Need Help!- Total Beginner & Project - DIY 3D 3-Axis CNC (foam)

[kaidomac]

First of all, I hope I'm posting in the right forum. I'm new to CNC and looking for some advice on working with my father (engineer) on a DIY CNC rig. Second, I have two general questions:

1. The project we have in mind is a DIY 3-D 3-Axis CNC milling machine for cutting Tooling Foam (RenShape). Budget is $5k - $8k. We want to be able to do profile contour milling with a profile tolerance of at least .030". The desired volume is 6'W x 6'L x 3'H. I guess my first two questions are, is this something that is possible to DIY, and second, is the proposed budget too much/too little/about right?

And just dreaming here, but would it be possible to do a DIY 5-axis CNC accurate to .001" (or .010" if that's not possible) with a material volume of 12'W x 12'L x 6'H? (Space for the machine is not an issue). If that IS possible, what would the ballpark cost be? (again, tooling foam is the material we will be working with, if that makes a different in materials/build structure of the machine).

2. I'm a total beginner at CNC and need some advice on where to get started. My background is computer hardware and programming. I'm working with my father on this project; his background is in Engineering. I don't have any experience with computer-controlled tooling, just standard hand tools, but I am more than willing to learn! Here's one resource I was looking at picking up:

http://www.abcsofcnc.com/index.html

Any advice on where to get started would be great, thanks! I'll have 15-20 hours to put in every week into learning & building, as this will be my new hobby for the next few years. Thanks

[awerby]

kaidomac First of all, I hope I'm posting in the right forum. I'm new to CNC and looking for some advice on working with my father (engineer) on a DIY CNC rig. Second, I have two general questions:

1. The project we have in mind is a DIY 3-D 3-Axis CNC milling machine for cutting Tooling Foam (RenShape). Budget is $5k - $8k. We want to be able to do profile contour milling with a profile tolerance of at least .030". The desired volume is 6'W x 6'L x 3'H. I guess my first two questions are, is this something that is possible to DIY, and second, is the proposed budget too much/too little/about right?

[Those are feet, right? Have you priced that Renshape stuff lately? The budget you've allocated for your machine would not be close to enough for a piece of Renshape that would fill your volume even once. (Renshape 450 costs about $0.156 per cubic inch - see:

http://www.kranderson.com/shop/Hunts...50-2X16X60-1BD

At that rate, a piece that's 72" x 72" x 36" would cost over $29,000.00.

Sure, it's possible (for some people) to build a machine that big with that amount of money. But are you those people? Do you have a well-equipped shop with welders and mills, lathes and grinders? Are you an expert electrician and machinist? Do either of you have a background in electronic assembly? Yes, if all that was the case, sure you could. If not, then you'd have to hire people to do at least some of this, which drives costs up.]

And just dreaming here, but would it be possible to do a DIY 5-axis CNC accurate to .001" (or .010" if that's not possible) with a material volume of 12'W x 12'L x 6'H? (Space for the machine is not an issue). If that IS possible, what would the ballpark cost be? (again, tooling foam is the material we will be working with, if that makes a different in materials/build structure of the machine).

[Machines like that go for hundreds of thousands of dollars. There's a reason you don't see too many home-built machines out there with a six-foot Z axis, let alone the 4th and 5th axis attachments. The longer your Z-axis, the more leverage you have working against you. The more leverage you're combatting, the heavier and stiffer everything has to be. The heavier those moving parts are, the more inertia they have to overcome in starting and stopping. So instead of the hobby-grade motors, controls and motion-control equipment that most people around here are using, you'd need the heavy-duty industrial equivalents, which cost a lot more, and are more complicated to set up.]

