Getting one-off CNC parts made

[Tom Henderson2]

Greetings-

I hope this is the right place for a "how do I get a CNC part made" kind of question...

I'd like to make a workbench top made with 20mm holes spaced at ~4" centers -- similar to the Festool "multifunction table" setup.
https://www.festoolusa.com/power-tools/work-tables/mft3-multifunction-table-basic-495888


To make the most of this system, the holes need to be very precisely located -- something that I would have trouble doing by hand.

Is it feasible to have a local CNC shop do this for me? The part will be ~3' x 6', 3/4" baltic birch plywood.

I realize there will be setup and design costs but I have no idea what the total cost will be for a 1-off project like this.

Any suggestions about how to find a shop that would be amenable to a 1-off design like this?

Any ROM guesses for the cost to do this? $100? $500? $1000? I'm completely clueless.

Any thoughts/suggestions are welcome. And if anybody is familiar with a shop between Santa Barbara and the 450 freeway that might be friendly to a project like this, please let me know.

Thanks in advance!

-Tom in Ventura

[Ross Moshinsky]

Material is $60+ markup. Machine time is probably 1 hour. Design time is probably 1/2 hour. These time estimates are on the higher side. Design is really about 5 minutes and machine time will vary between say 15-45min depending, but most people would round up since it's a one off.

Ball park would be between $100-250 I'd guess.

[Bruce Page]

I'm not sure what you mean by local CNC shop, most machine shops would not be interested, IMO. Someone with a CNC router would be the way to go. You could look in your local area for a sign shop, they usually will have a large CNC router or they could point you in the right direction. You could also ask your question on a CNC forum like Camheads or ShopBot and see if someone would be interested - you might even find someone close by.
I would guesstimate the cost at the high end of Ross's numbers, maybe a little higher. I did a quick sketch - on 4" centers your looking at 162 holes. when you factor in the material, number of holes, and 1 or 2 carbide cutters, I think $250 would be a minimum.
I'm attaching the sketch and drawing file.

[Nicolas Silva]

check in at cnczone.com
there is a section for work requests (RFQs) and employment opportunities.
Cheers!

[Jerome Stanek]

Check out 100k garages or you could post on the lets talk shopbot forum

[Phil Thien]

I agree w/ posting at CNCZONE or a ShopBot forum, post a drawing to get the maximum number of responses. Be clear on tolerances on those holes, so you don't get 162 holes that won't work.

Shops with larger CNC routers typically are businesses that want to make a return on their investment plus pay one to multiple living wages. So the one-off price may be high.

So consider getting quoted for 1, 5, 10, and maybe 25. Now advertise them on your local craigslist and see if there is local interest. If you find a bunch of other guys that would like one, you can get the per-unit cost down to somewhere everyone is happy. So the router shop can make some money and be happy, and a bunch of woodworkers can get a deal on an MFT-compatible table.

[Tom Henderson2]

I agree w/ posting at CNCZONE or a ShopBot forum, post a drawing to get the maximum number of responses. Be clear on tolerances on those holes, so you don't get 162 holes that won't work.

Shops with larger CNC routers typically are businesses that want to make a return on their investment plus pay one to multiple living wages. So the one-off price may be high.

So consider getting quoted for 1, 5, 10, and maybe 25. Now advertise them on your local craigslist and see if there is local interest. If you find a bunch of other guys that would like one, you can get the per-unit cost down to somewhere everyone is happy. So the router shop can make some money and be happy, and a bunch of woodworkers can get a deal on an MFT-compatible table.

Thanks for your thoughts Phil -- all good points.

[Ross Moshinsky]

A big production shop is going to price this well above market value because they don't want the work. You need to find someone who doesn't mind doing a one off job which is going to be guys in their garage or working out of small shops.

Just for reference, I threw Bruce's drawing in a machining simulation and it took about 30 minutes with a 1/2" end mill and 175ipm and 2 passes.

[Phil Thien]

A big production shop is going to price this well above market value because they don't want the work. You need to find someone who doesn't mind doing a one off job which is going to be guys in their garage or working out of small shops.

Just for reference, I threw Bruce's drawing in a machining simulation and it took about 30 minutes with a 1/2" end mill and 175ipm and 2 passes.

I had guessed ten minutes.

The cutting path for each hole is less than an inch, maybe 150 inches for all the holes, 300-inches for two passes. The drawing is now gone but the perimeter of a 4x8 sheet of plywood is 288-inches, double that for two passes, so 576-inches. Now, movement between the cuts will be done at the rapid speeds but even if we factor the same 175-ipm, you can just add the perimeter of plywood (288) again to the #'s above for a total of (roughly) 1164 inches. So that would be 6.64 minutes not including plunging.

Why is my math so far off your simulator?

I've got to be doing something wrong because if you figure five seconds per hole (which seems fast, doesn't it?) then 162 holes is 13.5 minutes.

I can tell you one thing: If you want to make money, don't use Phil Thien for your estimating. :)

[Ross Moshinsky]

CNC machines do not run at a constant speed. On straight cuts, you're going to see it hit the max speed more often than not but on lots of cuts on curves, it simply cannot run at the max speed all the time. It's kind of more a max speed/guideline speed. I believe I came up with around 3 minutes for the perimeter cut. The perimeter is 216in / 175ipm ~ 1.25min x 2 ~ 2.5min. When you figure in a bit of jogging, ramping, slowing for the turns, ect it works out. On circles, that math won't work.

[Phil Thien]

CNC machines do not run at a constant speed. On straight cuts, you're going to see it hit the max speed more often than not but on lots of cuts on curves, it simply cannot run at the max speed all the time. It's kind of more a max speed/guideline speed. I believe I came up with around 3 minutes for the perimeter cut. The perimeter is 216in / 175ipm ~ 1.25min x 2 ~ 2.5min. When you figure in a bit of jogging, ramping, slowing for the turns, ect it works out. On circles, that math won't work.

Of course! And here I am having recently played with the CV (Constant Velocity) settings on my little desktop router, I should have known.

Thanks for the explanation.

[Art Mann]

It is very hard to estimate cut time by hand because of acceleration/deceleration times and the slowdown cutting arcs. That 175 IPM speed that Ross simulated will hardly ever be even close to the actual average cut rate unless you are just routing straight lines. These parameters are highly machine dependent. It took a lot of fiddling with "Rapids Rate" and "Scale Factor" parameters to get to the point that my copy of Vcarve Pro tweaked for my machine estimates run time within even 10%.

Oops. I see Ross beat me to it.