Bench build, elongating holes for movement

[Tyler Keniston]

So I am going at a Roubo style bench build following (more or less) the benchcrafted split top plans. I haven't actually decided whether to split or not yet, but in thinking about this, I started pondering wood movement a lot.

I can't for the life of me find anywhere in the plans that the spax screws driven through the upper rails into the tops are in elongated holes? Am I missing that or is it just assumed we will do it (or does benchcrafted think it unnecessary?)

Speaking more generally, do you typically elongate (such as an oval) the clearance hole, or do any of you sometimes do a pivoting type of hole (creating almost an hour glass) so the screw actually levers rather than the wood sliding (can wood even slide past a tightened lag or bolt?)

Last question here concerning movement: At what point do you leave clearance in a mortise for tenon movement. I ask for the stretchers and the top mortises both. I have seen recommendations for double tenons when a tenon would be too large otherwise (I think its usually over 4") but I don't see how two tenons coming from the same piece of wood wouldn't move the same distance as one large tenon...? I wanted to make some really large legs, but if they go at 6" or so I want to ensure a tight tenon through all seasonal movements.

Thanks for any thoughts on all this moving
Tyler

[Jim Matthews]

There's no space needed for clearance in the mortises.

Those are glue joints - expansion forces are directed elsewhere.
My split top is held on with only two screws up through each half of the top,
in the space between the stretchers, toward the middle.

I want the top to expand away from the center, so that the center batten can be used as designed.
If you're not satisfied with the grip a few screws offer, bullet shaped dowels are traditional.

Tops like these are massive, and will resist movement due to their high moment of inertia.

Start with few screws, add them only if necessary.

[Chris Hachet]

Good to see someone else is considering a split top bench.

[Brian Holcombe]

I use a pivoting hole in table tops and such since generally I'll find a way to mitigate some of the change, such as pinning in the center and allowing expansion in the outside screw holes (such as Jim mentions). On large tables I've split the top then provided each half with the same treatment.

The benchcrafted plans show an upper stretcher, this locks in the wood which is the reason why they split the top. The top can expand from the outside tenons toward the center. I would prefer a bullet dowel like Jim mentions, but a screw would work as well.

If each side of your top is 10" and you have a 4" wide leg you are dealing with the 6" of top that extend past the tenons shoulder toward the center on each side. 6" of white oak over 5% relative humidity change will move about 1/8" so, to be safe I would plan around 1/4" of slack.

Solid top roubos, from what I have seen, do not have upper stretchers, they use the top as the stretcher so to speak. They also utilize a rather narrow top and simply rely on a slight bit of slack in the joints at the base to give it it's 1/8" of movement over the seasons in my estimation.

[Prashun Patel]

I lost love for Spax screws since I had a brand new one snap on me. I would instead use lag screws.

[David Weaver]

I lost love for Spax screws since I had a brand new one snap on me. I would instead use lag screws.


Ditto that on anything wood bench or table related. Either bolt and nut type bolts or lags.

(when specialty deck or wood screws have broken on me or bent and I've gotten a glimpse at the metal in them, it's never looked that fine.)

[Brian Holcombe]

I definitely agree. I drill and tap for regular bolts for bench related things. Table tops get heavy brass woodscrews but I cut the thread with a steel self-tapping thread first.

I've considering moving to acme threads for some bench parts that get put through the wringer.

[Pat Barry]

For a Roubo style bench isn't it typical that the front edge of the bench is coplanar with the legs to facilitate clamping. To accomplish this you will want to fasten the top securely at the front edge and of the bench with the elonagted screw holes toward the middle. The gap will then absorb the seasonal movement.

[Daniel Rode]

I understood that a true Roubo bench has the legs attached to the top with a sort of half-lap through dovetail / mortise joint. There's no top stretcher and therefore no wood movement issues. The leg angle changes slightly as the top moves, but I've read it's not an issue.

Barring that, I'd want mine to attach firmly at the front and expand and contract toward the middle and back.

[Curt Putnam]

Does use of a laminated top (edge grain showing for the top face) change the calculation at all since the greatest movement will be up and down? You then wind up with a leg tenon that wants to move at right angles to the top. Note too, that on a true Roubo with the dovetail and tenon connection to the top that the leg face is what the leg vise will bear against thereby eliminating the likelihood that the top will be pushed away by clamping force. Could do that on a split top but it seems to be overkill? Maybe a bridle to fit the leg tenon?

I'm thinking of using a spline running the length of the top stretchers to resist movement along the length of the top - the spline sits in a dado across the top. Haven't resolved movement across the short side of the front slab - and won't until I resolve the leg connection. For the back side, seems as though a couple screws would do?

If this post is a hijaack rather than discussion continuation then maybe Prashun will delete it for me?

Thanks, Curt

[Brian Holcombe]

That half dovetail joint that most use is basically a bridle joint.

[Daniel Rode]

I'm guessing here, but I don't think it makes enough of a difference to matter. Yes, the wood moves more one way than another but across 3" or 4" of thickness, it's trivial. The wood is often cut at tangential to the grain with somewhat varying angles. Rift-sawn edges are not really quartersawn even though some may be. Does 1/16" vs 11/128" matter? That's the kind of difference I think would occur (if any).


