I just finished (ok, I still need to poly it) a dining table top. I have not decided on the type of legs I will be attaching to it yet.

One of the two leg types I was considering was 4 "post" legs (3"x3"x29") attached to a 1" x 3.5" skirt with two lag bolts for each leg. The legs will be set 1' in from the ends and 10" in from the edges.

Will I need some kind of brace for the center of the table? Or will just the skirt + 4 legs do fine? I'm somewhat worried about sagging, but mostly worried about stability.

The tabletop is made from Sapele and Hard Maple.

  • You're not just supporting the top with the legs, you're supported by the 'skirt' (apron). Plug some numbers into the Sagulator and you'll easily get a good idea of how much deflection you can expect in your proposed design. As the top is made from sapele and maple my gut feeling is you'll be fine here as far as sag goes, stability will be down to how well you make the leg assembly. – Graphus Jun 9 '17 at 7:14
  • @Graphus looks like the sag is half of what'd acceptable. Can you add an answer? – Jacques ジャック Jun 9 '17 at 12:14
  • What figures did you use? With the numbers I picked (made some conservative assumptions about overhang because I forgot you'd listed what it would be) the sag is "acceptable" (7/100 inch total) even without an apron and that is with a 30lb centre load which is probably much more than the table will experience in normal use. Add the aprons and you have nothing to worry about as far as loads go — when I plug in the numbers for a 150lb person sitting in the middle (!) I get a total deflection of 1/10 inch which is nothing really, still "acceptable". – Graphus Jun 9 '17 at 13:24
  • To clarify, I didn't account for the overhang, and I did evenly distributed weight since there almost never be more than a few pounds (< 10) at any given point. The sag came back acceptable. The "half" I was referring to meant that the sag could double and it would still be "acceptable" – Jacques ジャック Jun 9 '17 at 13:52
  • Also, I did 30lbs per foot, if I do 30lbs total, the sag is much less. – Jacques ジャック Jun 9 '17 at 13:55

The short answer is "you are probably good to go"

The more elaborate answer is "it depends"

The depends part is based on the likely loading of the table top. An ideal tool for helping with the decision is Sagulator as @Graphus mentioned in his comment.

Before you use Sagulator, realize that it is based on structural engineering principles and guidelines. The allowable sag is determined by aesthetic considerations and even the "borderlne" sag is arrived at by the deflection being in excess of 1/360 of the span. Borderline does not usually mean that the structure is in imminent danger of failing catostrophically (i.e. breaking). When you get to "excessive" sag there is usually danger that the structure will actually fracture.

In computer programming there is an old axiom - GIGO - Garbage In Garbage Out. The same is true of using Sagulator which requires that only if you make all of the data selections accurately and knowledgably will you get a reliable answer.

The first piece of information that is needed is the Shelf Material. This allows Sagulator to determine the material stiffness of the wood by selecting a value for its modulus of elasticity. Two pieces of wood with identical dimensions will vary markedly in their stiffness (resistance to bending) depending on their species.

The next item, shelf Attachment is a tricky one and should not be overlooked. Engineers refers to this as "end conditions" which are either fixed or pinned ends. In steel structures a fixed end is welded and even a connection with many rivets is considered a pinned end. It is very difficult to get an end condition in wood that would be considered fixed (perhaps a well executed dove-tail might fit the bill.) The safest and sanest choice is the "floating" option because it is actually referring to "pinned". The "attached" option actually refers to a fixed end which in wood approaches being mythical. The difference in calculated results will be be in the range of 400% difference in calculated sag (e.g. 0.02 instead of 0.10.) Shelf load is pretty straight forward and it would be best to use a uniform load because then weight of the wood itself can be included easily. Not to include the weight of the shelf is to imply that a long piece of wood supported at its ends with no additional load will not sag.The dimensions of span, depth and thickness should be entered accurately.

Some times people get lazy when entering the Thickness and will enter 1" when the thickness is actually 3/4". This small difference will yield a difference of more than 125% in the expected sag of the shelf. Even using 1" instead of 7/8" would give an error of about 50%.

For the table in question I entered type of wood as sugar maple (which has the same modulus of elasticity as the sapele), floating ends, 72" span, 20" depth and 1" thickness. For loading, I allowed 5.5 pounds per foot for the table top plus another 10 pounds per foot for stuff carried by the top.

The result was a total sag of 0.15" which is quite acceptable.

But wait a minute how about the apron that is actually carrying the load (sort of)

The makers of Sagulator conveniently have an "edge strip" feature which allows for stiffening a shelf. That's exactly what the apron is doing for the table top. I entered sugar maple, 3.5" width, 1" thickness and got a total sag of 0.05" which is even more acceptable.

Having vented on Sagulator, I would like to suggest that there might be a bit of problem with geometric stability in that the end legs are only twenty inches apart (set in 10" from front and rear). I suggest that you move them closer to the edges, say 5" or 6". The cantilever at the ends of 1' should be fine. Make scale drawings to see whether it looks reasonable to you.

The legs themselves should do the job if securely fastened to apron. I assume you have something aesthetically pleasing in mind for that.

| improve this answer | |
  • Just two questions about the measurements you put into the Sagulator. 1) Why did you use sugar maple? Sapele was an available option. 2) Why did you use "Floating" shelf? Doesn't a floating shelf only have one edge with support?. Regarding thickness, it's is an inch thick. Was purchased "rough" from lumber yard, not big box store. – Jacques ジャック Jun 11 '17 at 11:50
  • I used sugar maple because I know it is hard - you specified hard maple and sapele. When Sagulator refers to floating, it is referring to points of support, not the entire length - both ends are supported. Floating or pinned means that the wood is free to flex at the support, whereas a fixed end is totally restricted from flexing at the support. As to thickness, use the actual thickness that you achieve after planing and smoothing, whatever it may be and wherever you might have obtained it. I assumed that was the correct dimension that you gave and made my comments for the benefit of others – Ast Pace Jun 11 '17 at 14:10
  • Excellent mini-treatise on how to use the Sagulator, I think we'll be pointing to this in future again and again! +10 Here, my first assumptions (before I remembered the overhangs had been specified) were based on a longer span and I still got "acceptable" even with some pretty unrealistic loading figures. One thing, I wish you could have included something about tables and the cantilever effect from the projecting ends which makes them differ from most shelves, but naturally we can't assume favourable conditions there to try to skew the results towards a desirable result. – Graphus Jun 12 '17 at 5:44
  • P.S. Did you double up the figure for the edge strip to account for the aprons? – Graphus Jun 12 '17 at 5:47

Upgrade the skirt to 2x4 on the 6-ft length would keep the top quite flat.

| improve this answer | |
  • 3
    I can't remember the last time I saw a hardwood dining table with an apron made from 2x4 material ^_^ – Graphus Jun 9 '17 at 13:30
  • @Graphus if he used treated 2x6s, it's be even more sturdy. :-P – CharlieHorse Jun 12 '17 at 14:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.