how to not overengineer woodworking projects
September 15, 2018 9:07 AM   Subscribe

We've been doing woodworking projects lately, but we're relative newbs with no idea how to decide how much load any given structure can hold. So we tend toward overengineering. Are there some standard calculations or rules of thumb people use when deciding what size wood to use?

This is currently for shelving, boxes, etc. with predictable weights. We are not to the "make fancy joints" stage yet.
posted by aniola to Home & Garden (10 answers total) 7 users marked this as a favorite
Are you concerned more by the potential of over engineering or under-engineering? Is your goal to build with materials as small as possible to avoid failure? I completely sympathize, as I would rather have to look at a thick shelf board than suffer a bunch of broken pottery due to underbuilding.
posted by klausman at 9:33 AM on September 15, 2018

I am more concerned with over-engineering. I biked through some amazing preserved and rescued forests recently as well as some utterly destroyed "managed" forests. I don't want to waste wood unnecessarily (but I also want what I build to last).
posted by aniola at 9:37 AM on September 15, 2018

Nomenclature check: you are under-engineering; and as a result you are over-building.

Overengineering would be spending weeks developing plans to get a table to weigh under 10lbs, when a 20 lb table would do just fine for its application.

Underengineering is when you avoid doin math and thinking and planning by using thicker heavier stronger materials and adding unnecessary bracing.
posted by SaltySalticid at 9:42 AM on September 15, 2018 [8 favorites]

It really depends on what you're building, what kind of wood you're using, and what your construction technique is.

1) Intuition. You'll get a sense for it!


If you're creating shelving, then there are a lot of resources:
The Sagulator
Shelving span tables
Stronger shelves


If you want a more general approach, you can look at deflection tables. Take a look at table 3 in this plywood deflection table. When the ply thickness is 3/4", and the span is 24", the L/360 (the load at which the plywood will deflect more than the span length divided by 360) is 154 psf, or pounds per square foot. L/360 is often a good rule of thumb for the maximum allowable deflection.

So if you had a 2'x3' piece of 3/4" plywood spanning 24" between supports, then you'd have 154psf * 6 sqft, or 924 lbs uniform load that the plywood could hold before it started to deflect more than L/360, or 1/15th of an inch.

You can use this approach to do a rule-of-thumb engineering process. If you're doing anything risky or where peoples' lives are in the line, then overengineer / add a safety factor / consult a real engineer!
posted by suedehead at 9:52 AM on September 15, 2018 [8 favorites]

I tend to guesstimate rather than calculate, but I've made great strides in getting my woodwork less bulky and heavy. I don't think it takes too long to start to get a feel for what thickness of plywood is right for a job, or where something might need a bit of bracing. I'm a big fan of using torsion boxes in preference to massive slabs of wood, for example. Also, learn to use wood in a way that maximises its dimensional strength - a relatively thin, bendy strip of wood may be incredibly stiff if you turn it on its side, for example. Notice how a cheap Ikea cupboard starts off flimsy, but suddenly becomes much stronger as soon as you tack that thin piece of board onto the back.
posted by pipeski at 11:00 AM on September 15, 2018 [2 favorites]

Making good joints will go a long way toward giving stability with less wood. Please allow me to recommend getting a biscuit joiner. It's pretty simple to use and the joints can be fantastic and very strong. It's also a small, easy-to-store tool that doesn't make you do a lot of ancillary work (it sprays some sawdust, and you do have to buy the little biscuits that go with it, but they're easy to get).

Basically, it cuts some special grooves in opposing pieces of wood (that you want to join) for you, then you glue the biscuits in the grooves (and put glue on the rest of the joined faces).

A huge bonus is that you can edge-join boards and make, say, a gorgeous long-lived desktop or even a door (I haven't tried making a door yet). You can also make invisible joints.

However, the reason I got it is that it makes strong joints.

Most joiners let you use three different sized biscuits/joints, too.
posted by amtho at 11:57 AM on September 15, 2018 [2 favorites]

Over engineering as a term is also very commonly used to mean that an engineer has looked at it, not had time to iterate what they are doing down to the minimal possible solution (or not had the wherewithal) and everyone has been lumped with twice the structure they need. Or an internal structure rated for wind load (true story). It really does mean "an engineer has looked at this and now as a result we're stuck building this ridiculous thing." That out of the way:

I work with both furniture and large engineered structures. In the latter I work with engineers and data such as the aforementioned deflection tables (and the related load tables and span tables) and it's all very certain.

I've never used any such thing in furniture land, it's all experience and a general sense of what looks good combined with a sense of what will actually work. When you are starting out experiment. Try using 6mm ply instead of 9 and see what happens. Look at what more experienced people are doing and mimic. I'd say 90% of furniture is overbuilt, you don't need to be able to stand in the middle of a dining table but it would be rare to find one on which you couldn't.

It is reasonably rare that you can hurt too much if your furniture bends more than it should, the same can't be said for a house.
posted by deadwax at 2:59 PM on September 15, 2018 [3 favorites]

To make it easy, when you want to build an X, look at simply designed examples of X that are sold in stores -- keeping, of course, to products that are designed to last more than an undergraduate degree. Follow those examples for your own design.

Like, if you want to build a desk. Tour IKEA and see how they build their desks, and what materials they use. Consider that it's okay to design "up" from particle board to plywood or board lumber, as long as you keep the thickness comparable.

That's about it. For every X you could want to build in just about any design you could imagine, there will be a similar X for sale and designed by people who know what they're doing. Follow their lead and you won't go wrong.

(It's called standing on the toes of giants, it's the way almost everything is designed anyhow, it's all good!)
posted by seanmpuckett at 3:05 PM on September 15, 2018

you don't need to be able to stand in the middle of a dining table

It’s important to make furniture as strong as typical furniture items, because furniture sometimes gets used in unusual ways. Perhaps someone will stand in the middle of a dining table to change a light bulb. Perhaps an eight year old will try climbing up a bookcase.
posted by monotreme at 4:14 PM on September 15, 2018 [4 favorites]

Oooh, I do this. There's also a website, let me find it.

Found it. The Sagulator. Does this answer your question?

And, on scrolling up, it seems I've been scooped.

Anyway, it certainly beats doing load, shear, bending diagrams instead of woodchips or shavings.
posted by RolandOfEld at 8:20 PM on September 15, 2018 [3 favorites]

« Older Recommend me a bass amp for noodling around at...   |   Non-Racist Halloween Costumes Newer »
This thread is closed to new comments.