2. I'm a total beginner at CNC and need some advice on where to get started. My background is computer hardware and programming. I'm working with my father on this project; his background is in Engineering. I don't have any experience with computer-controlled tooling, just standard hand tools, but I am more than willing to learn! Here's one resource I was looking at picking up:

http://www.abcsofcnc.com/index.html

Any advice on where to get started would be great, thanks! I'll have 15-20 hours to put in every week into learning & building, as this will be my new hobby for the next few years. Thanks

[Instead of leaping into the design and construction of these monster machines, don't you think it would make more sense to build something somewhat smaller and simpler, like a Mechmate, first? The plans are pretty much worked out, there are people who have built them who can help with advice, and you'd gain confidence and experience in the process. Plus, when you're done, you'd have something that might actually work, and you might have enough money left over to get the CAD and CAM software you'd need to use it. (I can help with that part). Who knows, you might even save enough for some pieces of Renshape...]

Andrew Werby
www.computersculpture.com

[Zygoat]

well i think the first machine you are talking about should be very possible. i would put the second machine to the far back of your mind for now. get going with the first one, after its built and running for a while then you can start thinking about the second.

the first one is on the big side. but there are people here running machines with about the same foot print. Length and width. your z axis (hight) is quite a bit bigger than just about any of the machines i have seen on here. but your tolerance is loose enough that i think this project is very do-able.

i think you will want to build a cnc router. (thats what they are called) it uses a normal router (like for wood) as the motor to power the cutter (the "bit" if your will.) this is where i would start my research. look at some of the threads on here to get on idea of what is going on.

you will have to spend considerable time planning and learning before you start your build. (but it sounds like you are ready for that.)

i am finishing up the design and research phase of my machine and i am getting ready to start the actual build of my machine.

you can check it out here.

http://www.cnczone.com/forums/showthread.php?t=99824

many other very good threads on here. read up and ask questions. people here have been very helpful to me.

feel free to ask me anything or pm me if you have questions, ill do my best to answer them.

[YZF]

My 2 cents:
- As you are cutting foam your cutting forces should be very light which is a big help.
- The 0.03" accuracy spec is almost 1mm which is really not that accurate at all. So that one is not a problem if you follow some basic engineering principles.
- You haven't told us how fast you want to be moving/cutting. I.e. do you care if it takes 2 days to cut?
- Given those I think your budget is fine or this could even be done for cheaper.

To start:
- Review some other designs in these forums. In particular see how the linear drive systems work (screw/motor/guides) and how the different frames are designed.
- Read up some on stepper motors, there's a lot of information on the web.
- F = ma :-)
- You'll need to figure out how you lay out your design. Paper, SketchUp, CAD?
- Do some system's engineering - figure out your speeds, acceleration, masses, forces...
- Read on some of the available control software Mach, TurboCNC, EMC and how they interface with your machine.

If you've done your research and have some questions I'm sure there are a lot of people here who would help...

In terms of your dream machine. A rule of thumb that I've learnt is that accuracy of 1um/1m is the limit of what can be done without some very exotic technologies. That accuracy is also very hard to achieve. As you go up in order of magnitude it becomes easier and easier. At the higher accuracies various compensation techniques are applied and very expensive equipment (interferometers) is used to calibrate the machine. Other things that pop up are thermal issues, stiffness, vibrations etc. I've seen people here build incredible machines, DIY just means that you do it yourself, there's no real reason why you can't build anything that anyone else can build. Just a combination of time, money and knowledge.

[kaidomac]

Originally Posted by awerby Those are feet, right? Have you priced that Renshape stuff lately? The budget you've allocated for your machine would not be close to enough for a piece of Renshape that would fill your volume even once. (Renshape 450 costs about $0.156 per cubic inch - see: http://www.kranderson.com/shop/Hunts...50-2X16X60-1BD At that rate, a piece that's 72" x 72" x 36" would cost over $29,000.00.

Correct, that is in feet, and excellent point. The idea is to use a faceted plywood substrate - so that only an inch of foam or so would be needed to approximate the countour with the plywood facets. For some applications, we can also use less-expensive expanded polystyrene (EPS).

Instead of leaping into the design and construction of these monster machines, don't you think it would make more sense to build something somewhat smaller and simpler, like a Mechmate, first? The plans are pretty much worked out, there are people who have built them who can help with advice, and you'd gain confidence and experience in the process. Plus, when you're done, you'd have something that might actually work, and you might have enough money left over to get the CAD and CAM software you'd need to use it. (I can help with that part). Who knows, you might even save enough for some pieces of Renshape...]

