[Surveybot]

Hi
I am new to this site. I am based in Geelong vic.
I want to build a CNC machine initially for cutting
marine plywood parts for boat building.
Ideally a full size sheet but half sheet would do.
I want to narrow the field of choices.
Thanks for any help.

[pippin88]

What equipment and skills do you have?

Steel is cheap. If you can weld, you could build a reasonable machine for 4-5k

Look at BST automation (Fred) on Aliexpress for Hiwin linear rails. 15mm rails are enough.
I've had a number of good purchases from BST. Most recently had a VFD front panel die. Fred sent me a new one quite promptly.

[hanermo]

Use size 25 rails.
Look at mechmate for inspiration and pics of real machines suitable for *professional use* making money.

The right machine will be about 1000 kg for a full sheet, and about 600 kg for half sheet.
Many chinese machines will be suitable - use an agent.

For *real use* you maybe need a toolchanger.
Boatbuilding as such can be done with manual toolchanges where / if the work is mostly structural, in frames, longitudinals, cabinetry frames etc.
Aim for about 3 kW spindle.

Where a cnc SHINES is panels, cutouts, custom cabinetry etc. with NO FITTING ie dimensionally exact first time.
You will want to acquire sw for doing auto-fits of pieces ie layout sw meant for routers.

Another part that is potentially very lucrative is dimensions based modelling and cutouts for families of parts.
Again, typical simple sw wont do this.

Your 2 good options are steel or a chinese cast iron (or steel) machine frame.
You wont be able to do anything commercially useful for 4k.
Plan for 8k and up, plus tools (1-3k) plus sw (anything).

[hanermo]

Further to my last post ..
It was meant for professional use, where someone is not getting paid except when producing stuff.

A light (cheap, not rigid) machine with small spindle will work fine, slowly.
Fine for hobby use.
End result is making 10$ / hr, ie losing money. A basic salary burdened cost in australia is 50$.

You comparison is NOT, weather it works.
Your comparison is everyone else, charging 50-80$ / hr, running a 10kW and up spindle at 30.000 - 40.000 rpm, cutting at 700" / minute.

If you cannot get to a similar level, you are much better off only doing the cad work, and getting someone else to cut the pieces for you, at 80$/hr.
So, your minimum usable level is about 200" / minute vs outsourcing the cutting.

If you do 200" / min, your cost is now 80 / 4 = 20$ / hr- your cost, minimum 20$ / hr.
You are not getting paid, you are not breaking even.
You are working for free and losing money.

Servos, toolchanger, 3 kW and up spindle.
Rigid heavy machine.

My recommendation to everyone has been about 15k$ minimum for boatbuilding, cabinetry, or anything meant to make money.
Sorry, but that is how it is.

[ger21]

Use size 25 rails.

20mm rails are more than sufficient.
25mm rails are used on $100,000 machine weighing thousands of pounds.

You comparison is NOT, weather it works.
Your comparison is everyone else, charging 50-80$ / hr, running a 10kW and up spindle at 30.000 - 40.000 rpm, cutting at 700" / minute.

If you cannot get to a similar level, you are much better off only doing the cad work, and getting someone else to cut the pieces for you, at 80$/hr.
So, your minimum usable level is about 200" / minute vs outsourcing the cutting.

If you do 200" / min, your cost is now 80 / 4 = 20$ / hr- your cost, minimum 20$ / hr.
You are not getting paid, you are not breaking even.
You are working for free and losing money.

A lightweight, inexpensive machine can easily make money for a small 1-2 man shop.
For two reasons.
1) Even a slow machine can cut complex parts, especially curved parts, far faster than anyone can manually cut them,
2) While the machine is working, you are working on something else. So the CNC becomes another employee, that works for free.

Now, if you want to do production cutting, where time= money, than yes, you'd want a much better machine. But you wouldn't be doing this in a small shop.
Plenty of one man shops have made a living cutting parts on lightweight machines like ShopBots, and even similar DIY machines like Mechmates.

