Thread: One ring to rule them all?
I'm a new member here, a little about me, I worked at a college for 10 years and retrofitted several manual machines to CNC. Mostly old Bridgeport style knee mills and a couple lathes, so I'm no stranger to CNC however I've never built a machine from scratch. I no longer work there so now I want to build my own machine. I've wanted to build one for years and now I'm finally going to do it. I realise that this is an ambitious project, especially for a first build and it won't be cheap but I'm willing to do whatever it takes, within reason of course. So I'm really looking for a bit of advice and guidance before I start the project, I've got plenty of ideas floating around in my noggin but I've not set my heart on any particular design yet. Basically I'm trying to squeeze as many features into one machine as possible...
My design goals are;
-Ability to machine hard materials, steel and maybe even titanium on occasion.
-Router/ Mill & Lathe combo
-Large Z travel
These are absolutely essential and non negotiable, if it's not feasible then I will have to look at another plan.
-Ideally It would be small enough to fit inside a Ford Transit van or similar for transport, my garage space is a little bigger but not much so, I could sacrifice transportability for size but because my garage space is only a little bigger anyway I think the advantage of mobility is preferable. I'm not planning to move it around but if I ever needed to it would be whole lot easier if it fit in a van. So I'm thinking 1300 wide by 2000 long by 1600mm high. Ideally I want a lot more width, by my reckoning I'll get about 700-800mm machine travel out of the 1300mm width which is pretty pathetic.
-5axis machining, not really essential but would be nice. I plan to build a lathe into the machine with an indexing ability so that would be one way I could get a 4th axis. Great for small round parts but not ideal for large flat parts. I think it would be quite useful to route/mill and drill large flat pieces at angles other than 90degree to the bed. Not sure exactly the best way to go about this though. I don't think a traditional 5axis head attached to the Z would be particularly good for drilling. What you really need is another axis attached to the head that plunges.
-Automatic tool change
Because I'm working with a small footprint I think the only way I can get a decent work area is to go with a moving gantry design.
I'm thinking the construction will be mostly 100x100x5mm box steel fully welded. Gantry could be 200x100x8mm box steel or even two welded together. Probably use 25x5mm ball screws, HIWIN HG25 or similar linear rails, 200 to 300mm carriage separation. Does this sound about right?
I don't think it will be particularly a problem to machine hard materials vs soft materials as long as I keep within the limits of the machine which just means going slower.
I plan on making my own servo drives using Arduino, I've already started looking into it and I'm confident it will be a success.
I've read a bunch on this forum and a couple things keep popping up, a long Z travel is bad but I really want it. What is the problem specifically with a long Z? I had a few ideas how I might stiffen up a traditional design. I don't really see how the frame could be the weak link but if it was I could just go with bigger box section, same for the gantry. If the rails where the weak link I could use 4 rails per axis, expensive but if its got to be done then its got to be done. I did have an idea though, what about in addition to the Z lowering the spindle another axis could raise the bed? Similar to a Bridgeport you raise the bed and then plunge with the quill. This would require a lot of linear rails mounted vertically and at least 4 ball screws so it certainly wouldn't be a cheap way of doing it but I think it would be a pretty useful feature anyway so I might add it regardless.
A little unconventionally I'd like the longer axis of the machine to be open for ease of use. So the gantry will be the long axis and run along the short axis, which would make the gantry 2000mm long or so. This wouldn't be ideal from a rigidity perspective but I feel the benefit of easy access outweighs the negatives. 2000mm isn't that long anyway. Any problems with this?
I've checked out some of the calculators I found on the forum and it seems feasible, very low deflection with 200x100x8mm box steel.
Ok that's it for now, any advice and guidance much appreciated. Thanks Rufe0
Wow!! you sound like a man with a plan.
I have nothing positive to contribute, and can only recommend you start sketching up some plans, which will start the feedback coming..
I made a servo, (of sorts) once, using an arduino... this wasnt for cnc use though (it was used for power steering on my lads toylander)
intrigued as to how you intend to tackle some of the design challenges you have suggested (ie, milling steel and titanium), no doubt you have a plan
look forward to your responses... keep the ideas coming. Sometimes new threads dont attract many replies until you start posting pictures :D
You are really planning a beast!
2000mm long gantry with a dual 200x100x8 steel box section.... I cant really imagine it inside Transit van.
It will be really heavy and hard to build at least in conventional home garage. But it is doable, no questions there...
Dont start buing stuff until 100% finished design!!!
Anyways take a look at http://www.mycncuk.com/threads/6619-...an-Unusual-one
Mister Silyavski is into beasts as well :D
I've been going over some of the numbers, I'm thinking the torsional load on the gantry is the weak point.
Using this calculator
http://www.kennametal.com/en/resourc...and-power.html a 16mm 4 flute endmill in steel with recommended chip load feeds speeds I get a tangential load of 1000n / 100kg.
Assuming a Z travel of 1m, 1000nm torque on gantry.
Using HIWIN HG25 linear rails http://cnc4you.co.uk/resources/Hiwin%20General.pdf
basic dynamic load rating (lowest rating)
26kn / 2600kg. At a separation distance of 300mm and 1m Z that would be 780kg or 7800nm (but there is 2 of them so this times two really). Clearly the linear guides are adequate.
So then there is the torsional deflection of the gantry beam.
θ = TL/GJ
θ radians deflected
T torque = 1000nm
L length = 1m
G Shear modulus = 80 steel
T Torsional Constant = 180 m4
= 0.006944444 radians / or 0.12114 mm at 1m Z travel
This is the calculation of a 1m gantry supported at only one end, not sure of the calculation for a torque at the mid-point. I assume it would be less. I assume it would be at least halved with 2 box sections welded together.
So I'm assuming it would be in the range of 0.05 which is acceptable I feel. It most likely wouldn't be enough to matter in most cases and even if it did you could work around it.
What do you think? Have I made a math error?
There are some vertical machining centre ideas here on this commercial machine that might be useful . . . these are purpose built cnc machines, not the converted manual machines that you are used to.
And best of luck with the project !
Z travel 1meter on a moving gantry? I can't imagine it milling steel let alone titanium.
Get some drawings down then post them up for better advice.
Last edited by Tenson; 03-11-2015 at 12:56 PM.
Ha just realized I made 2 maths errors, actually it's
0.0694 radians or 3.978874 degrees which is 69.4mm deflection at 1m. This is a serious problem.
Stocking more goodies than just Smoothsteppers
To machine titanium, steel or stainless, the machine needs to be extremely rigid and 5mm box section is not even close to getting to a point where it could accurately cut.
1M of Z would make the machine flex like a rubber dildo if you tried to cut steel, let alone titanium.
Will all the naysayers please let the man show us his ideas before rubbishing them?www.emvioeng.com
Stocking more goodies than just Smoothsteppers
By Wal in forum Linear & Rotary MotionReplies: 23Last Post: 13-05-2014, 10:49 PM
By mekanik in forum General ElectronicsReplies: 1Last Post: 15-10-2013, 04:50 PM
By jcb121 in forum Open Source SoftwareReplies: 4Last Post: 26-07-2013, 08:00 PM
By crimsonred in forum Probing, Digitizing & ScaningReplies: 4Last Post: 27-09-2010, 01:47 PM
By Pmac in forum Tool & Tooling TechnologyReplies: 4Last Post: 28-01-2010, 06:55 PM