper jamb and lockset specifications. What I mean is, the slab will either remain stable enough to be serviceable or it won't...not much you can do about it. Doesn't seem to me that your intended rails will have enough strength to resist the forces of distortion that (and this is important) will be modified by the fact that one side is hinged to a rigid jamb and the other isn't. Likely what you'll end up with is an overall twist that will prevent full seating against the jamb-rabbet on the strike side. This in turn will, at minimum, put additional stress on the lockset as it will be called on to flatten the twist each time the door is "forced shut." My main point is that the rabbets will have to be deep enough to accomodate a slab that will undergo as many changes as the weather. The lockset will also have to be able to "reach" and "plunge" as the seasons change.I'd be prepared to build jambs that conform to the needs of the slab, meaning, the strike-side rabbet may end up not being so much plumb, as parallel to the door edge. Maybe you could fab. the slab and hinge jamb in the shop then hang the door in its opening and see what you get. If the strike edge runs plumb with the hinges hung, then cool, but if it doesn't, no problem, just mill the strike and head rabbets to match the reality and nobody will be the wiser. If it were me, "empirical approach" would be my mantra. Accuracy is something quite esoteric on a project like this and the eye is the final arbiter, meaning, legerdemain and trompe le oeuille rule the day.
*I've seen a pair of doors like that only they were made from a white oak tree that had sunk on the floor of Lake Superior for about (est.) a 100 years. The tree was brought up and air died for three years and then a slab cut. They were carefull to not use the heartwood. The original tree must have been at least 7 feet across. Apparently the water pressure at the depth they were at "dried" the wood without rotting it because of a lack of oxygen at the depth it was at. Removing it from the water was enough to dry it without the normal movement that can cause checking. I can't say how much it moved because it hung about an inch away from the jam. The pivot point of the doors was a good foot away from the wall so part of the door was counterbalanced. The door wasn't weathertight, that job was done by the outer doors. Sounds like a fun project, usual warnings of which you've already heard many.
*Freelance,You bring up some interesting points, especially that about making the jambs conform to, at least, the original shape of the door. I've also been thinking of some "less than common" lockset - something with an adjustable strike. 'course, that's just in my imagination right now. What comes to mind is a lock with a drop down lever, like a gate latch, but then that's also in my imagination.I'll certainly leave more than the usual 1/8-inch gap around the door. I've considered calculating possible wood movement, which is a task I've done successfully on other large projects, but never on a single slab like this. My best guess is that the slab should react the same as a glue up, but with a range of grain structure, from flat to vertical. Cupping is really my greatest concern - I think. Twisting in a long piece, I think, might be controlled by anchoring one edge on the hinge side.By the way, I've already insisted that this is a chancy project, one that I've never done before, and that I don't want the liability. Will that hold up in court?
*PEG -- polyethylene glycol
*I have built door made from boards no more that 3 inches wide.. Stability will be a problem. There has been some problems with the doors made this way. And walnut will bleach out when the sun gets on it.. Be sure to stain it with a aniline dye for color fastness. I have made mantles out of big slabs. I got an air dried slab and had the sawmill saw it again to make the slab straighter.
*I've got a one slab black walnut table that has never cupped ( a cherry one that did) but a door that size out of one slab would certainly be a risky endevour. I think I would at least rip it a few times and rejoin to help control the cupping. Good luck.
*I hope they are paying you a mint....and that you have "no warranty, period, capiche? no warranty of any kind whatsoever" in print somewhere, and everyone has signed off. I would anticipate greater wood movement than in a similar sized glueup (across the width), and as mentioned before I would try and anticipate as much winding and cupping as possible in the design of the rebates, or however you plan to do that. I think it was mentioned around here or at Knots too, when a big slab question came up...might be worth a call to the George Nakashima people...if anyone has done more than one door in this style, my guess is it would be them. I hope it works out.
*I've never seen a walnut tree more than 3 feet across. Are there really logs that big available ?
*One thing I might do on a deal like this is discuss with the arch./builder/owner an alternative, more conventional design that would be acceptable should the original approach be deemed unacceptable. A cost could then be determined for removing the orignal and fabbing and installing the alternative. The total monies involved, meaning, cost of original work and the potential alternative, could be deposited into an escrow acct. Upon install of the primary door assembly you get paid that amt. Then, for a year or two, or three everybody watches. If necessary, the alternative can be built and installed without a pall of concern about getting paid in full.In general, I'd put as much thought into the contract as into the door, with ample enough payment up-front to allow you to research all considerations -exhaustively- before you accept the job.Also, don't count on the hinging to control twist, inherent tendencies to do so will merely manifest entirely on the strike side.
*What would happen if the door slab was resawn through the center of the plane, the two thinner slab pieces reversed and placed face to face, glued and vacuum bagged? Then jointed and finished normally. Would the twisting forces cancel? Would the glue hold? Would that allow the full slab view without stiles? I don't think the glue line would be noticed on the edges.Also, could you kerf the top and bottom edges of the slab and insert a steel T-bracket, with the edges dadoed deep enough to cover the T with a filler? (Instead of a top and bottom rail to prevent cupping).Huh??
