Flattening Roubo workbench top vs. front side

[John Schroeder]

I'm how in the home stretch with my roubo build. I've got it assembled on its side and just flattened the front legs and side so they are all in a single plane. The top is pretty flat pre-assembly, but unfortunately it is not at a perfect 90 degree angle to the side now, although it's fairly close. Not good enough for furniture joinery, but I'm at a loss as to whether I should worry about it. I can certainly fix this when i flip it right side up, but my question is how much does it really matter? If the front side with the leg vise is flat and coplanar, and the top is flat, how important is it to have the front and top at precisely 90 degrees? I've certainly gone the extra mile with the fit and finish, but I just don't have enough experience to know whether from a practical view it will matter enough to worry about. I haven't found much in either Chris Schwarz's books or online, so I'd appreciate any thoughts from folks with more experience. If it needs to be a precise 90 degree angle, any tips on techniques would be great as well.

[Chris Griggs]

Well, I don't think it matters as long as the surfaces are flat - unless they're way way off, or you are somehow going to use them as a right angle reference. But that's just my opinion, and it will be interesting to hear what others say.

I can tell you that the aprons on my Nicholson bench are most certainly not a perfect 90 degrees to the top and it has not seemed to affect anything.

In fact, the aprons aren't even all that flat on my bench - they are flattish... - the reason the top needs to be flat is so that you can flatten a board and not have the out flatness of the top be reflected in the board you are trying to flatten. To me as long as the front is flat enough to have a piece be clamped tightly against it, that is flat enough.

[Mike Siemsen]

Depending on how you built your bench the angle between top and legs could change with seasonal movement anyway. I am sure it is fine.

[john brenton]

It matters when you go to plane the dovetails on your drawer sides. It'll never be right if the joint isn't held square. I struggled with this on my 8' bench and shaved off a bushel of wood to get it right. Its still not perfect, but I can't tell that its not perfect....I just know that it isn't because I did it!

[Derek Cohen]

Hi John

It depends on whether the bench-leg intersection is going to be your reference for joints, such as dovetails. If this is the case, and you want to square them, then simply shimming/tuning the tenons (legs into top) should sort that out. You do not need to plane the top to the appropriate angle (you do need to plane it flat in all directions).

Since the advent of the Moxon vise for dovetailing I think that this intersection is less important. My reference is the vise on top of the bench, not the side of the bench.

Regards from Perth

Derek

[john brenton]

I say a bushel, and although it was a lot (and it was hell in 100* weather with 5,000% humidity), it wasn't that bad and was worth it to have a decent reference.

I'm no stranger to gettin' by and definitely no stranger to half-a**ed work. Shimming is a fix, but why not just have it square? I've grown fond of being able to clamp one board in the twin screw vise with the other resting square on the top and knowing it's a fairly accurate 90* angle. We don't know how out of square he's talking here. If it was anything like mine was, clamping the pieces in and planing would have left a mess.

[Tony Shea]

It depends on how far out of square they are. As others have said, it is going to change with seasonal movement, especially depending on what % moisture the lumber for the bench started out at. Even in the Schwarz's book I think he says something about using the high moisture lumber for the legs and apron so as the pull tighter once they become dry. I believe there is a picture in his original book with him holding a square to show how out of square his legs/apron is to the top. Having said all that, I would certainly strive for perfect right from the start if it was my bench. If you have gone the extra mile for everything else it may bother you knowing that this part of the bench is not square, it would me. So it boils down to YOU.

[Chris Fournier]

One, the "Roubo" benches that are so popular right now are inherently flawed in that the bases movement messes with the tops stability. Perhaps this wasn't in the magazine article but it would be obvious to folks that have taken the time to understand our medium - wood. Understanding wood is rarely a topic of today's magazines and the "quick fix" media. This is too bad as the current crop of woodworkers are worse off for it.

Two, this is your bench. If you can't see your way to seeing this piece square and flat, I think that the rest of your work may be doomed to the same compromises and dithering. Making things "right" is hard work especially when we do a poor job of it at the start (I have done plenty of this!). I suggest that you make your bench as perfectly as you can, once you beat this task down, the others that follow will seem so much easier.

Four components compromise the seasonal stability (a bench top will expand and contract but need not go out of flat on a better designed bench) of the Roubo bench given the design. What are theses components and how does their overall dimension affect this instability? This is a skill testing question to which there is no prize except perhaps a better understanding of designing wooden structures.

[Rob Fisher]

Chris Schwarz does actually talk about how the wood will move. He even shows a photo of how the top shrank in width slightly, enough to slightly A-frame his bench. He uses a framing square to show that the angle is no longer 90 degrees but is now slightly more. That said, I do believe that it would be best to be as close to 90 degrees as possible initially. Doing it right initially is almost always easier than trying to fix it later.

[John Schroeder]

Thanks for the thoughtful replies. I'm definitely going to get the top dead flat, as it is just critical for so many things as a reference point. I'll mark it out to square it 90 degrees to the front side and get a planing workout, but as others have pointed out, it's going to move anyway even though i've left a bit of room in the back mortices. FYI the side and top are definitely off square - not terribly with a small square, but with a 9" down the legs it's obviously not right. Two of the legs are made out of solid 16/4 maple, and even though they acclimated for a year I'm worried they've moved a bit after surfacing relative to the laminated ones. That might also doom me to "slightly off," which while I'll know it's there, I doubt I'll ever notice functionally.

