Shopsmith Forums

Poking around on line to see what others might think about drawer bottom thickness, I came on this thread from 12 years ago:

viewtopic.php?t=12276

as well as an interesting video showing some destructive testing of small drawers:

https://www.youtube.com/watch?v=co3EH8WZsX8

Dennis,

Take a look at some of Ramon Valdez' videos, https://www.youtube.com/@ramonartful/videos. He has a lot of great tips here and there from his long professional experience.

twistsol wrote: ↑Wed Feb 05, 2025 12:30 pm Thank you, the drawer fronts are single poplar boards that span three drawers horizontally each. I finished the cabinets with Varathane Cognac stain and sprayed three coats of satin poly. There is a bit of trickery here as well 24 handles on 20 drawers.

As you've shown, poplar can look great when it's nicely stained and finished. All of the doors and most of the wood trim in my house are wood-stained poplar, and it looks a lot more expensive than it is.

As for your extra handles and faux drawer-face gaps, I plan to do exactly the same on my CNC base. In my case, it's to make the one double-height drawer look like the two adjacent single-height drawers.

One day is very impressive for making all those dovetail drawer boxes! I strongly suspect that would take me more like a week. But I do plan to add a vertical workholding station to my CNC, which has the potential to make that go a lot faster. Of course, it also has the potential to turn in to a enormous time sucker.

roy_okc wrote: ↑Wed Feb 05, 2025 2:06 pm Dennis,

Take a look at some of Ramon Valdez' videos, https://www.youtube.com/@ramonartful/videos. He has a lot of great tips here and there from his long professional experience.

Thanks Roy, I see that he has a couple devoted to making strong drawers. I'll be sure to watch those before I finalize my drawer design.

It occurred to me last night that I could use my CNC as an apparatus for load testing it's own drawer design, and that went quickly and easily today. I'm basically just using the Z axis of the machine as a press, very much like Matthias Wandel did with his apparatus in the video that Ed linked above. But Matthias was pressing on his miniature "drawer" bottoms with a flat pad that was almost as large as the bottom. So he was really testing ultimate strength, not drawer-bottom deflection. My own setup applies all the load to a small patch in the center of full-size test specimens, to simulate deflection under worst-case loading conditions.

To simulate the perimeter support of the drawer bottom by the drawer frame, I simply laid four 1/4" MDF spacers on my spoilboard, with their edges just inside the 23-7/8" square perimeter of my 1/4" Baltic Birch test specimen. (That specimen measures considerably closer to 1/4" than to 6 mm thick.) My 12 mm specimen was slightly longer, and I simply let the extra length overhang the inside supporting edges of the side spacers, as that shouldn't affect the test results materially.


For scale, that's a 4" fixturing grid on the spoilboard. Below is a shot with the 12 mm specimen in place atop the spacers.


Next, I placed a small pressure pad on the center of the test specimen. The blue rod in the photo below is a 24" long piece of T-track that I placed on its side to use as a compact straightedge. The pressure pad is taller than the T-track, so the track is merely lying on the test specimen, even under maximum test load.


An ordinary digital bathroom scale served as the instrumentation for these experiments. To deal with its rather large footprint, I stacked a piece of 18 mm BB plywood on top of the pressure pad, to serve as a base for the scale. Then I placed another scrap piece of 18 mm plywood on top of the scale, to spread the force from the CNC spindle so as to not break the glass top of the scale.


To apply the test loads, I simply lowered the spindle until its collet nut pressed on that top pad with the desired force. I tested both specimens at 50 lb. and 100 lb. Below is a photo that shows the deflection of the 18 mm specimen under a 100 lb. load.


To quantify the deflection, I simply measured the gap between the top of the test specimen and the bottom of the straightedge, using setup blocks and/or feeler gauges.

And the deflection results are:

• 12 mm plywood @ 50 lb. load: 0.045"

• 12 mm plywood @ 100 lb. load: 0.087"

• 1/4" plywood @ 50 lb. load: 0.156"

• 1/4" plywood @ 100 lb. load: > 0.25" (my spacers were only 1/4" tall, and the specimen bottomed out on my spoilboard)

Given my shallow drawers and the need for close drawer spacing, I don't want the drawer bottoms to sag by more than 1/8" under full load. The 1/4" plywood fails that standard at even a 50 lb. load, whereas the 1/2" plywood has a good safety margin with a 100 lb. load.

So it appears that you guys have been giving me sound advice!

Nice to see science, opinion, and experience in agreement.

Interesting "fixture grid"!

Anyone surprised by the 'result'?

Not surprised by the result, but impressed by the testing apparatus. Very slick set-up!

I'm not sure if you are using the term 'spoilboard' correctly. That's a masterpiece occupying the location of a spoilboard. I guess those 'tiles' are individually replaceable.(?) Did your CNC cut all of the WoodAnchor grid?

When you get to that point in your project, I think it would be interesting to compare how much the drawer frame reduces drawer bottom deflection.

Looking at your original two proposed designs, I'm wondering why not 12mm BB plywood for the HD drawer frames.(?)

- David

DLB wrote: ↑Wed Feb 05, 2025 9:29 pm I'm not sure if you are using the term 'spoilboard' correctly. That's a masterpiece occupying the location of a spoilboard. I guess those 'tiles' are individually replaceable.(?) Did your CNC cut all of the WoodAnchor grid?

When you get to that point in your project, I think it would be interesting to compare how much the drawer frame reduces drawer bottom deflection.

Looking at your original two proposed designs, I'm wondering why not 12mm BB plywood for the HD drawer frames.(?)

- David

Thanks David — the spoilboard has been working out great. It’s a composite construction, with an 18 mm Baltic Birch bottom layer and a 1/2” MDF sacrificial top layer. The load-bearing portion of the slots is milled all the way into the plywood, for strength, and you can surface up to 1/4” off of the MDF before you need to renew it.

Yes, all the spoilboard machining was done by the CNC itself. In addition to the fixturing slots, it has a grid of 5/16” smooth-bore holes on a 2” spacing, into which you can insert precision dowel pins for very accurately positioning workpieces on the spoilboard. The larger holes are merely counterbores for the screws that fasten the spoilboard to the torsion box.

I suppose that you could replace the “tiles” individually, if you were motivated — at least on the MDF layer. But the design intent is to simply glue a new sheet of 1/4” MDF atop the old sacrificial layer, when the time comes to renew it. Then re-machine the fixturing slots and fixturing-pin holes, resurface the top, and it should be as good as new. That part is all done by the CNC, of course. To make the glue-up easier, you can tile the new top layer from smaller pieces of MDF — just place the edges over the fixturing slots, where they’ll be remachined in any case.

As for using 12 mm BB for the drawer frames, I’ve been wondering the same thing. After cutting the drawer bottoms from full sheets of plywood, I could still make all the frame components from the cutoffs. Does anyone know of any reason not to do so?