FWIW, you don't need a high rpm spindle to cut at 700ipm. I cut over 800ipm every day at about 16,000 rpm.


Now, be aware that $4000 buys very little.It's very difficult to build a 4x4 machine for that price, and nearly impossible for a 4x8.
There are also not many plans worth buying for larger machine. You basically have two plan options, imo.
Mechmate, and Joe's 4x4. With the Hybrid, there have been many updates and improvements to the original design, and I think that you need to get all of that information in their forum.
If you can find a kit in your price range, it's probably not worth buying.

[Surveybot]

Hi
I am new to this site. I am based in Geelong vic.
I want to build a CNC machine initially for cutting
marine plywood parts for boat building.
Ideally a full size sheet but half sheet would do.
I want to narrow the field of choices.
Thanks for any help.

Thanks for the info and commercial realities.
When I say boat building I mean repairing my boat projects
and bits and pieces. I do some laser scanning and cad modelling for my small
surveying business and I can see some applications cutting out ply to suite.
I do a bit of hobby welding, cabinetry so steel frame sounds like the go.

What affordable hobby software do you recommend if I want to use my own cad files?
Thanks for your help.

[pippin88]

Fusion360 is by far the best free CAD/CAM. Free for hobbyists and startups / low turn over businesses.

Hanermo likes to dream big. He has never posted a photo of his alleged machines, I look forward to seeing one.

[hanermo]

Unlike most, I used to sell commercial machines, avg price 73.000€, and have factory training in machine building.
I sold many millions of € (60+) in machines, and visited over 200 shops myself.
I know of 19.994 shops, here in Spain.

To reply to ger21..

The OP make a Q. regarding commercial cutting for boatbuilding.
This is very much what I replied to, and I tried to be specific about it.
At least tried to be.

The smallest guiderails used on Haas machines are 35 mm.
The midsize and large machines use 45 mm rails and guides.
The GSR routers use, iirc, 45 mm rails.

The *right* choice is *more rigidity*.
The cheapest, easiest way to do so is to use bigger rails (vs ultra precision rails. Almost no-one here could make or mount them sufficiently well, and the price is very much higher than simply using larger rails).
The right choice, ime, is 25 mm rails for commercial routers, at a very minimum.
Larger rails allows you to have a more sloppy build, while achieving better results, easier, faster.
Carrying capacity of the rails is not the point. Rigidity is the only point.

"Making money" commercially means having a positive net income .. not having "some income".

This is why the comparison is not "can I make any money", the comparison is can I make more money versus someone else cutting the stuff.

For example, I have a Bridgeport M head as spindle, bought when I started way back in 2003.
Now, I am changing it for a new ISO30 spindle, with auto toolchanger.

Business case example.
With an auto toolchanger I can make money, and without, I cannot.
Likewise, I will be doing 20-200 parts-at-once, because this is how I can make money.
(I try to focus on small parts, less than 1 kg finished mass, steel.)
The cycle time is irrelevant, to me, the unattended running is what makes it profitable.

Expected running times will be hours to tens of hours.
The extra costs (about 5k€, latest upgrade) are immaterial, to me, as long as the end result is profitable.

My business case is simple.
Single load - run set of parts, between 1-2 days till next load.
I don´t need high rpm, high power (3 kW at spindle is plenty).
Running costs are essentially zero.

I make pieces with 1-10€ margin/piece, in qty 10-200.
Essentially, every morning goal is another 200-1000€ in gross margin, done.

The auto industry looks at this differently (wrong), and the current flavor of the month is drill-tap machines like Brother.
Very fast machines (very good, as well), making one piece after another.
This leads to very complex, very expensive, load/unload systems, robots, auto doors, apl, etc..
And large numbers of machines, lost of space, power, etc.

The other option is making large numbers of similar/identical pieces in the same machine, and long unattended running times.
I get this by having a relatively huge volume, as the work cube area (1600x1000x500 mm).

Cycle-time advocates ignore market size realities, where you have a finite number of possible pieces to make, before prices start to go down, due to market saturation.
The better pov is optimising for how can I make x pieces as cheaply as possible.

