Thread: Base-to-top joinery versus timber expansion

Originally Posted by genericuid

Whenever we see dining tables being built, the top is fastened to the base using some method which allows the top to expand: "Buttons", elongated screw holes, figure-8 connectors, etc..

All 4 of the workbenches and a few assembly benches I have made use steel or wooden clips to hold the top on, and I was to make another I would probably do the same
The oldest bench is 17 years old and is still functioning just fine.
The main reason I have done this is so the benches can be partially taken apart so it can be more easily moved.
In one case I also did it also because I used different timbers in the base than I used for the top.
On a large work bench I use around 24 clips but even on a small one I would use about 16-18 clips.

In terms of expansion for a fixed base/top construction, if the wood is dry and quarter, or close to quarter, sawn there won't be much expansion/contraction AND if the same wood is used for the base and top, in most cases the amount of you expansion/contraction will be the same or similar enough not to be off concern.
Some timbers may expand/contract more than others and these are mixed with different timbers some problems can arise.

Thanks Bob! I'm pretty surprised that expansion wouldn't be a problem, considering how much you hear about it online. Even when I make little boxes I take special care to keep it in mind.

Do you by any chance have a photo of the clips you use? I'm wondering how the connection would be made to prevent movement during planing.

Similar to what Bob did I used steel brackets and screws so it can come apart to move. The base frame is also knock down. Being split top the movement issue is also reduced. I can get mine in the back of a car.
You can see two of the brackets in that pic.
Regards
John

I have been wondering the same.

Is it something to do with table tops being thin and liable to cupping or doming if rigidly held near the edges, whereas a bench has a very thick and robust top that will not?

I don't think Roubo benches as shown in the original photo have a stretcher binding the legs together at the top. I suspect any movement in the benchtop would be accommodated by the legs flexing along their length down to any bottom stretcher where it probably wouldn't resolve to a significant movement.

Where leg frames are constructed independently of the benchtop with top and bottom rails, like Shaker style benches, I tend to think the heavy tops are positioned onto the frames using tenons or dowels thus allowing movement.

A common solution for bench tops is to have two 50mm planks with a recessed centre 25mm panel fitted into rebates. I have four benches in excess of 35 years old, all with solid tops. My work bench has end frames made from 100mm square legs with two horizontal rails. The top is full width, about 750mm wide, by 50mm thick and bolted to the end frames as well to a central bearer. The difference is that the holes for the bolts in the end frames and bearer are 12mm for 10mm bolts. It also has drawers etc. fitted between the end frames. The hex bolt heads are visible in the top and have never been responsible for an "incident", not pretty, but functional. They could have been recessed and plugs glued in to conceal the bolt beads. Timber movement in the bench top has never been an issue.

I would post a photo but the bench top is always covered, after all it is a horizontal surface in a workshop.

Jim

When I built my workbench, I accommodated for wood movement by employing a split top design.


20210711_172735.jpg














The two top slabs are secured to the legs with dovetail tenons. The section of the top from the legs to the centre are allowed to expand toward the gap in the middle, and the part of the top from the legs to the front and back edge (i.e. the overhangs) are allowed to expand outwards.


20210101_141430.jpg20201231_181622.jpg













The endcap is secured at the dovetailed corners and pinned with dowel to ensure it stays tight against top with only the pins at the end glued and the ones toward the middle have elongated holes (see above) to allow for expansion and contraction.




20210501_162755.jpg
20201129_144917.jpg



















At the other end, again the dovetailed corners are fixed but this time I used barrel nuts and bolt to keep it tight against the top. The holes for the bolts, as well as for the twin screws for the moxon vise, are enlarged to allow for movement of the top,


It's been about a year since I installed the top and so far I haven't noticed any problems due to seasonal expansion and that includes the very wet summer we just had.


Hopefully that gives you some ideas on how to plan for wood movement when it comes to workbenches.



Cheers,
Mike



EDIT: Pic below how I laminated the top around the dovetail tenon of the legs....

20201104_154314.jpg













I also used these "thingamajigs".....

20201201_182649.jpg












.....to secure the tops down towards the middle of the bench. I'll see if I can find a pic with them installed.

20201213_205036.jpg













Here's a photo of the "thingamajigs" installed on the underside of the top.....



20201019_175701.jpg




















....and this pic shows you their mating slots in the rails to give you an idea of their installed locations.

That is a sensational bench Kahoy!!
I like the sliding dovetail approach, makes total sense.

Hi G. KK did make a gorgeous bench. I think the movement in a Roubo style bench is vertical only. If the top absorbs moisture it will only get "thicker" because the grain runs up and down.

Originally Posted by Mountain Ash

Hi G. KK did make a gorgeous bench. I think the movement in a Roubo style bench is vertical only. If the top absorbs moisture it will only get "thicker" because the grain runs up and down.

I might be seeing things wrong, but doesn't the grain run horizontally? I've never seen it configured in the style of an endgrain butcher block.
I would expect all expansion to happen vertically as well as front-to-back; but not left-to-right.

Originally Posted by genericuid

I might be seeing things wrong, but doesn't the grain run horizontally? I've never seen it configured in the style of an endgrain butcher block.
I would expect all expansion to happen vertically as well as front-to-back; but not left-to-right.

What MA is referring to is what results when you "face" laminate boards, i.e. glueing the boards face to face as opposed to edge-laminating boards. When you face-laminate boards you effectively get a big quartersawn slab in which case most of the movement would result in the slab becoming thicker or thinner as opposed to becoming wider or narrower. Face-laminated boards are also less prone to cupping or twisting. This is more likely to occur with edge-jointed boards which is why it's important to pay attention to the orientation of the growth rings and alternate the boards accordingly. See below.


download.jpeg solid-wood-panel-component-orientation-2.png

Edge jointed boards like these (most table tops) will expand and contract along its width.

portable_moravian_workbench_WID8085.jpeg














Here's a pic of a typical laminated Roubo top. Notice the orientation of the growth rings.

Tangential expansion is greatest. In a bench this will often mean the top gets thicker, as posted by others above.

But radial expansion is around half tangential expansion (does vary substantially between species)

Longitudinal movement is essentially zero.

Even if radial expansion is only 4%, that is 36mm on a 900mm wide bench.