I am hoping for some suggestions on a problem I ran into recently. I built a conference room table, a picture of which is attached. The table is appx 5’x6′ cherry. The table is constructed out of solid 8/4 cherry, with a thickened edge (3.25″ +/-). The table was finished on all sides. As you can see, the table is supported be a structure that runs the long length of the table. There is also a piece of 1/4″ steel plate that is welded to the base on which the table is secured with lags through elongated holes. The table recently cracked on one end. I can only guess it is due to the sheer mass of the wood and differential movement plus the load of folks leaning on the table.
I am not opposed to fixing it, but I would prefer to redo it to eliminate any future problems. I am looking for suggestions on how others would construct this table. I considered veneered mdf, a torsion box, etc…. Thoughts?
Thanks,
Replies
I can think of two possibilities. If the lags through the elongated holes was over-tightened, or if they were at the inner ends of their slots, wood movement will cause the crack.
Second, if you builtup the end with cross grain, that could cause your problem.
Third - if there is a flex problem, then an angle iron type support would provide more resistance to moving.
Fourth - Explain that wood is a natural material that naturally cracks :-)
Bill
You're missing the most likely one. Moisture loss through the end caused shrinkage that did not take place in the center. It's heating season and humidities are low. Lee
I agree, Lee. Any thoughts on how to minimize this problem? Perhaps slots in the bottom to increase exposed surface area in the middle of the slab, or raise the table slightly off the metal plate to increase air movement in the center of the table?
Thew,
What is the finish? It might be possible to soak the endgrain with a penetrating sealer enough that the moisture exchange will not be enough to cause cracking. But that's hard to do. Wood in the lumberyard is sealed with latex paint or wax. Checking is almost inevitable in thick slabs. That's one of the reasons why "engineered" construction with veneering (plywood, MDF) came into being.
I have a picture of a teak bench hewn from a solid log at least 24" in diameter which is now the diagonal measurement of the square ends of the bench. The wood has large, deep "checks" in many places. They give it a lot of character.
Not so "tongue-in-cheek," maybe your table can develop a few more cracks for that "effect." It is really not inappropriate to claim that such events are a normal part wood slab tables.
I recall a very large, thick slab design, featured in one FWW issue, that was accomplished by building a torsion box and veneering. I can't recall the designer/builder. It was a very effective solution. Of course, such designs have no real end grain, which is always a feature I like about thick slabs.
Rich
Thew
I agree with Lee that differential drying of the ends vs the interior is the cause, actually, a "check." But do examine the lags in the elongated slots. Are the lags being driven up against the slot limits by wood movement?
Rich
How far apart are the holes in the across the grain direction and how much did you elongate them? Also, are there large humidity swings in the environment especially from where the table was built and where it is now located and also seasonally?
Have you taken out the lag bolts to see if they are at the ends of their slots? Also how big is the steel plate?
Presumably the the grain of the built up edge runs the same way as the top and you oriented the slots across the grain.
John White, Shop Manager, Fine Woodworking Magazine
Yes, the grain of the built-up end runs with the grain of the top. I have not yet checked to make sure the bolts have room in their slots, but I recall the elongations being 1" or so on both sides. I agree that the mass of the wood is allowing the end of the tables to acclimate faster than the core. I am used to accounting for this from a surface drying perspective, but the additional dimension is new to me.
Thew,
I know you didn't design the table to have any extraneous elements, but you might entertain using a "butterfly" key to control the crack, a la George Nakashima's designs
Rich
I was just thinking along the same lines. I just worry that without solving the problem I will end up with a veritable flock of butterflies!
A quick calculation, 48" between bolts and a 12% change in MC, shows that the top could easily move 1 1/2" or more so you could easily have exceeded the capacity of the slots if you started out centered in them, giving you only an inch of movement in either direction.
All of the other possibilities mentioned could be factors but I would start out by seeing if you have run out of room in the slots.
John W.
Thew
There's a lot of guessing going on here. To help solve your problem, an actual picture of the underside of this table would be extremely helpful. We need to see:
1. How EXACTLY you fastened the steel plate to the table.
2. How much clearance was left for these lag bolts for seasonal wood movement. That table is gonna move quite a bit.
3. A picture of the crack/check, and it's relationship to the fastening bolts.
As a side note, I would recommend that you experiment with a new fastening method. I've built 2 tables similar to yours in design, with steel bases. I fastened the steel plate to floating dovetailed stock in dovetail runners in the underside of the table. As the table moves, it's free from the fasteners, so it's not restricted. And, the wedge shape of the dovetail holds the table very firmly. I tested this by having some of Lee Grindinger's sheep fornicate on the top, to see if it's sturdy enough! (Just kidding, Lee!) Seriously, though, this setup has worked well for me.
