26
Stumbled on a stat that 30% of custom cabinet jobs fail because of wood movement, and I'm debating if we all over-engineer for it or not enough.
Saw that number in a Fine Woodworking survey from last month and now I'm questioning if my standard 1/8" gaps are too tight or if others are seeing more callbacks than they admit.
3 comments
Log in to join the discussion
Log In3 Comments
jamesm4818d ago
Back up for a second, that 30% number might be a bit misleading. I think that stat includes every kind of failure from cracked panels to doors that won't close, not just simple wood movement. A lot of those callbacks are probably from folks skipping basic stuff like accounting for seasonal humidity changes in their region. So you might be fine with 1/8 gaps in a stable shop, but it could be tighter than you want if you're shipping installs somewhere with wild swings in moisture. What kind of climate are your builds going into most of the time?
4
taylor.amy3d ago
Hey, have you looked into using a hygrometer and tracking the actual humidity in your shop for a few months? That's what I ended up doing after a similar headache. I was aiming for 1/8 inch gaps religiously, but after logging the wet and dry seasons here, I realized my shop swings about 15% relative humidity year round, so I adjusted my gaps to account for the worst-case scenario. Now I just design around whatever the highest moisture month looks like, and I stopped getting those "the door drags in July" calls. It took the guesswork out completely, and honestly it's way simpler than overthinking material mixing or fancy joinery tricks.
7
kimr1018d ago
Maybe it's just me but I wonder if part of that 30% is from folks using materials that don't move well together, like mixing solid wood panels with plywood frames that expand differently. A lot of the over-engineering focus is on gaps and hardware, but matching the expansion rates of your components might matter more than any single number. Slowing down in the design phase to check how each part's movement lines up could cut down callbacks without adding any extra work.
2