Long Post-Automotive CNC applications

[HarmonicBoiler]

Dreaming of a machine with a large envelope, but I have no CNC experience. (Computer Science - background)

In hopes of fabricating some custom body parts for a car, I had created the parts in 3ds and started calling companies around florida who mill hulls for boats. Upon emailing the files, they said it would be outrageously expensive to mill small body parts like bumpers, etc. at their shops. So one company gave me the phone number of a guy who had custom built a 5-axis machine in his garage. I believe he was some where in North Carolina and his name was Chris. Literally, the envelope of the machine was his entire garage! Upon sending him the files, he said it would be around $15,000 to mill the foam molds for the body panels. Knowing down the road I'm going to need lots of molds done, this led me to this forum.

Ultimately, I would like to build a 4'X8'X2' envelope, 5-axis machine which would be utilized to build foam molds for automotive body panels.

The conventional (non-CNC) route for creating custom body panels is rib construction:
DDR Motorsport
http://www.ddrmotorsport.com/process.html




As you can see, the steps required to get from the digital version (the ribs are created from profile cut outs of the CAD car) to the physical part are numerous. The molds for each part are created from the finished car. This ensures that the car and its parts will look/fit correct. This is way too cumbersome and the individual panels are created in a way that doesn't provide "bolt areas" where the edges of the body panel can then be bolted down into the unseen areas of the car, such as the frame. Example of this (look how the edges of the mold go up instead of down):


Milling molds of each panel is a better alternative. This would allow us to produce a female mold in foam of the body panel, say a bumper, in its entirety. We could then lay fiberglass in the foam female mold and the finished fiberglass part would be a direct bolt on to the car. Out of each female mold we might be able to get 50-100 parts.

From searching this forum, to start the journey, it looks like the best option would be to start small and cheap. Something like:
The Phoenix
http://www.crankorgan.com/phoenix.htm


The next step looks like:
Cad Cat INDUSTRIAL ROUTER
http://www.inshorepowerboats.com/cnc.../CNCROUTER.htm


The envelope of this machine is 4' X 8' X 9". We would need to extend the Z-axis and beef up the frame rails, etc.

Although unreal, it would be great to have something of this caliber, garage-size, which would leave the ability, if need be, to mill the entire car out of foam:
TITAN
http://www.cncmotion.com/titangallery.htm



Knowing what we would like to get done and the conventional methods, what suggestions do you guys have? Would more experience be needed between the Phoenix and the Cad Cut machine? What modifications would be needed to increase the envelope of the Cad Cut machine from 4'X8'X9" to 4'X8'X2'? Any other thoughts or ideas?

Any help appreciated!

[High Seas]

To get the extra Z, it seems like the challenge is one of keeping the whole system rigid. Seems like the side rails on the CAD-CUT could be higher and then the Z is larger. Sorta, "don't raise the gantry/bridge - lower the work surface."
Thats the approach I plan on for getting extra Z in System2. Its not yet 5x9 (to cut a 4x8), but I'm thinking of going to rack and pinion drive when I move up to that size.
Like the car bits idea - I'm on the same path eventually as well as boat bits.
Jim

[pstockley]

I think the problem you are going to have is generating the necessary GCode to run a 5 axis cutter. Software to do this is very expensive $15,000+ Without good software you will have an expensive toy.

[turmite]

You mean carving something like this? This was carved using a 4.75" z 3 axis machine. It can be done like this but not as well as with one of those large 5 axis.

Mike

[ger21]

Originally Posted by pstockley I think the problem you are going to have is generating the necessary GCode to run a 5 axis cutter. Software to do this is very expensive $15,000+ Without good software you will have an expensive toy.

There is this one. http://www.rainnea.com/cnc_toolkit.htm

As for jumping from the Phoenix to the CADCUT, operating them both is the same, it just depends on tour fabrication skills.

[chuckknigh]

What level of precision are you going to require? For a fiberglas mold I would think you wouldn't require much more than the nearest .5 millimeter. If so, then your project is much more achievable by a home builder. I've heard of projects that used garage door rails as linear rails...definitely something to think about, if you want to build it into your garage. But, this approach does not lend itself to ultra-high precision.

If you want precision to .001", be prepared to spend a LOT of money on some VERY sophisticated parts. If you can stand some slop in the system, it'll be much more do-able.

-- Chuck Knight

[HarmonicBoiler]

chuckknigh, you are exactly right. These are going to be body panels and not crank shafts so the precision won't be too much of a detriment. Do you have any contact information for any of the guys who used the garage door rails? If you know of anyone, I'd be interested to even fly over to where there machines are, just to see how it all works. Any other tips or recommendations? Do you think 5-axis would be neccessary or would 3-axis do? I'm thinking the front/rear bumpers are going to be the most complex shapes which need to be milled.

Any help appreciated!

[turmite]

Originally Posted by HarmonicBoiler Do you think 5-axis would be neccessary or would 3-axis do? Any help appreciated!

Well I tried to help! The top shown above was cut with a 3 axis! Do you need 5 axis or not? It would have been much easier to make with 5 axis as would the bumpers, but it can be done with limited z and some pre-planning.

Mike

[chuckknigh]

No contact information, but I remember seeing it mentioned on this site, a LONG time ago. There's a search function on this forum...

I think it's pretty obvious, though. The idea is to make a "block," which we call a gantry or a table (depending on the design) move in a straight line. And, to do it with the least amount of mechanical effort and wobble, possible.

Go outside and look at your garage door. Get inspired! It's a large "block" which is moved linearly by a screw or chain drive, on 2 linear rails. The rollers are of low quality, but that can be taken care of with some more creativity.

Fair warning -- a garage door rail system is NOT going to be as precise as a "properly built" DIY mill. It may be adequate for your needs...but examine your needs before committing to one design!!!

Back to small machines. The Phoenix, as well as the rest of John's designs, are very good starting points. They're well thought out and well executed machines that actually work, and which are buildable for fairly little money. You can also use them to build the precision parts for your big machine. We say it often -- build small first.

-- Chuck Knight

[Snipes44]

Would have to agree. Build small first. There is a lot to learn and its way better to crash a small mill then a large one :-). Not that I have ever crashed mine. lol

I have built John's "The Brute" for milling and drilling out circuit boards. Works very good once I got all the bugs worked out. Just a heads up on John's plans. I have seen a few people complain about John's plans in that they don't give you an exact parts list. His theory is more giving you the concepts on how to build the machine using what you have.

I highly recommend his plans.

Hopefully you will be posting progress reports as your project progresses.

Good Luck!

[buglerroller]

5 AXIS is the way to go. HarmonicBoiler what software are you familiar with? have you heard of a company called 5 axis models in Huntington Beach, California? My old company manufactured bumpers, hoods and other fiberglass and urethane products but after i couldn't find more info on the cnc process i threw in the towl and sold the company. i found this site how i don't know but when i saw this thread i had three legs! i have alot of time on my hands now and willing to help in research or even funding based on this post! do you have an email or im we can talk on?

[Glidergider]

Hey HarmonicBoiler
You might want to consider using cnc foam cutting rather then machining. It's much less expensive to build a large volume cnc foam cutter.

To get complicated 3D forms with a wire cut, you would have cut slices of model in foam, much like your plywood shapes. Except your end product in foam would be solid foam. You would need to do some final sanding and shaping.

I saw a stealth fuselage of a military unmanned airplane get shaped with this method. On top of this would be carbon, kevlar and fiberglass coatings.