Stability in a Large Solid Wood Table Top
Pros discuss moisture-related movement and the associated design issues in a big table top. July 24, 2006
Question
I am making a table which is to be 18 feet long by 6.5 wide at the middle, tapering to 5. The material is to be walnut or mahogany. Thickness should finish at about 1 1/4 inch or more. Top will be able to float on support frame. Straight taper, but may evolve to a boat curve along the length. Two 9 foot lengths join with splines and tite joint hardware at the beltline.
The concern I have is the size of these two slabs, and the need for stability. Also how to manage the movement that normally occurs, maybe even to the extent that the design is a top that is joint- or crack-repairable in sections without cutting into a 6 x 9 slab.
Questions:
- Flatsawn or rift/qtrd. stock?
- Grain orientation with or across length?
- Preference of mahogany over walnut for stability? I like walnut for appearance.
- Option of making up each half out of say, three sections that are spline and tite jointed the long way together like shirt buttons, to accommodate movement and maintenance down the road.
Called a conference table, but will get a lot of food service. All suggestions welcome.
Forum Responses
(Furniture Making Forum)
From contributor D:
The standard answer is that these things are more difficult than you may think. The potential problems don't magnify or multiply, they cube. Solid wood, field joints, and food service add to the inherent problems and make things truly unruly. I'm not questioning your ability, since I don't know your ability, but I have seen several of these things go horribly wrong. This has been discussed at length before, so a Knowledge Base search will yield good info.
From the original questioner:
Thank you for the words of caution. Latest plan: Since the effect called for with this tabletop isn't formal or high-end, my interest is in solidity and function. So I plan to use qtr./rift 8/4 stock, orient the grain with the width, not the length, and make it in 6 glued-up sections that will be dry-joined with the tite-joint connectors. A bigger shop will shape and profile the sides (CNC), sand in the crossbelt machine, and finish as a unit. Thorough finishing on underside and edges of sections. Entire baby to park on lengthwise steel rails. Comments?
From contributor J:
Have you calculated the total seasonal movement you can expect along the 18' of length? If so, how is the apron of the table going to allow the top to move that much? Boards usually run lengthwise on a table top because then the amount of movement is much less across with width of the table, and can be dealt with. In the case of this table, the 6.5' width would be much easier to deal with than the 18' length - almost 3 times easier, regardless of wood selection.
From contributor C:
The "Wood Shrinkage Calculator" (link below) at WOODWEB's calculator page will provide a quick (and accurate) value for the predictable movement to expect.
WOODWEB's Shrinkage Calculator
From contributor P:
I think the plan is feasible as long as you have a floating attachment to the steel. If it's an I-beam, then a series of simple L-shaped wood blocks which are screwed to the underside of the table and reach over the flange would work fine. The calculator isn't exactly helpful for seasonal movement, but seems to imply that the total change in length could be 2" or more. I think that running the width across the short dimension is actually the best way to go - as long as the client doesn't object to the look. It's a lot easier to restrain cupping if you do it that way, and the tite-joints will be easier to install. Incidentally, you might consider going to 5 panels instead of 6, as tite-joints are extremely slow to install and adjust, and it will save you. You didn't say whether you had decided on walnut or mahogany, but the end grain of the mahogany would be relatively soft. I'd use walnut. Building large tables can be very complex, but this one should go well as long as you finish both sides the same, and don't fasten it too tightly to the beam.
From the original questioner:
I grant you it is unusual to have grain orientation across rather than with the length. In domestic dining tables our visual vocabulary permits extension leaf joints and sometimes also wood grain across the table's width. The room has natural/window light coming from the long wall opposite the entry, across the width. So the section joint lines won't show up as shadows, and wood grain will display well. Expansion and contraction will be ferocious, but the L-blocks would keep it close - not tight - to the steel. Maybe some UHMW-faced tape on the steel would give some extra slip. Thank you a lot for the benefit of your experience.
From Professor Gene Wengert, Sawing and Drying Forum technical advisor:
Mahogany is about 20% softer than walnut.
The joint in the middle will not be tight no matter what hardware you use. Joints will gap in the winter and summer. In any case, you need to coat the entire table with a thick, vapor-proof coating to slow down moisture changes and gradients that can develop.
From contributor P:
Why wouldn't the joints in the middle stay tight? If he secures the top to the beam at one point in the middle and lets it float at each end, the joints should stay tight even as they move around.
In my opinion, the biggest danger with a table like this is warping caused by temperature differences from top to bottom - think sunlight shining on the top for a few hours while the bottom stays cool, or forced air heat hitting the bottom of the table and not the top. Or, god forbid, if the client puts glass on top of the table. I would use a pair of steel I-beams, put them 18" apart, secure the top in the center, and let the top move where it wants to. It's important that the base design doesn't seal off the bottom of the table from air circulation. The only downside to running the grain short way would be at the ends, where the top will be cantilevered. With 8/4 stock you would be okay, but with a 1 1/4" finished thickness, it would be vulnerable if someone sat on it. Perhaps an auxiliary cleat with slotted connectors to stiffen up the end of the table?
From Professor Gene Wengert, forum technical advisor:
Moisture regain or loss in the table will be more likely on the edges before the center. This is because there is more exposed wood on the edges. Hence, in the summer, the edges will swell and there will be a gap in the middle. In the winter, the edges will dry and shrink and there will be a gap on the edges.
I do agree that there is a big risk of cupping if the moisture on the top is different than the bottom. This is why a good vapor coating is needed, top and bottom. With a good coating, you will minimize the solar heating affect (which is really caused by the lower RH and therefore lower MC on the heated side). (Technically, heat does not cause wood to expand or contract any appreciable amount. It is the moisture change brought about by humidity changes.)
From contributor C:
That was an excellent point made about the edges tending to move more than the faces. Perhaps impregnating the end grain with epoxy might help even it out. I also think the UHMW tape on the steel is a good idea. One thing, though - how are you going to transition the steel, or feather it out toward the ends? The ends will need supports nearly out to the edge because of the cross-grain construction. You might consider building some sort of wooden structural element to take over for the steel in the outermost 2' or so. A piece that spans between the steel rails, and goes from the steel hgt to very thin at the edge, leaving the last six inches or so unsupported, but still able to slide.
From Professor Gene Wengert, forum technical advisor:
I should add that if you attach the middle joint really well, the shrinkage or swelling force can often crack the wood elsewhere in order to relieve the stress. Note also that the larger the table, the greater this stress.
From contributor P:
I agree, and I have experienced this kind of movement in tables we have made. (Smacks self in forehead.) Extra sealing of the edges will be critical. The client should be warned to use humidifiers during heating season.