That sounds excellent - again I'm admittedly a total newbie to the world of DIY CNC here, so a beginner project would be great. I'll take a look at the Mechmate this week, thanks! As far as CAD software goes, we're covered in that aspect (we use a variety of software from AutoCAD to UG NX), but I will need some advice on CAM software, especially stuff that will play nice with a DIY rig.

[kaidomac]

Originally Posted by Zygoat you will have to spend considerable time planning and learning before you start your build. (but it sounds like you are ready for that.) i am finishing up the design and research phase of my machine and i am getting ready to start the actual build of my machine. you can check it out here. http://www.cnczone.com/forums/showthread.php?t=99824 many other very good threads on here. read up and ask questions. people here have been very helpful to me. feel free to ask me anything or pm me if you have questions, ill do my best to answer them.

Wow awesome!! Do you have a schedule in mind as to when you'll have it finished? That's a pretty admirable project for the budget!

Yeah, I've got a lot to learn, so any books/websites/video training you could recommend for a beginner would be great! I just need to get pointed in the right direction so I'm not off chasing rabbit holes, you know?

[kaidomac]

Originally Posted by YZF My 2 cents: - As you are cutting foam your cutting forces should be very light which is a big help. - The 0.03" accuracy spec is almost 1mm which is really not that accurate at all. So that one is not a problem if you follow some basic engineering principles. - You haven't told us how fast you want to be moving/cutting. I.e. do you care if it takes 2 days to cut? - Given those I think your budget is fine or this could even be done for cheaper. To start: - Review some other designs in these forums. In particular see how the linear drive systems work (screw/motor/guides) and how the different frames are designed. - Read up some on stepper motors, there's a lot of information on the web. - F = ma :-) - You'll need to figure out how you lay out your design. Paper, SketchUp, CAD? - Do some system's engineering - figure out your speeds, acceleration, masses, forces... - Read on some of the available control software Mach, TurboCNC, EMC and how they interface with your machine. If you've done your research and have some questions I'm sure there are a lot of people here who would help... In terms of your dream machine. A rule of thumb that I've learnt is that accuracy of 1um/1m is the limit of what can be done without some very exotic technologies. That accuracy is also very hard to achieve. As you go up in order of magnitude it becomes easier and easier. At the higher accuracies various compensation techniques are applied and very expensive equipment (interferometers) is used to calibrate the machine. Other things that pop up are thermal issues, stiffness, vibrations etc. I've seen people here build incredible machines, DIY just means that you do it yourself, there's no real reason why you can't build anything that anyone else can build. Just a combination of time, money and knowledge.

As far as the lightness of the foam goes, we'll be using low-density foam on the order of 1-pound to 40-pounds per cubic foot. No wood or metal cutting at all. I'd like it to be fast, especially since it's just cutting low-density foam - should be pretty quick. So if that could be kept in the ~$7,000 range, that'd be awesome! For the system design, we can draw it out on paper, we can use Sketchup, we can use CAD, no problems in the design area

Starting out with a smaller, simpler 3-axis project would probably be more up my alley to begin with, as I am not only new to DIY CNC but to CNC's in general. Computers & electonics are no problem, drafting/design is no problem, so I think the two areas I need to look into in detail in my research are (1) machine design (frame, components, etc.), and (2) CAM software to interface with the machine (assuming a model in step file or IGES is ready to go). Are there some commonly-recommended books or good tutorial websites I could take a look at to get a basic foundational knowledge, or is that mostly gained from browsing the forums here?