For local industrial carpenters (ie manufacturers of custom wood products) getting into the cnc game, my suggested minimum investment is 15-25.000 €.
I will finance the machine for them, if needed.

For example.
ISO30 toolholders - 24 of.
Collets for same.
End mills, face mill, taps, drills, etc. all tooling.

Toolholder, collet, tool, all in about 130€ each x 24 = 3120€.
Some metrology - 1000-5000 €.
Lifter/stacker, secondhand, 2-5000€.
Material racks - 1-4000€.
(Auto) saw, secondhand, 2-4000€.

Pick, and mix and match from above as-needed.
Finishing stuff as needed.

I prefer not to sell a 8000 € "machine" with no tooling, no sw, no training.
End result is customer not making money- unhappy.

Only possible choice, imho, is to only sell all needed parts, and invest the needed 2-4 weeks in training, if the person is new to cnc business.
End result is customer making parts, profitably.
Machine is paid for in 3-6 months.

Anyone buying a machine for 10.000€ more vs a lightweight "kit" makes many multiples of money, and much higher payback, and much higher monthly profit.
IE machine 10-15.000€, and 5-10k in ancillary equipment, as needed.

For previous industrial customers they had, on average, 6-10 machines, in 250 sq m workshops, often with 1-2 people.
Most machines were old, 10-15 years, and most were manual and not much used any more.

The comparison is industrial routers, like GSR series from Haas, about 70k in the EU (not really sold here, regulatory issues).
But smaller.
A GSR for example would be excellent for boatbuilding - ie making boats as a manufacturer.

Final point.
I expect to get about 40-70 n/um rigidity in the next iteration.
Using 35 mm rails, and 32 mm ballscrews.
Previous version was 16 N/um, iirc.

Deflection was 0.05 mm/80 kgf (I stand on the spindle, about 80 kg. 800 N/50 um => 16N/um).
I never expect to load the frame/machine to the 500.000 N theoretical capacity (12 trucks x 35 mm / 4000 kgf capacity (static load)).

I expect the max load at 500 kgf to deliver only small deflection, thats all, at fairly low cost.

In practice, I hope to only need less than 100 kgf loads (2-3 kW spindle power).
Rails cost about 2000€ (6 on z, 4 on x and y).

[ger21]

The OP make a Q. regarding commercial cutting for boatbuilding.

He never said "commercial cutting".

[hanermo]

Ger.. maybe he did not.

He did say "for boatbuilding" and that is usually commercial cutting.
I tried to be clear my advice applies for commercial use.

If the use is occasional, or as an adjunct to other business, many of the drivers change greatly.

IE an architect, making models, can use a very, very light machine, where the main business is making designs, and the models are only an aide, and productivity of the model is not all that important.
Likewise, often even in industrial use, very vey light cuts are the order of the day.

Some moldmaking is like that, and the cuts are with 1 mm endmills, running for 3-5 days, 24x7.
The mold can be 3 x 0.6 x 0.4 m, 10 tons of steel (auto stuff).

The finest spindles I have held, and hope to buy, some day, are NSK or nakanishi.
40 mm outer D, 200-400W, 60.000 rpm.
Near silent in use.
Cost is an issue, 4-6k$.

Air cooled.

I just try to express my rationale .. which is correct as far as it goes, for it´s intended setting.
No disrespect is intended, just helping out.

[Surveybot]

Thanks for your help guys.
I did not mean to start a fight.
Cheers

[quinnkyle]

There are some really cool templates you can cut directly from a wood CNC, using 1/4" plywood. Basically, each section has 1/4" holes every 10" or so. You zip tie the pieces together, using ordinary plastic zip ties. When all sections are together, it's sort of a floppy boat design, but once fiberglass paint is applied liberally to both sides, you end up with a small boat that still looks like wood! Pretty amazing. My husband build one and used it for fishing each year - it requires annual fiberglass touch up. The designs are around $40 if I remember, and created by someone who worked and built boats in the Royal navy if memory serves.