[joeavaerage]

Hi,
my mill has travels of 350mm x 350mm x 350mm with screws that are 650mm long overall. They are 32mm in diameter and 5mm pitch. I direct drive them with 750W servos
at 3000 rpm for a max cutting speed of 15m/min and spin them at 5000rpm for a max rapid traverse of 25m/min and have acceleration of 0.27g at rated torque or 0.75g at overload torque.

The axes weigh about 150kg to 200kg. As it turns out I could have used 25mm diameter screws but 20mm would have been too small. I used 32mm because that was what was available
second hand (near new old stock) at the right price....and often whats available (affordably) trumps whatever it is that you want.

The size of your machine (proposed) is too big for 16mm screws, in fact I think 20mm is marginal as well. I'd say 25mm is the sweet spot, while 32mm is great for stiffness and accuracy but will
require bigger servos and obviously cost more in the first place.

I'm sorry to say that I believe the size of your machine is too ambitious, its going to cost a small fortune. I think travels of 250mm x 250mm x 250mm are more practical...and certainly cheaper
and much more readily achievable. My machine has cost well over $30,000NZD and that's without a spindle, nor the fourth and fifth axis. I think you are in danger of saying 'I want this and that at such
and such a size' only to find that you cant possibly do it. I've been there, and have to avoid the thinking that leads there daily.

Craig

[peteeng]

Hi Sus - Yes you can use a 15mm for any axis, just that may limit you in terms of speed or buckling. You have to start with what you want to achieve vs making up front cost decisions on parts. Identify costs and where you can get things but don't commit to them until the design is complete. Your mill is not a cheap exercise so you need to go thru the specs, figure out the cost of the machine then go around the cycle again until it fits your budget. If you make individual cost or performance decisions along the way (pick this, pick that) the machine will not be commensurate with itself or your expectations. The best approach is to design the machine in total not in pieces. Then at the end figure out costs, then start again and trim it down to fit your $$$....

Now your into the Z axis and spindles are you going to balance it? are you going to use a brake on it? The size spindle/Z axis your describing will self drive down due to its own weight on the ballscrew when the power is off. So this has to be managed somehow. Peter

What are your current parameters and I'll run a one page calc for you. Are you using steppers or servos? What torque rating are they? Have you used the spreadsheet I posted? Peter

[joeavaerage]

Hi,
you may have noticed that the Z axis servo on my machine is still 750W but is longer than the others? Its because it has a built-in electromagnetic brake.

With the little spindle I got at the moment the Z axis is not inclined to back drive, but once I pit my BT40 toolchanger on there it will, so I bought a servo with a brake.
Cost and extra $150USD ($588USD vs $438USD), but its there when I need it.

Craig

[pippin88]

Ball screws - I don't think our machines approach buckling load (unless you are way undersizing the screws), and don't think you need to calculate that really.

Critical speed / whip speed matters. Very dependent on length
There are easy calculators for this.

Then there is a max rpm related to the speed of the balls.
Often termed DN or Dm.N. Diameter of screw times RPM = Dm.N value. Need to keep this below manufacturer value, often 50,000 or 70,000. 20mm screw times 3000rpm = 60,000.

So there is sometimes an advantage to smaller screws for Dm.N value.

Larger screws are stiffer

A lot depends on rpm you will be driving screw at, which depends on steppers / servos, direct vs reduction (belt/gear) drive etc etc

Rotational inertia becomes a big issue with larger screws. At 30mm and above, screw rotational inertia dominates (much more than inertia of moving parts). At 25mm screws and below, rotational inertia is less of an issue. Depends on relative mass / inertia of moving parts and screws of course

[suspension]

Hi,
I'm sorry to say that I believe the size of your machine is too ambitious, its going to cost a small fortune. I think travels of 250mm x 250mm x 250mm are more practical...and certainly cheaper
and much more readily achievable. My machine has cost well over $30,000NZD and that's without a spindle, nor the fourth and fifth axis. I think you are in danger of saying 'I want this and that at such
and such a size' only to find that you cant possibly do it. I've been there, and have to avoid the thinking that leads there daily.
Craig

peteeng[/COLOR];]
Hi Sus - Yes you can use a 15mm for any axis, just that may limit you in terms of speed or buckling. You have to start with what you want to achieve vs making up front cost decisions on parts. Identify costs and where you can get things but don't commit to them until the design is complete. Your mill is not a cheap exercise so you need to go thru the specs, figure out the cost of the machine then go around the cycle again until it fits your budget. If you make individual cost or performance decisions along the way (pick this, pick that) the machine will not be commensurate with itself or your expectations. The best approach is to design the machine in total not in pieces. Then at the end figure out costs, then start again and trim it down to fit your $$$....

