5

In a Maloof-style low back chair (https://americanart.si.edu/artwork/low-back-side-chair-34536), the back legs attach directly to the seat and to the back. The grain in the seat runs back to front, while the grain in the back runs sideways. It seems like the expansion of the seat would pull apart the back joints. Why doesn’t that happen? Do the legs flex enough to accommodate the pressure?

I am trying a similar approach (not in style, but in that the legs attach directly to the seat) and wondering if I should align the seat/back grain in the same direction. Wood movement calculator for my area (Portugal) says that a beech seat could expand by up to 1cm…

Below is a prototype in cheap fir, without the back installed as I agonize over this issue. (Yes, I know the leg joints are weak, the plan is to reinforce them with wooden or brass pins.)

chair prototype

stanch
  • 159
  • 8
  • Congrats on seeing through to a potential problem before you built your chair. You could align the grain in the back to be the same as the seat so there were no worries about differential movement, but this places the weakest axis of the wood in a position where it might split (where if the grain were horizontal in the back it's maximally strong). There are simple solutions to this problem however, depending on the proposed dimensions of the seat back. – Graphus Jun 25 '23 at 18:57
  • Maloof's chair appears to defy good wood sense, and while there is obviously a conflict there, various factors that Maloof was building for probably mean it's perfectly fine in that context. One can't really compare Maloof's chair with anything else with the same basic plan, because of so many different details. I suspect the 2 most major factors are 1) that ziricote has a lower movement coefficient than many other species and 2) the overall dimensions mean that max movement isn't excessive. Related to 2, what numbers did you input to get a potential expansion of up to 10mm for the seat? – Graphus Jun 25 '23 at 19:04
  • @Graphus thanks! My back would be about 40x10x2 cm. I tried making one with “vertical” grain, but it flexed like cardboard! Part of the problem is that while my actual chair would have panels made of 2 25 cm wide boards, here I am using thin rips of fir. Is the design just better off with “horizontal” grain everywhere? And can I still do anything sensible for this prototype given its seat grain orientation and dimensions? The seat is about 45 cm wide. I plugged it here: https://kmtools.com/pages/wood-movement-calculator. Tbh I have various beech furniture and I doubt it moves that much. – stanch Jun 25 '23 at 22:24
  • @Graphus I know it’s not related to my original question, but since you brought up good wood sense, I have to ask. This is my first chair. Is this leg joint stupid (even when pinned through)? It’s kind of an oversimplification of https://www.woodreview.com.au/how-to/redesigning-maloof1. The reason I went for it instead of an apron is that 1) I don’t like the extra visual thickness an apron would add to a thick-enough-to-be-scooped seat; 3) I find it easier to cut compound-angled notches than to precisely assemble an apron with weird angles. – stanch Jun 25 '23 at 22:33
  • "Tbh I have various beech furniture and I doubt it moves that much." Yes, I suspected that part of the issue here is merely that the movement you think you need to account for isn't so great (1cm does seem quite a lot doesn't it?) but on the other hand one does need to account for a worst-case scenario. I thought I knew which movement calculator you'd have used but I was wrong so I'm glad you included the link! I'm using the one on WoodBin, which doesn't make so many assumptions, allowing one to play with the numbers more... but I do get about the same result as for Indeterminate/Unknown :-/ – Graphus Jun 26 '23 at 04:28
  • So, if we're both right that 1cm is excessive what's going on? Well the first thing is finish makes a big difference, and neither movement calculator factors it in. Even a so-so wood finish limits moisture transfer; and a better one, a film finish, limits it a lot more. So going with an assumption of an improvement of only 25% (and I think you can actually expect better, but again worst case) movement of course goes down to 7.5mm from 10mm. If you could get it down to 5mm I think the worries go away would you agree? And that's possible with careful selection of the seat wood — all quartersawn. – Graphus Jun 26 '23 at 04:35
  • "since you brought up good wood sense, I have to ask. This is my first chair. Is this leg joint stupid (even when pinned through)? " To be honest I do think this won't work well. However, I think it's something that needs to actually be tried in wood to actually see. So prototyping this in softwood is a great way to get at least some idea of how this'll work real-world, so major kudos there — this is what furniture designers do. Good ones at least. There are tons of things that seem like they won't work/won't work well that actually do seem to work adequately in real-world applications. – Graphus Jun 26 '23 at 04:42
  • 1
    Thanks @Graphus for your help and encouragement! I will see how soon this prototype falls apart :) Perhaps in the next version I can try an actual Maloof joint. Instead of cutting an angled notch and tilting a straight leg, I think I would need to make the joint straight and cut the leg at an angle from a larger blank. – stanch Jul 01 '23 at 21:38

1 Answers1

2

The issue here is your calculator.

If you look at how Katz Moses explains his little widget, it assumes your furniture is outside all year round and/or your shop space is completely without climate control.

I'm guessing your beech furniture isnt also going to be beach furniture. A joke haha... so you won't actually see that much movement.

A more accurate method to calculate the movement in your piece is to calculate by hand using the actual conditions of your workspace/where the furniture will live. It's pretty simple in terms of calculation. The only work is looking up all the pertinent climate values. But even that is easy because it's all online very accessibly.

Expandong on that, if you're building in a shop that is climate controlled, as is the room where the piece will live, it is completely overkill to be working off the assumption that the piece will be subject to the full annual temperature and humidity swings of your outdoor climate when in fact it would only be subject to annual swings in the ambient moisture content of the air at a stable temperature. (JKM even notes this... but as a relative footnote for some reason. It's a non-trivial caveat as far as I can tell.)

There can still cause considerable movement though. For instance, I work in a Chicago basement. It's very dry in winter and and very moist in the summer. Even with the temperature held stable at about 64F my wood moves noticeably on wider workpiece. So JKM's "best month to assemble" recommendation is probably salient -- even if his calculation method tends towards over estimation.

YoStephen
  • 167
  • 7
  • 1
    Excellent Answer. – Graphus Jun 29 '23 at 00:39
  • Thanks! My workshop is unheated/uninsulated, so I measure up to 90% humidity there in winter However, summers are pretty dry around here, so I think if I finish in the next couple of months it should match my home environment. – stanch Jul 01 '23 at 21:33
  • 1
    @stanch you're probably in luck then because high winter humidity and low summer humidity means your absolute moisture content is relatively stable. – YoStephen Jul 02 '23 at 06:25
  • It also means your summers are much more comfortable than mine! – FreeMan Jul 03 '23 at 17:37