I am at the last cuts for a shelf which my wife asked for. The whole thing is made out of cherry. The shelf top is 7/8 inch thick.

The supporting brackets which are mounted to the wall studs are planned to have a pin running perpendicular to the wall which the shelf (having matching dovetail slots) will slide onto up to the wall.

The supporting brackets are not as wide as the top shelf, so the sliding dovetails will be stopped at about 2/3 of the width of the shelf.

So my real question, after all the setup information, is what is a safe thickness for the dovetail slots in the shelf? My gut feeling was about half of the thickness of the total shelf (so 7/16 of an inch or so).

I am sure there is a lot of slack either way but just wanted to make sure, or see if anyone had a good rule of thumb for this sort of thing.

edit: added rough sketch enter image description here

  • @Ecnerwal: I added a rough sketch. The picture auto-rotated the wrong way, I will try to fix it later.
    – jbord39
    Mar 13 '17 at 19:29
  • I don't know that I've ever read a given size for the dovetail for a housing joint in relation to the stock thickness, but I think the general way they're done is to remove as little wood as needed to make the dovetail shape, so as thick as possible could be taken as the perceived advice. Obviously this is one of those things where you do have plenty of slack though since the dovetail isn't under a lot of strain.
    – Graphus
    Mar 13 '17 at 21:22
  • 1) How long is the shelf? 2) How deep is the shelf and bracket and what is the nature/approx. weight to be placed on the shelf?
    – Ashlar
    Mar 14 '17 at 3:10
  • The dove tails will bind, a T-track will glide much better and be invisible under the shelf. Mar 14 '17 at 10:47

Considering the low loads and the dimensions you established there should not be any problem with making the dovetails extend 1/2 the depth of the shelf.

A More Technical Discussion: My statement above is based solely on my subjective experience, so I tried finding some more factual engineering information for evaluating this type of condition in wood. If you look at the connection of the shelf and bracket placing a load on the shelf will result in the back of the shelf (against the wall) wanting to tilt up and the front to tilt down The bracket resists this force by holding the shelf down at the back and up in the front. This means that the shelf wants to pull out of the bracket at the back and the dovetail flanges must resist this. At the front the shelf pushes down on the bracket/dovetail top edge and the dovetail sides do not contribute to the support. In both conditions the strength of the connection is measured by shear strength, which can be loosely considered as the ability of the wood to hold together when faced with adjacent up and down forces. Woodbin.com provides a table for strength for various wood species. For cherry the shear resistance is established as 1700 psi (pounds per square inch of stressed surface. In this case the depth of either the shelf or dovetail at the connection location.) That means that if your shelf is 1" thick and the dovetail reaches upward halfway, the load transfer between the shelf and brackets can resist 850 lbs in shear for every inch of length of the bracket/shelf joint. Clearly failure of that joint is not going to be much of a problem in most woodworking projects.

Added Note: The shear strength from the table is for parallel with the grain which would apply to the dovetail. The shelf grain would run along the length of the shelf and perpendicular to the grain at the brackets. The tables did not give that shear value, but I assume it would be greater still. In addition since the dovetail sides are sloped, even though the shear stress is parallel with the grain, the actual strength of the wood in the joint would be more complicated to calculate. It is my assumption that it would be greater but I have no engineering proof.

  • Thanks. So with that in mind, would a dowel/pin that dropped down into the support (towards the end of the stopped dovetail) be a good idea?
    – jbord39
    Mar 14 '17 at 22:52
  • @jbord39 Not sure what you mean - With What in mind? The shelf - bracket connection you propose is already more than adequate to carry the load transfer. Incidentally, even the dovetail is overkill. A simple dado on the underside of the shelf would also receive your bracket. You could also skip the dado and rest the shelf on the bracket. In that case a dowel forward and back ends between the bracket and shelf would hold it in place.
    – Ashlar
    Mar 15 '17 at 2:51
  • 1
    Thanks. Well I am not gluing the joint so the dado won't work. It's near the entry way so it will need to be dissembled anytime we move furniture or refrigerators. That's also the only reason I did the stopped dovetails (sawing those was a real pain ... ). Just finished assembling it today, thanks for your help! Ended up going 3/8 of an inch into the top shelf with the top dovetail.
    – jbord39
    Mar 15 '17 at 4:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.