This is true, I understand this could be too ambitious in terms of build and cost as this is my first one. Given that this requires lots of work and patience I like to use that to build what I would ultimately want and hence the size. But if the cost becomes too much, I will certainly have to make it smaller. I am curious as to why it cost that high though. To my understanding, the highest cost would be for ball screws and spindle. Servos will probably be from JMC, the metal work will not be that costly (for stand/etc). A rough calculation would be:

Ballscrews/nuts/supports - say 25/32mm new-open box -> 2000
Motors/brackets -> 750
Spindle ATL -> 2500
Frame -> 1000
Precision milling of surfaces -> 500
Counter w eight/etc (Air cylinder based) -> 200
Linear rails/etc -> 1500
Other -> 1000

Total -> 10,000 approximately.


BTW, I have much of the stuff needed for ATC already so its not included above.

[pippin88]

https://www.engineersedge.com/calcul...crew_15635.htm

Safe value for actual machine use is max rpm = 80% of ball screw critical speed

[joeavaerage]

Hi,
to be honest I'd say that C5 or C3 screws are the most expensive part of the build and Sus is leaning towards dyGlobal selling second hand and new old stock
ballscrews. I got $9000NZD of screws and support bearings for $1800NZD including shipping. The amount that can be saved just can't be passed up.
It also means that you take whats available and match your machine to what you can get.

I'd say to OP take a very critical look at whats available and get what you can....and let that determine the starting point of the build.

I would advise against trying to make the machine as big as you might like, but keeping it small means you can affordably build a super rigid and accurate machine.
Accuracy and rigidity are by far and away more satisfying than size, and those properties will characterize all the work that you'll ever do with that machine.

Craig

[joeavaerage]

Hi,
just as a comparison, all these prices I've converted to NZD including taxes where due:

three cast iron axis beds 115kg each $3000
Machining/stress relieving of cast iron beds $6500
three servos (750W Delta B2's, one braked, including shipping to NZ) $3250
three BNFN C5 ballscrews and FK support bearing including shipping $1800
three sets 20mmHD THK linear rails/cars, new old stock, including shipping $1100
Profile cut steel for frame 32mm med tensile $1800
Profile cut steel for saddles (3) $190
surface grinding saddles $90
Profile cutting blocks for supports and ballnuts $345
Carbide tooling for machining $205
Hire of mag-drill for frame drilling $195
4 x 150kg castoring wheels $470
Sheet steel and bending for coolant tray and way covers $360
Aluclad sheeting for enclosure $145
Aluminum extrusions for enclosure $75
Bolts Various $120

TOTAL $19,645NZD (or about $12,376USD)

Please understand this doe not include any electronics, some of which came from my previous mini-mill, nor does it include either of the two spindles, which likewise came from
my previous project. Nor does it include the $5700NZD I'v invested into the fourth and fifth axes to date, with still another few hundred to complete the fifth axis. Nor does it include any
of the Fusion 360 subscription that was used in its design and build.

All up this is an expensive hobby. If it were not for the fact that I use this machine daily to make a living I could in no way have justified its expense. You may also be thinking
'He's just gone and bought what he wanted without regard to the cost', I assure you I have not. I have to consider carefully every decision say with regard to the cost/quality trade-off
at every step. I won't say I always buy the cheapest, but try to buy the best affordable quality that matches the overall objectives of the design.

I am curious as to why it cost that high though.

Oh, I bloody well know why....good things cost......and cost a bomb!

Craig

[pippin88]

Still cheaper than fast cars

[joeavaerage]

Hi ,
I notice in a previous post I posted 'My machine has cost well over $30,000NZD and that's without a spindle, nor the fourth and fifth axis.'

As you can see from this list of costs , that I can remember anyway, that quote was not accurate, had I said $30,000NZD including spindles and fourth/fifth axes would be accurate.
Apologies, I do not intend to mislead.

Craig

[joeavaerage]

Hi,

Still cheaper than fast cars

and booze and women folk.

Craig

[ardenum2]

Precision milling of surfaces -> 500

you're missing one 0 in there. Why don't you use DWH from Diamant for the rail surfaces, then it would cost 500



...$30000 hahaha my sides hurt.

[NordicCnc]

you're missing one 0 in there. Why don't you use DWH from Diamant for the rail surfaces, then it would cost 500



...$30000 hahaha my sides hurt.