*Whew, good luck w/thatone
*Straight grained Walnut is well behaved wood, BUT, 1. it's an entry - unless it's under a porch or alcove or in some respect protected from direct sun, I don't think you can keep it stable enough to keep it working year round.2. a figured slab air dried eight years is quite stable - until you have different temperature/humidity conditions on opposite sides of the piece - 3. if I had the choice, I'd take the slab sawn out of the middle of the tree and not be scared of it containing the center - quarter sawn walnut is very stable - it is 2.5" thick after all, I would be wary if the piece did not exhibit a similar ring pattern on both ends of the board.4. if you go with the breadboarding technique, I'd suggest you attach the hinges to the breadboards rather than to the body and be very generous with the rabbet - 5/8" or so. attach the slab to the beadboards at the center - then the seasonal movement of the slab at the latch will be 1/2 what it would be if attached at the hinge side.5. if the situation is such that the door will not have to endure extremes, you might consider rabbeting the ends (top and bottom)and installing 1 1/4" x 3" boards - straight grained and stiff - into the rabbets to keep the panel straight. would be more discrete than the breadboards.6. previous tip on building the jamb to the door is right on.I have built a door this size out of 2 boards of northern pine, battened, installed on the north side of an 1840 log cabin under a porch - finished with boiled linseed oil and stain - heated/AC inside, mid west -20 F to +95 outside- 8 years old - works well and consistently - dimension changes only about 1/8" over the course of a year - the thumb latch (drop latch) is a good idea, but commercial ones are too puny for this application - custom units are available to fit the grand proportions of a door such as this. good luck.
*I've built 3 doors with "floating" 2" thick oak/walnut/teak slabs inside a 3" thick frame, and the slabs still move and crack (Seattle area). Only one that did not move was the oak one totally covered with polyester resin, which your client probably would not want to do. Have not tried the polyethlene glycol treatment, but I've read for that to work you need to start with green wood and let the poly replace the moisture over a few month period, after which the house would smell like antifreeze for a door sized object.
*Although no one on Taunton's or other sites has actually made an entry door of a single slab, I've received a lot of good advice. I'm going to Windsor, CA next Wednesday, where the owners will purchase the slab, and then I'll build the door. I've issued all the disclaimers in my signed contract, I gave them no guarantee, but I said I'd return, at $85/hr., to make adjustments.Here's how I plan to do it. In an attempt to arrest cupping, I've decided to use steel instead of wooden breadboard ends. They'll give the door more resistance to cupping, and they'll satisfy the architects desire to have the slab naked, without ends. I plan to have two, 3/8-in-thick, steel T-bar weldements made at 2-inches wide across the top and 4-inches high. The hinge side of the weldement will have two bolt holes, at about 3-inches from the top, and the latch side will have similarly placed a slot. I'll bolt the bar into a matching groove in each of the door's ends - solidly at the hinge side, and loose enough to allow for movement at the latch side. I'll recess the bolt's the head and a thin jam nut, and then plug the holes.To allow for possible twist, the jamb's stops will be adjustable: the latch side stop will be anchored at the latch bolt, but will pivot, top and bottom, at to adjust for possible future twist. The top stop will be adjusted similarly, the hinge side stop will be fixed, and the door bottom will have a semi-weathertight sweep. Weather stripping will be inserted in a groove on the door side of a wider than normal stop. I've written a spreadsheet wood movement calculator that takes into account the wood's current moisture, relative humidity, and predicted RH where the door will be hung. I've come up with results close to TAB's. I'll adjust my figures after measuring the workpiece after it stands in my shop for a month.This project is exciting, because I think I'm breaking new ground. The architect and owners have agreed to take the liability for future problems, and I have agreed to amass as much knowledge and experience about the subject before proceeding to use my best woodworking skills to give them a good product. That's what this business is all about: learning something new always requires that we take a chance, and I've taken quite a few, but not without giving the subject a lot of thought.Anyway, I'll let you know how things work out. Thanks again.
*
After a long discussion regarding the possible pitfalls of a walnut slab entry door, my clients and their architect want one. We discussed problems, such as wood movement, cupping, cost, and weight. Then, a few moments ago, I was awarded the commission to add this to others of this Silicon Valley remodel’s artistic projects. Before we all travel up to Peter Laing’s sawmill in Windsor, California, and especially before I take tools to this expensive slab (at least $1,500 for the single piece), I hope one of you made a similar door and has some advice for me.
This’ll be a 2½-in-thick, 3’0” by 7’0” door, with a squared, but live edge on the lock side, and a straight, plumb edge on the hinge side. Although the architect hoped for a one piece door, I believe breadboard ends are appropriate to arrest cupping. I’d put a 6 to 8-in-high rail on the top, and a 9 to 12-in-high rail on the bottom. These rails would be joined to the slab via sliding dovetail joints, glued at the hinge side, to allow the slab to move while keeping a straight edge against the jamb’s weather stripping.
Wood movement, I believe, is the single most important problem to overcome – others, like weight, I can deal with. These questions come to mind:
1. Although we want a figured slab, what movement problems can we expect with full width slabs from the butt of a walnut tree?
2. Should I look for a piece cut from a particular part of the log, i.e. close to but without the center?
3. Laing’s walnut slabs are air dried eight years (I’ll take my Delmhorst along) – is this a problem? I’ve never made a door of anything but kiln dried lumber.
4. Should I expect more wood movement at one end of the slab than the other?
5. Aesthetically, sapwood is o.k., but will it present movement or finishing problems? I usually avoid it.
6. What questions am I forgetting?
I once made a door with 289 parts – a 4-part frame, ¾-inch plywood panel substrate, and 142 pieces of ebony and white oak epoxied on each substrate face. That project seemed simple in comparison to this one.
Can anyone help, please.
Gary W.