Roubo suggestions?

[JD Dolan]

I just got a benchtop made of ash. It's 96" long, 30" wide, and a shade under 2 3/4" thick. (It shipped at 270 pounds). My plan is for bench that's 24" wide and the full 96" long.

Is this sucker thick enough for a Roubo-ish bench, or do I need to think about something with a skirt and/or upper stretchers?

I'm still trying to decide between a relatively straightforward Roubo and Bob Lang's 21st Century Bench, so any suggestion there would also be welcome.

Thanks in advance,
J.D.

[David Keller NC]

"Is this sucker thick enough for a Roubo-ish bench, or do I need to think about something with a skirt and/or upper stretchers?"

I'm not sure about the correct answer to this, and I'm not sure that any other creekers could authoritatively answer it either over the internet. It depends a lot on the grain orientation of the plies, whether the stock was defect-free when it was laminated, etc... You may well get "I have a bench that thick, and it's fine" sort of answers, but again - it depends greatly on the wood species, stock orientation, whether there are any end-grain splices, etc...

So... You should test this yourself (which is easy). Simply block up either end of the slab, joint a board 8' long to use as a straightedge, and measure the deflection when a friend sits on the bench. That'd be sort the absolute worst case - my guess is that if it deflects no more than 1/4" with 180 lbs on it, it will easily keep its shape well enough to plane on. If you'd like to get a little fancier with this test, and have a bench that allows you to work while you make your bench, consider making a couple of sawbenches ala Chris Schwarz. They're obviously too low for a permanent solution, but if you're going to do any stock dimensioning with handsaws, they're extremely handy to have, and if the Japanese can get away with a very low bench for planing, you certainly could for a few weeks until you build the framework for yours.

[george wilson]

JD,I don't see why your top wouldn't be perfectly adequate for a bench top. I assume it is a maple bench top,such as are commonly available. I made my workbench out of 4" beechwood,but have another bench with a 96" X 2 3/4" top. It has a metal working vise mounted on it,and lots of heavy stuff piled on it.It is quite rigid.

[John Schreiber]

The Sagulator says . . . You're good.

With a 400 lb load and the supports at the very ends of the 96" length, you get 0.01" of sag.

Since you will seldom have 400 lbs on your bench and the supports will probably be closer to 60" from each other, you can expect less deflection than that.

[Adam Cherubini]

John,

I don't think you used that program incorrectly. I just ran the numbers by hand. The deflection should be at least 10 times what you reported.

I think you may have clicked on a uniform load. To mimic the force of planing I think you need to look at a point load. Also, I think you reported deflection per running foot. Of interest here is the total or max deflection.

Also, 400lbs is way too much load. You could put 100lbs on a bench. But just for apples to apples, 400 lbs would deflect that bench about 1/8". A 100lb load would be 1/4 of that or a 1/32". Personally, I can't get excitied about either number. I think the bench will be plenty stiff. I'm working on 2by doug fir and I like it fine. And I'm bigger and heavier than most. So I just don't think it's an issue.

Now I used 1.5M for young's modulus (E). Depending how this top was glued up, E could be considerably less. MDF which is essentailly glue with wood filler is 1/3 of that which is why I don't recommend it for use as a workbench. Plywood is about 1M. So same bench, same dimensions the mdf bench would deflection 3x more or 3/8". The plywood bench (that's a thick plywood) would deflect 1.5 times more or 3/16". I think MDF is also heavier than solid wood which woudl exascerbate the problem.

My assumption here is that there is 0 deflection when not in use. I'm not suggesting this is the International Space Station's work bench. So static deflection due to gravity alone needs to be considered. For the wood bench, a 300 pouind top will cause a deflection of 3/32" or so. So the bench will need to be built up, then flattened on the legs.

Running a bench top through a thickness sander or planer doesn't take this into account. (ditto for plywood or mdf which arte flat in the pile at teh lumber yard, but not flat on your bench's legs and not generally flattenable). That's why commercially made benches (or tops) may not be more accurate than if you just did the job yourself.

A reasonable question is whether any of this matters. My feeling is no. But we often talk about this issue imprecisely.