Jeff
Excellent idea on the dovetail fastening. Thinking back to the installation, I think I can make that work. I will be checking it out tomorrow and will report back.
Thew,
I don't have the technical expertise that these other guys do, but, picking up on one suggestion, tell the owner that this is part of the experience.
You could explain something like, Wood moves during the seasons and it was your plan all along to add those advanced design touches of butterflies (for which you are not going to charge him as that is how you develop for him a one of a kind table) but they had to be in harmony with the table and until the table adapted to its new environment and began its next stage of aging that you didn't know where to put the butterflies. And, now you know where to put some!!!
No charge for the marketing ---- :-)
Alan - planesaw
Again, thanks for the input. As an update, I looked at the table this morning, and there is still room for expansion and contraction in the fasteners, and none seemed to be overly tight. The lags are about 3.5' apart at their greatest separation, and I allowed 1" of travel in each of the slots, so I think I'm good there. I guess it must just be the moisture differential at the center of the slab. Anyone have any ideas on how to make sure this doesn't happen next time or continue to happen on this table?
First of all, this is a risky design for precisely the reason you've found but,Sprung joints might help because they put the ends under slight compression. However, sprung joints can behave unpredictably when there are that many involved. You'd need to work from the center out with each joint to make sure there isn't too much tension.Quarter sawn lumber is the most obvious solution if it can work in your design. Use veneer with a stable substrate.Put about 5 more coats of finish on the end grain as on the rest of the table. This will help with the fact that wood can lose moisture 15 times faster out the end grain than long grain.As some have suggested, use Nakishima's technique. I will say, though, that both George and, now, Mira let very large slabs thoroughly dry in their showrooms and sheds. The movement is predictable and the butterflies were added before the pieces left for client's homes. These splits were a design feature and this won't work with a glued up slab.Here are a couple of questions that might help predict future behavior of this check, Where is the table in this country? How dry was the lumber when it was used? What sort of glue did you use? What sort of finish did you use? How old is the table? During what time of the year was it built? Delivered? Is the house air conditioned? Your workshop?LeeOh, and Jeff, those sheep have been acting funny since you used them, they are awfully skittish about anything approaching from behind. Heh, heh, he...
Actually, these were not rhetorical questions. Where is the table in this country? How dry was the lumber when it was used? What sort of glue did you use? What sort of finish did you use? How old is the table? During what time of the year was it built? Delivered? Is the house air conditioned? Your workshop?I could ask more, like "How was the glue surface prepared? What is the grain configuration around the split?" and so on. Lee
Sorry, Lee.I'm in southern Minnesota, so the ambient humidity inside varies quite a bit over the seasons. The table is in an office that is climate controlled , but not humidified, so it is still drier in the winter than the summer. The table was built and delivered during the winter last year, but ambient humidity in my shop was probably greater than at the office. I cannot recall for sure the moisture content of the wood was, but I routinely check lumber in storage and in the winter is around 7-9%.I used Titebond II, and glued the table up in manageable sections to try to maintain enough clamping pressure. Edges were machine jointed, but due to the board weight I handplaned each board after to ensure I didn't waver on the jointer. The split began at a joint, and then diverged slightly into a section of wild grain. The table was finished with a shellac washcoat for color and a waterbourne lacquer from Target Coatings (USL).I realize there are a lot of variables here. Thanks for the help.On a side note for those interested, the USL has held up quite well over the year. I was a little concerned about scratch resistance. Although there are some errant ballpoint pen indents (which I expect would happen with most finishes) it looks good.
It seems clear that shrinkage at the ends is the problem. Slabs like this can grow without too many problems because the stresses during swelling are the length of the joint in the center of the slab. Moisture loss through the ends, on the other hand, exerts the most pressure where the joint is most vulnerable and weakest, on the ends.It's not possible to see how much clamp pressure you needed to pull these boards together but a rule of thumb is that if it requires more than hand pressure it's poorly fit. The location of the split says volumes to those that suggest PVA is stronger than wood. Initial strength might be but down the road it's not, you had a short road.I'd say the entire end was under tension the wild grain of the one piece had stressed the PVA previously weakening it. In a perfect world the wood have all had identical numbers of growth rings and identical grain direction. Unless you buy quarter sawn material, kinda ugly is cherry, you're going to lose joints like this. A mechanical joint such as glue joint, tongue and groove, or a spline would put the split elsewhere but it would probably still split unless you make sure the moisture loss and gain is equal throughout the slab and that the comings and goings of the individual pieces do not stress the joint by being contrary to it's neighbor.Well, enough blathering,Lee
Thew,
Couple of questions:
1) the slots in the steel plate are running crosswise to the length of the top, right?
2) what means did you use to attach the wider banding at the end-grain to end-grain joint, on the ends of the table?
Regards,
Ray
This forum post is now archived. Commenting has been disabled