Does use of a laminated top (edge grain showing for the top face) change the calculation at all since the greatest movement will be up and down? You then wind up with a leg tenon that wants to move at right angles to the top. Note too, that on a true Roubo with the dovetail and tenon connection to the top that the leg face is what the leg vise will bear against thereby eliminating the likelihood that the top will be pushed away by clamping force. Could do that on a split top but it seems to be overkill? Maybe a bridle to fit the leg tenon?

I'm thinking of using a spline running the length of the top stretchers to resist movement along the length of the top - the spline sits in a dado across the top. Haven't resolved movement across the short side of the front slab - and won't until I resolve the leg connection. For the back side, seems as though a couple screws would do?

If this post is a hijaack rather than discussion continuation then maybe Prashun will delete it for me?

Thanks, Curt

[Reinis Kanders]

I have used timberlok screws for years in outside projects (sheds, etc.) they are sort of an upgraded lag bolt. They seem to hold up. I used a couple for my workbench top.
My HD sells them.


Ditto that on anything wood bench or table related. Either bolt and nut type bolts or lags.

(when specialty deck or wood screws have broken on me or bent and I've gotten a glimpse at the metal in them, it's never looked that fine.)

[David Weaver]

Thanks for the heads up. At some point, I'll break my bench too and need to build, but I'll be interested in quick bench and not bench book bench.

I looked them up, and they look like they are a slightly narrower replacement for specific traditional lags. Do they seem as strong (for example, the general use one says it replaces a 3/8" lag)?

[Brian Holcombe]

Thanks for the heads up. At some point, I'll break my bench too and need to build, but I'll be interested in quick bench and not bench book bench.

I looked them up, and they look like they are a slightly narrower replacement for specific traditional lags. Do they seem as strong (for example, the general use one says it replaces a 3/8" lag)?

Why not acme threaded studs?

[Tyler Keniston]

Thanks for all the reply's.

This is the first time I have heard construction lags (i.e. spax, timeberlox, etc.) down talked over traditional lags. I have always understood it that construction lags/screws are a harder steel, therefor thinner gauge, but with sharper and larger threads. Old lags just have a sloppy look compared to construction lags to me. Thick, dull, and with flaky galvanized coatings. But I am no metalurgist so I don't really have a clue.

I do know that the Benchcrafted plans calls for tapped threads and bolts in the vise installation, but for the top to upper rail joining calls for spax. Not sure the specific reasoning to use tapped bolts in one application and spax in another, but that is what's called for.

"There's no space needed for clearance in the mortises.
Those are glue joints - expansion forces are directed elsewhere."

I am not sure I know what you mean Jim. The tenon joining the stretchers to the legs, and the tenons joining legs into the top have grain oriented 90 degrees to their accompanying mortises . So movement in the cheeks of the tenons are at 90 degrees to movement in the mortise walls and could over power a glue joint if sufficient.


The point about tradition roubo's with their 'bridle dovetail-esque' leg to top joint and how the leg vise would push on the leg itself is interesting. I never thought of that. With the Benchcrafted design the top does attach with large mortise tenons though, which should be more than sufficient to stay put (so long as those mortise tenons stay relatively tight, hence the question about tenon movement).

[Pat Barry]

The tenon joining the stretchers to the legs, and the tenons joining legs into the top have grain oriented 90 degrees to their accompanying mortises . So movement in the cheeks of the tenons are at 90 degrees to movement in the mortise walls and could over power a glue joint if sufficient.
This is an interesting point with regard to modern glues. Those glues are not cement, they actually willl creep to some extent. I don't know that I have ever seen it documented just how big a mortice and tenon would have to be for that particulate cross-grain situation to be an issue though. It will depend on wood type, wood fiber orientation - radial/tangential vs longitudinal, moisture content, the degree of temperature and humidity change with season, etc. Not an easy thing to figure out. The beauty is that the tenon is captured in the mortice and cannot really grow relatively speaking because it is restricted. It can however shrink a bit but my feeling is that with as large a glue surface as you would have it should not ever completely fail. I suppose you could always add a couple of cross dowels to lock it all up if you have concerns

[David Weaver]

This is the first time I have heard construction lags (i.e. spax, timeberlox, etc.) down talked over traditional lags. I have always understood it that construction lags/screws are a harder steel, therefor thinner gauge, but with sharper and larger threads. Old lags just have a sloppy look compared to construction lags to me. Thick, dull, and with flaky galvanized coatings. But I am no metalurgist so I don't really have a clue.


I think that has more to do with some of us having used lags, but not any of the new whiz bang thinner hardened stuff. Sometimes additional hardness comes at the cost of toughness, especially in commodity goods, and that's what would concern me. I wouldn't be concerned in something like attaching a top, though, just in a situation where something had racking forces.

[Daniel Rode]

Harder is not necessarily better. A thinner, harder fastener might be ideal for one application but a poor fit for another. For example, nails bend pretty easily. This can be a good thing when paired with wood movement.

[Prashun Patel]

My criticism of Spax was based on a sample set of 1 data point. So, take that for what it's worth. Also, I was speaking of the construction screws, not the lag screws, which are considerably thicker.

[David Weaver]

I get the sense reading about the spax that they are popular because on a paying job, you can drive them without pilot holes.

I'm sure a well made bench out of any of them would be fine.