Adam

P.S. 2 easy solutions to using thin tops or springy bench top materials is to:

1) Increase your moment of inertia with an apron: My 2by doug fir bench, made up of 2by construction lumber has a 12" apron which effectively doubles the stiffness of the solid bench (at least on paper- I know the center span is not as stiff.)

2) Add addtional legs. In the deflection equation, length is cubed. A third set of legs decreases deflection by a factor if 8. That said, if you know yoy need bench stiffness, working closer to the legs (old fashioned common sense) works too!

[John Schreiber]

John,

I don't think you used that program incorrectly. . . .

Interesting. I haven't run numbers like that by hand for about 25 years, so I'd like to know if I can generally trust "The Sagulator."

I specified white ash as a representative ash.

Fixed attachment - A bench on trestles isn't really fixed, because the trestles can bend in and out around the stringers, but I reasoned that it seemed more fixed than floating.

400 lbs as a uniform load - I doubt I'll ever put anything heavier than my bench top on my bench top, so I guesstimated that the benchtop itself, plus material on top of it would be the greatest load. I don't think impact loads would be a factor for this purpose.

96" span, 24" deep, 2.75" thick - the span would have to be less than 96" and probably no more than 60", so that gives an extra safety factor.

Based on that the Sagulator says .01" total sag or .002" per foot.

If I put in your figures, as above except for a 100 lb center load, I get the same results. If I change to floating ends, I get .03" total sag. To get .10 of total sag, I've got to put in a 400 lb center load with floating ends.

I think we can safely say that JD's bench is going to be plenty stiff enough, but I'm curious about why I getting different results on the Sagulator than you do by hand.

[Adam Cherubini]

That's right. Simply supported, load in the center. I'm picturing him pushing down on the plane. That equation is PL3/48EI which is simply supported (pinned) ends, load at L/2.

Adam

[JD Dolan]

Thanks for all the input, guys. Roubo it is!


J.D.

[Billy Chambless]

I think it's hard to beat the Keller Empirical Method for determining sag.

[John Schreiber]

I think it's hard to beat the Keller Empirical Method for determining sag.

Just guessing, but some guy named Keller lifts the material in place and measures sag with a ruler?

[David Keller NC]

"Just guessing, but some guy named Keller lifts the material in place and measures sag with a ruler?

Ha! Yeah, that's what I had in mind. Though I'm an engineer and like math (yeah, I know, that's weird), there are so many uncertainties in "exact" calculations that I figured it's just easier to prop the benchtop up on blocks, sit on it, and measure the deflection. Funny thing about calculations - when I was drawing up overhead bridge cranes in the early 1980's, the calculations for failure point, deflection under load, etc... were required, but one then multiplied by 5 to arrive at the correct web height for the main beam. So even engineers are ruled by "fudge factors" (they called it a "safety factor").

[Billy Chambless]

Just guessing, but some guy named Keller lifts the material in place and measures sag with a ruler?

Exactly!

Of course, I only called it the Keller because Dave was the first to mention it in this thread...

[Chris Friesen]

Is this sucker thick enough for a Roubo-ish bench, or do I need to think about something with a skirt and/or upper stretchers?

The essence of a Roubo is the thick top and tree-trunk-like legs, with minimal stretchers. I think your top might look a bit thin, though it would probably work just fine.

I went with a modified knockdown Holzhappfel design. In this design the stretchers are substantial because they provide the racking resistance.

[Chris Friesen]

I don't think you used that program incorrectly. I just ran the numbers by hand. The deflection should be at least 10 times what you reported. ... But just for apples to apples, 400 lbs would deflect that bench about 1/8".

According to the sagulator...white ash, floating attachment (so simple supports), 100lb total load, center load, span of 96", 30" wide, and 2.75" thick gives a total sag of 0.02", or about 3/128".

White ash has a modulus of elasticity of 12000MPa, so plugging that into the formula for a simply supported beam gives a deflection of 0.37" in the center.

I think the sagulator has issues.

EDIT: I double checked everything and the sagulator is correct. The deflection for the above scenario is given by
(W*L^3)/(48*E*I)
Where W=100lbf, L=96in, E = 12000MPa = 1740453 lbf/in^2, I = width * height^3 / 12 = 52 in^4

I think I must have mixed up my unit conversions.

[JD Dolan]

Chris--do you have a photo of your bench?
Thanks.