[vwdevotee]

SO, I am planning on designing a 5-axis router for my universities senior design project. For the various projects my department (mechanical engineering) does, I would like the working envelope to be about 3'tallx5'widex12'long. I am thinking a gantry style would be best, and I am expecting to use Mach3 to control a Geckodrive G100 and then some steppers beyond that. I want to use a readily available motor to run the spindle, so I was looking at standard 2.5hp router motors. Also, I was looking around the forums and liked the idea of a mill bed ATC to speed things up. I would have access to a fully equipped machine shop with a 2-axis CNC lathe, several 2-axis CNC mills, a 3-axis CNC mill, and a 3-axis abrasive waterjet cutter, as well as to a small aluminum foundry to complete my project. Anyway, I was hoping everyone would make there comments and suggestions.

[vwdevotee]

No comments after 57 views?

[randyf1965]

I think the size of it is beyond normal DIY'ing.... but with the machine tools you have available it would be possible.

3' of Z axis travel will be the tough part. alot of force on a 3' lever moving on the x&y axis

I would build it with a truss type framing... lots of triangles.

[vwdevotee]

OH, woops, I forgot to mention the 3-axis plasma cutter, and the shop master who is a fricken genius as resources I have avaible to me.

[wcarrothers1]

Most comments are slow in cases where the project seems to big for the initial question ie: where do kids come from? and big questions like that.

Course then again my first mill (other then my mill conversion) was a 4x8' which I'm pretty sure most people think that is to big to be the first,.

b./

[TCSpooner]

Hi

It seems pretty ambitious to biuld a 5 axis machine but you only get to go round once. I am a mechanical engineering student and three of us recently completed a 3-axis CNC router with a huge work envelope 132" x 60" x 32". So do your statics home work for the structure then do the dynamics. Most good suppliers of cnc parts and equipment have the right specs for you do do a first pass analysis of what will be involved. Before you even start make sure you know what you wnat the machine to do and try to make it realistic. My rule of thumb is 3.14:1 design:build ratio. Design it then design it twice again. Making the machine is the fast part it's getting the design right that should take the majority of the time.

[JerryFlyGuy]

VW, Well if your serious about doing this then go for it, however I will say this. Before you start, set down what your spec's are, just like any other project of this type. Nail down those spec's and then in case you want to move them later.. screw,bolt and glue them down .

Why do I say that? Well.. I'm building a 20' x 10' x 4' milling machine. I'm to the stage where I'll be starting to install the electrical components and doing the final assembly of the major parts in the next few weeks [ at least I hope to be]. I've been working on my project for nearly 12 months, about 8 of which were actual "hands on" working, the rest has been design and research. I've averaged about 10-20hrs a week of work on the project, since it started.

Initially I suffered from "Design Creep-i-tous" where I wanted to do it "this way", then realized that for only a little bit more money or work I could do "that way". This lead to continual changes and higher and higher tolerances and higher and higher expected performance, and hence cost.

So, if your wanting to machine wood or foam or light plastic's, you'll have a totally different set of spec's than if you want to do alum or steel or Ti.

A couple points to ponder, typically linear rails come in 10ft lengths, if your going this route, a 12 ft machine isn't the best mat'l usage [10 of 15 would be better], 5ft wide is fine as you'll need two rails @ 5ft or one 10ft rail cut in 1/2. The 36" is probable fine as well. You can buy the rails cut to length, but you'll usually get a better deal if you buy raw lengths and cut them yourself.

If you plan to use a standard 2.5hp wood router, the ATC option is fairly remote. [I've never seen it done, but thats not to say its impossible.. depending how complex you want to get.. oh.. and how much $$ you wanna spend ] If you go w/ a cheaper draw-bar style ATC, your looking at about $5-8000Usd however its about 20-24" long, a smaller air powered one is over $12000. [those numbers include the needed VFD's and cooling equipment].

Finally, 5 axis rolls off the tongue easier than it happens in real life. I've planned to use 5 axis on my machine as well, but for starters I plan to use a simple 3 axis setup just to get things working. CAM software for the 5 axis is going to be more fun.. and cost. Mach3 doesn't currently do the needed kinematic tool offsets needed to run most 5 axis CAM program code [the CAM usually puts out tool tip position and index position and lets the control software calculate where the machine position is due to index position and tool length.

