Chippendale (Newport RI) Desk Legs

[RichCournoyer]

So the guy who hates wood is making an heirloom circa 1800 modified (one-sided) Partner Desk (John Goddard Designer). I’ve had many people asking how do I do the 4-sided, 3-tiled programming/machining on the legs, so here are (I hope) all your answers.

Timeframe:
Research 2 months (Complete)
Leg Model/test: 2 months (Complete)
Leg Machining/final: 3-4 weeks
Remaining desk machining: 6-9 months
Cost: Wood/hardware: $1000 (est)

Programing/Machining. How: In Fusion I create 4 different Setups (on each Square Block/ centered on the dowel hole) (the middle block will be removed during the 3rd tile move, and the end blocks manually removed via a miter saw).

While not a difficult project, it’s also not a for the novice. Strict bookkeeping must be maintained due to 32 difficult programs. A good program loaded in the wrong position will scrap the leg. I still need to write the various programs for the 6 Mortices on the leg top. The other photos should answer most of your questions. Let me know if I’ve left anything out.

.

.

The Setup. The cap screw head is being used as a locating dowel pin, and the 123 block (in the back) aligns the leg 90º to my table.
.

.
Finish leg (as modeled in Fusion (Missing the 6 Mortices, which I will design and program while the legs are machining)

Material: African Mahogany.
Cost: $50 per leg (No I didn’t buy any extra material!)

Six Mortices

Test Leg (Pine/Fir) (Without the upper portion)

.

Philly Inspired Partner Desk (There are NO photos of a Newport inspired Desk) Differences are subtle.

[Griff]

I’ve been contemplating doing something like this…as soon as I learn how to model a leg from scratch. Thanks for the description and photos.

Looking forward to updates as the project progresses.

[fiero1]

Very nice work around/work flow. Looking forward to the finished product.

[RichCournoyer]

They ARE Done!
Here are the numbers
37 days (26 actual machining days), 51 individual programs (x) 4 legs = 204 program runs, 127mb of program data (largest program was 8mb, smallest 3kb), 48 Setups… and NO Shapeoko or Program Errors = Miracle!

PS1: I still need to program and machine the six mortise slots in each upper leg post.

PS2: The 1.0" x 1.0" notch on the upper leg will receive a fluted 1/4 column (I really could use a 4th axis to make these but will make do with my 3 axis). You can see a sample of the 1/4 column above (finished desk photo).

Lots of sanding and hand carving to do and many coats of homemade shellac.

Now on to the other 64 parts…

Note I made: 1Right Front, 1 Left Front, 1 Right Back, and 1 Left Back.

[RichCournoyer]

Turned and milled Quarter Columns (they get inserted into the legs) fresh off my CNC Shapeoko Lathe:

Richard

[RichCournoyer]

Desk update:

[gdon_2003]

Very nice work. Are you making this for your home, office or a commission?. If this is for your home is it the style of furniture you already have or is this a show piece. On another forum I had detailed a rebuild project of a lathe and some one asked why I did all that work, I answered simply “Why do men climb mountains, because they are there.” I also do segmented turnings and there are sometimes hundreds of pieces you have to keep track of. During glue up a misplaced piece can ruin a project. I appreciate the level of detail and the massive data management you had to have to get this done. Excellent work.

[RichCournoyer]

Home…actually for my wife’s retreat room. If you know me, you’ll know how much I dislike working with wood. This project was to test my metal craftsmanship skills to see how it translated into wood…on a challenging heirloom piece of furniture…and so far, I’m pretty pleased. It will be used as a guide (centerpiece) to decorate the rest of the room.

It is far from finished. I figure about 4 more months of CNC machining…(and a month of sanding and fine-tuning the carvings/edges)…lots and lots of edge (gadroon) molding to make…(almost 40 feet)…The center drawer is next on the long list of things to be made. It has a shit-ton of carvings, along with the two support brackets. It is VERY difficult to get these shapes correctly into CAD…but my goal is to have the Shapeoko machine (about) 95% of the center drawer artwork.

(Photo above) I had just roughly installed the four drawers, and haven’t made any adjustments…I (now) see from the photo that the upper right drawer needs a little tweak to the left about 1/16" and 1/32" up. (Yes those locks actually work)

[luc.onthego]

What did you use as joinery for the drawers and the rest of the desk?

[RichCournoyer]

I use the Shapeoko (of course) and an 8º Dovetail Cutter:

(Questions? Ask away)

[luc.onthego]

Nice! Did you used straight evenly spaced toolpaths for your dovetail cutter like a dovetail rake jig would? Did you use F360 to design your joinery or a something like tailmaker? I was wondering if I could get away with Carbide Create for this but I think it would need to handle open vectors.

[RichCournoyer]

First, let me explain…since I’m a trained Tool and Diemaker…and an expert CADCAM guy…I approach my wood projects like a metal project. I know my grandpa (Dziadzu) is shaking his fist at me (he was a cabinet maker/home builder), but it works for me.
Having NEVER cut a dovetail in my life…I (my brain) needed to see it modeled in Fusion. I needed to see how it looked, how it fit, clearances for the cutter, etc in CAD. Now having cut these drawers…look at the two attached photos. I first rough out the center material, then cut the angles with the dovetail. I don’t see any reason these simple cuts couldn’t be cut with Carbide Create.

One word of caution, the dovetail cutter had a tendency to pull out a little (no matter how much I tightened it). I replaced my collet and this fixed that problem

[WillAdams]

Making dovetails using Carbide Create was discussed here:

Sharing Carbide Create Dovetail Files

Actually, it’s been a long time since geometry in high school. Instead, it’s more straight-forward to just draw this up — we start with the endmill itself since it and the stock dimensions as discussed above are the fundamental constraints. [image] Since the dovetail bit we have selected can only cut to a depth of 1/2", we have to leave 1/4" uncut for the blind dovetails, and we have to determine if we will draw up the pins or tails first. Folks do it both ways, and there are numerous argumen…

CC400 and later seems to handle open vectors well enough to allow it to work.