All that aside, after working on a converted desk as a workbench for years, I'm thrilled to have a flat top and side surface with reliable, fast clamping and holding. The 90 degree side is nice, but not critical in my view and just wanted to make sure I wasn't missing anything in my thinking. FWIW my previous workbench/desk creaked, rattled, and slid, and I've managed a few nice pieces in spite of the work surface. So for me the stability of a 300+ pound hard maple tank with draw bored M/T's outweighs any movement of the top against the legs, which given the quartersawn grain orientation should be more up and down than across the width. For dovetails i'm going to use a moxon vise anyway, so as Derek pointed out the top will be the reference, and the top is too wide to hang drawers over to plane down anyway (personal preference). Thanks again for the input, much appreciated.

[Chris Fournier]

Hey, none of us are woodworking on granite inspection surfaces! My cheap shot at the "Roubo" revolution is to point out that a floating top can move and stay flat, a Roubo design simply cannot which to my way of thinking is a design flaw of the first order. In principle I may have a point, in practice perhaps less so.

[john brenton]

Flat is such a hard thing to get to on a bench. Which side do you take from? The back, the front or the leg? It is discernable, but I know in my case I had a really hard time with it. You have no real reference as you certainly can't rely on your garage floor. All I could do was get the top plane, and I had to trim down the edge and legs. To make it easier I made some reference cuts wide enough for the blade of the straightedge to fit (the old school rosewood squares have thin blades). That helped get it dialed in. In the end, my bench is mostly out of square, but flat and square where it counts.

[Hans Braul]

Four components compromise the seasonal stability (a bench top will expand and contract but need not go out of flat on a better designed bench) of the Roubo bench given the design. What are theses components and how does their overall dimension affect this instability? This is a skill testing question to which there is no prize except perhaps a better understanding of designing wooden structures.

I did that calculation for my Roubo bench. The top is essentially quarter sawn ash, since it is laminated from flat sawn stock. Distance between front and back legs is 11". Seasonal movement of the top between the legs is about .05", resulting in a change in the angle between top and sides of about 0.1 degree. This is woodworking, not precision metal machining. I maintain that this amount of movement is insignificant for its function.

Regards,
Hans

[Jim Foster]

Out of curiosity, how much is it out?

[Chris Fournier]

I did that calculation for my Roubo bench. The top is essentially quarter sawn ash, since it is laminated from flat sawn stock. Distance between front and back legs is 11". Seasonal movement of the top between the legs is about .05", resulting in a change in the angle between top and sides of about 0.1 degree. This is woodworking, not precision metal machining. I maintain that this amount of movement is insignificant for its function.

Regards,
Hans

Your numbers seem to be out of the realm of my experience Hans. What were your starting and finishing MC values taht you used when you calculated the tops movement? What is the distance of the top of the bench top to the top of the stretcher on your base?

[John Schroeder]

According to my feeler gauge it's out .030 inches at the end of a 9" square. I took a picture of the leg with two squares on it, you can see the gap/shadow at the end of the longer one. I've now got the side flat, within .002" over a four foot straight edge across the legs and side, so now it's on to the top. If I can get it to 90 I will, but a dead flat top with continuous shavings from my jointer plane is the main objective.

Just for fun I put a starrett protractor on it - it might be a hair off 90, but if you didn't know I'm not sure you could tell. It's also square according to my larger framing square. I think it will be plenty good enough even if I can't get it perfect, which will be annoying but as Hans said it isn't a machine shop. Any At some point I also need to get going on my wife's new chest of drawers, as that was the quid pro quo to building the bench. :) Thanks again for the comments.

208782

[Hans Braul]

Your numbers seem to be out of the realm of my experience Hans. What were your starting and finishing MC values taht you used when you calculated the tops movement? What is the distance of the top of the bench top to the top of the stretcher on your base?

Hi Chris,
I used a wood movement calculator with the following assumptions:
Black Ash
initial moisture content 15% (summer)
final moisture content 5% (winter)
quarter sawn (radial expansion)
Distance between front and back legs 11"
Distance from bottom of bench top to top of stretcher 22"

Just an approximation, but it seems reasonable to me. This says expansion will be .05" over the 11" distance, which is consistent with with my experience, which admittedly is not much. Of course, if the bench is wider, the problem becomes worse, but even if it's twice as wide, it seems like no problem to me. If the bench was built in summer and the angle between top and legs was exactly 90 degrees, then in winter the top will shrink relative to the stretchers and the legs will be pulled in to make <90 degrees. Assuming deflection is the same on both front and back legs, the angular change will be about .05 degrees on each joint (0.1 degree total, divided front and back). Even at twice or three times this amount, are we really talking about anything significant for woodworking? Certainly not the way I work!
Regards
Hans

[Chris Fournier]

Hi Chris,
I used a wood movement calculator with the following assumptions:
Black Ash
initial moisture content 15% (summer)
final moisture content 5% (winter)
quarter sawn (radial expansion)
Distance between front and back legs 11"
Distance from bottom of bench top to top of stretcher 22"

Just an approximation, but it seems reasonable to me. This says expansion will be .05" over the 11" distance, which is consistent with with my experience, which admittedly is not much. Of course, if the bench is wider, the problem becomes worse, but even if it's twice as wide, it seems like no problem to me. If the bench was built in summer and the angle between top and legs was exactly 90 degrees, then in winter the top will shrink relative to the stretchers and the legs will be pulled in to make <90 degrees. Assuming deflection is the same on both front and back legs, the angular change will be about .05 degrees on each joint (0.1 degree total, divided front and back). Even at twice or three times this amount, are we really talking about anything significant for woodworking? Certainly not the way I work!
Regards
Hans

When I get a chance I'll dig out my formula and plug in the numbers you have posted. If anything, 15% is high in my experience! I have a bench that would be a cousin to yours and it moves more like 0.25" + from winter to summer and it's in my house.