Finally, something this size is going to take you at least twice as long as you think, and thats not counting any glitchs that come up, and it will take twice as much money as you initally guess as well.

One thing you'll notice through all of this is money.. the more you have to spend, the more options and possibilities there are, on the cheap its still possible but your spec's will have to reflect this. [ie; DIY rails will give you +/- 0.01" at best not +/- 0.001" ]

At first glance it may appear that I'm pretty pessimistic when it comes to all this, I'm not. However, the real facts might as well be presented before a person jumps into a project of this size.



Hope that helps...
Jerry

[chad123]

I 100% agree with Jerry, Not just in theory but in practice also. Mine is 10' x 12' x about 38" . It has been quite a ride. The first 5 axis head was a little rickety and has been converted back to three axis pending a complete 5 axis head redesign. I also am waiting for art to get to the 5 axis kinematics (hopefully some day) running
.
When you build something this big you learn all about heavy mass and resonances. Feature creep has also been a problem, some just by wanting to do it better and some by necessity. As of this month it has been almost a year . Lots of design- redesign. It ended up costing around triple from the first draft. and isn't done yet. Also count on around ~$20.000 for a continuous 5 axis cam package and custom post for mach. Maybe you can get mastercam to pony up a student version..

I am the last person to advise not biting off more than you can chew, just be realistic of what can be done in the scope of a senior project.
Big iron, big motors, big drivers, big rails, big ball screw- belts- rack, money, time. But it is possible i am about 80% there and cutting, and boy it is fun to watch it move.

Chad

[Switcher]

The links below should help as far as 5-axis DIY.

Software:
http://www.rainnea.com/cnc_toolkit.htm

Machine:
http://www.rainnea.com/cnc_5axis.htm

Desktop Machine:
http://www.rainnea.com/cnc_5axisMill.htm



.

[chad123]

VW why don't you lay out your plans so far for us and and we will throw out some suggestions and try to get you staerted on the right track.

Chad

[turmite]

VW why don't you lay out your plans so far for us and and we will throw out some suggestions and try to get you staerted on the right track.

Chad

Chad are you the same Chad from the Mach support group? If so detailed build photos would be great. I saw the ones on the support group, but they weren't detailed enough.

Mike

ps if you are the same Chad I'll be calling you in the nex couple of days.

[chad123]

Yea that's me.. I know i need to take more pictures, it is all i can do just keep adding a couple now and then.

Too much stuff to do

Chad

[vwdevotee]

At first glance it may appear that I'm pretty pessimistic when it comes to all this, I'm not. However, the real facts might as well be presented before a person jumps into a project of this size.
Jerry

I appreciate your candor. I have been thinking about my project a lot more lately, and considering everyones input along with my faculty adviser. For my senior design project I am thinking that I will do just that, design the system and specify all of the parts, but not actually build anything. I also think that a little revision is in order. I have been reading various ATC project on the forums, and I think that they themselves could be project without the rest of the CNC machine. So, I am going to abandon building that in. Maybe someone else can design that if anyone wants to pick up where I leave off. I am also thinking, after talking with the interested groups in my department that building 5'x10' would be plenty large enough, so I am going to go that direction. I am planning to design it with the intention of cutting foam up to maybe wood. I would like to be able to cut aluminum, but I don't think it would be very realistic to design a stiff enough gantry structure to withstand the cutting forces. For the projects in my department I think that having a 0.015-0.025 inch accuracy should be fine, especially given the size parts being made. I am probably going to go with using rack and pinion for my drive because I think over 10 feet the whip in a leadscrew would be a bit scary when I try to rapid from one end of the table. I am hoping to build the machine with the 5-axis head right away, and not hook up the 4th and 5th axis steppers right away, but just get the 3-axis base working and then hook up the other axis when I get to that point. Incidentally, rather than MasterCAM I will be using FeatureCAM to generate all of my G code since that is the one that my school has purchased licenses for. Are the kinematics of my cutter head something that I will have to design and build into a post processor for FeatureCAM? Also, what type of millbed material is commonly used for a project of this size? For some reason I can't see to many catalogs having a listing like "millbed, cast-iron, 5ft x 10ft'. Do most people just use steel plate, or REALLY thick plywood or what? Thanks for all of the good thoughts and advice. Please keep the ideas coming.