[Zygoat]

Originally Posted by awerby [Instead of leaping into the design and construction of these monster machines, don't you think it would make more sense to build something somewhat smaller and simpler, like a Mechmate, first? The plans are pretty much worked out, there are people who have built them who can help with advice, and you'd gain confidence and experience in the process. Plus, when you're done, you'd have something that might actually work, and you might have enough money left over to get the CAD and CAM software you'd need to use it. (I can help with that part). Who knows, you might even save enough for some pieces of Renshape...] Andrew Werby

im gonna have to disagree only slightly here. the physical size is not that big of a deal, if they have the space for it (meaning also the space to build it.) and the tools for it. im reading in between the lines here, but it sounds to me like they have a decent shop set up to handle the construction. at that point it becomes about the mechanics of the machine. it will require bigger motors and beefier electronics. that is for sure. and these can be more complicated.

for those reasons it might be a good idea to build a smaller one first. BUT. i would not want to build a smaller one just to get an idea of how to do it. if it will be functional for you (ie you can actually use it for something productive) than i think its a great idea. but to build it just to learn, spend $$$$ (big time money) and then still have to build one that you can actually use $$$$. then i think it is more practical to just spend a little more time on the design and research phase and save the money and have a big one you can use.



also just a note. i think it could be done for less than $5k but there is a ton of variables there. could be done for less but could end up being 5 figures easily also.

[kaidomac]

Originally Posted by Zygoat im gonna have to disagree only slightly here. the physical size is not that big of a deal, if they have the space for it (meaning also the space to build it.) and the tools for it. im reading in between the lines here, but it sounds to me like they have a decent shop set up to handle the construction. at that point it becomes about the mechanics of the machine. it will require bigger motors and beefier electronics. that is for sure. and these can be more complicated. for those reasons it might be a good idea to build a smaller one first. BUT. i would not want to build a smaller one just to get an idea of how to do it. if it will be functional for you (ie you can actually use it for something productive) than i think its a great idea. but to build it just to learn, spend $$$$ (big time money) and then still have to build one that you can actually use $$$$. then i think it is more practical to just spend a little more time on the design and research phase and save the money and have a big one you can use. also just a note. i think it could be done for less than $5k but there is a ton of variables there. could be done for less but could end up being 5 figures easily also.

Yup, tools & space are no problem, neither is the design/drafting aspect. If I could jump into a 12x12x6' 5-axis CNC beast, that'd be great, but I'm also willing to build a smaller, less-complicated model to learn on, if necessary. Ideally I'd like to build my dream machine right off the bat, of course, so if all that is needed to do that is more time and research, that is fine. I don't know how much we'd use the smaller machine after the big one is built, so if going straight to the monster is doable, I'm all for that!

The budget is flexible. However, I want something that will work, not a toy or something that will break easily, so if spending a bit more is in order, that's okay. Obviously the smaller the budget the better though

[YZF]

The CAM software takes in your 3D model and spits out G-code. G-code is fairly simple. There's a wide range of CAM software with different capabilities and prices. If you're a software person you really shouldn't have a lot of issues there.

Machine design wise, I would start with the forums, just look at a few different designs. Check out the awesome 5-axis build that was recently posted here. Also linear motion products manufacturers will have application notes on their web sites.

You'll find some more awesome machines here:
http://www.cnczone.com/forums/vertical_mill_lathe_project_log/

And check out this build:
http://www.5bears.com/cnc.htm

[YZF]

Also worthwhile taking a look at some commercial CNC machines...

There's also the question of whether or not you're looking at this as a fun/learning project or looking for the fastest way to get something up and running without caring too much about the details. I.e. there's shortcuts you can take duplicating elements of other designs, buying vs. building, plans etc.

[Phife]

Originally Posted by kaidomac Yup, tools & space are no problem, neither is the design/drafting aspect. If I could jump into a 12x12x6' 5-axis CNC beast, that'd be great, but I'm also willing to build a smaller, less-complicated model to learn on, if necessary. Ideally I'd like to build my dream machine right off the bat, of course, so if all that is needed to do that is more time and research, that is fine. I don't know how much we'd use the smaller machine after the big one is built, so if going straight to the monster is doable, I'm all for that! The budget is flexible. However, I want something that will work, not a toy or something that will break easily, so if spending a bit more is in order, that's okay. Obviously the smaller the budget the better though

If you want to build a 5axis machine expect to blow your entire $7000 budget on the software alone!