[Madclicker]

Glad to see you are throttling back. I did my senior design many moons back when ABET first started to make it part of the accredidation requirement. In my senior design class we were judged as much on realistic results as our ambitions. If it didn't work it was a seriously awkward demonstration!

My school was a pretty highly rated engineering school (CU), but i doubt the entire graduating class could have built a viable 5 axis machine in one semester.

[JerryFlyGuy]

VW, good to hear I didn't step on your toes too hard . I know where your coming from, you have an great idea but reality kinda pulls ya back to earth. I think your definatly going in the right direction by doing things on a bit smaller scale. I think your tolerance is very acceptable and should easily be achievable. As to FeatureCAM, I don't know if it does kinematic offsets or not for sure. But I think I've heard it does. It's a question I've asked their rep, but never got a reply back on. The difference between FeatureCAM and some other 5 axis CAM packages is that FeatureCAM isn't true 5 axis, its 3+2 or possitionable 5 axis. This makes a huge difference to your code generation and how the machine controls your tool. As to the base, it depends on what you want. Because your tolerance is alot lower than say, a Bridgy you could get away w/ just putting a floor plate under it, either of plywood or steel. On my machine I'd been planning to put cross members across between the two x axis rails and then screw down the mill base plate to these, I was just going to use two layers of 3/4" plywood, the first is screwed down to the cross members and the second is screwed to the first. The second layer then becomes replacable as need be [ you'll nick and gouge it for sure over the course of time], I don't think you'd want a steel plate there simply because if you run your machine down into it by accident.. something is gonna bend or break.

Btw, rack and pinion is a great idea for stuff of this size of course I'm bias as I'm doing the same thing myself I'm using 20DP instead of 14.5 as you'll get better strength and tooth engagement.. [so they say.. I'm no expert in this area.. well no expert in ANY area if the truth be told ]

Keep us posted as to how your 5axis head comes out.. I haven't got mine nailed down as to how I'm gonna make it yet.. but I did buy my spindle for it today..

Jerry

[turmite]

Guys I probably shouldn't post this, but I know for a fact that there is a manufacturer who is about to hit the market with a axis 4 and 5 to be attached to a 3 axis machine. I'm curious if this could be included into the design and not have to build the 4th and 5th?

BTW I have a Digital Tool machine that has been running rack and pinion since 1991, and it's never missed a second due to the r/p. I'm hoping to build a machine much like has been proposed here basically for the same reason, foam. My problem is I need about 48" of clearance to do what I want to do. I may already have my first $$$$$ job for it so I need to get started. If you guys have not seen Chad's machine over on Mach support you need to go check it out. It is super nice.

Mike

[vwdevotee]

check out www.5-axis.com I think that they are the company you might be thinking of. I was actually looking at their patent documentation to see how there head is design, so that I could see what the usual and easiest way of designing mine would be. It ended up being really really simple, just a set of spider gears.

[JerryFlyGuy]

Guys I probably shouldn't post this, but I know for a fact that there is a manufacturer who is about to hit the market with a axis 4 and 5 to be attached to a 3 axis machine.

Mike

So Mike, you've got my ears perked up.. don't hold out on us now! Whats the scoop? I've looked at that site that VW listed, in the past and it looks like a solution, however I'm sure the cost if it is still in excess of what I want to spend.

Curious...

Jerry

[turmite]

No, not the same company. This is an individual in NZ. I will email him shortly to have him come introduce himself.

Mike

[chad123]

Turmite: Thanks!

As for the 5 axis heads there are some you can buy right now from Colombo and HSD. They are really nice the problem is they start at $25.000 not including the vfd or cooling.
The one from www.5axis.com are around $30.000 not including servos or spindle motor. The picture that they show looks nice and compact but the servos actually hang off the back and is kind of a hack, doesn't look good and i would think would get in the way of doing large pieces. Remember that these are made to retro fit a vmc.

Let us know who this NZ guy is. If the price and performance is right i would rather buy than build

Chad