Whoch DWH from diamant is recommended? Myself I am planning to use R&G epoxy mixed with metal powder, I saw Scratchbuilt Designs use that approach on Instagram. I have seen on youtube also that metal epoxy called Weican A was used, CHIPLOAD lathe build.

Skickat från min SM-A528B via Tapatalk

[ardenum2]

Whoch DWH from diamant is recommended? Myself I am planning to use R&G epoxy mixed with metal powder, I saw Scratchbuilt Designs use that approach on Instagram. I have seen on youtube also that metal epoxy called Weican A was used, CHIPLOAD lathe build.

There are 2 products, DWH for non moving parts like rail surfaces, moglice for parts that slide, like ways in grinding centers, metal to metal contact. Here's a guy that used it on his 3 ton machine: He then went onto building a business with the machine and now makes smaller ones on it, like the one I linked to on page 5.

https://www.youtube.com/watch?v=xtFzb6r6Np8

[peteeng]

Hi - If you machine surfaces correctly then no epoxy shimming is required. To create/build a true surface using epoxy putty is difficult as you need a true surface to cast from. No point in "building" surfaces and use high spec ballscrews. Have them correctly machined. If you can't afford that then back down to C7 and then the poor surfaces don't matter as much. Quality machines and quality parts are not cheap....

Work through your budget, get budget quotes from the people you are going to work with and see how it all comes together. Once you have a true picture of the $$$ and the sticking points then you can assess the situation. Keep the exercise on paper, do not buy anything until you have resolved the design and the $$$. Everything is connected, change something and it can flow thru the whole machine. Sus your list is too short and not enough detail to estimate true cost, as you say its rough so keep at it... There are 100's of hours in designing a good machine from scratch...Peter

[ardenum2]

Hi - If you machine surfaces correctly then no epoxy shimming is required. To create/build a true surface using epoxy putty is difficult as you need a true surface to cast from.

you guys don't have these?



Flatness down to 7um for 500mm($400), 12um for 1000mm($600), compared to $5000 for machining...

[peteeng]

Hi Ard - Until the quote is in writing I don't know if $5000 is accurate. Plus you may get one side "flat" but is it parallel to the other side? Have to look at all angles on these sort of things. Peter

[NordicCnc]

Hi Ard - Until the quote is in writing I don't know if $5000 is accurate. Plus you may get one side "flat" but is it parallel to the other side? Have to look at all angles on these sort of things. Peter

To achieve that they use a ground rod on one side of the component, which they rest the straight edge on. Then they put a spirit level on top of the straight edge, and adjust intil level with a adjustin screw attachment or shims in the other end. You achieve very precise result like this.

Looking from the other direction how do you ensure parallellism etc by machining? If you machine there may be tension in the material causin warping and what not. Also the machine may be not so precise, and if it is rly precise, it will cost you. I had my alu profiles milled by a local company and they had both warping and bending afterwards. It was really poor result, but that may have been due to several reasons. My point is that by casting and having the right tools, you dont copy the existing machine but you can adjust "until happy" and as good as you can measure with your equipment.

Thanks ard for letting me know, apparently it is the same guy that sells those FS3/4MG machines. Same approach is used on those and I am planning to do the same. Seems like I need to put in an order for the epoxy then to make some tests. I should get my DIN874/0 1500mm 100x40mm assembly straight edge in the mail this week.

Skickat från min SM-A528B via Tapatalk

[ardenum2]

To achieve that they use a ground rod on one side of the component, which they rest the straight edge on. Then they put a spirit level on top of the straight edge, and adjust intil level with a adjustin screw attachment or shims in the other end. You achieve very precise result like this.

Looking from the other direction how do you ensure parallellism etc by machining? If you machine there may be tension in the material causin warping and what not. Also the machine may be not so precise, and if it is rly precise, it will cost you. I had my alu profiles milled by a local company and they had both warping and bending afterwards. It was really poor result, but that may have been due to several reasons. My point is that by casting and having the right tools, you dont copy the existing machine but you can adjust "until happy" and as good as you can measure with your equipment.

Thanks ard for letting me know, apparently it is the same guy that sells those FS3/4MG machines. Same approach is used on those and I am planning to do the same. Seems like I need to put in an order for the epoxy then to make some tests. I should get my DIN874/0 1500mm 100x40mm assembly straight edge in the mail this week.

Skickat från min SM-A528B via Tapatalk

for cold casting inserts you'd use a steel piece that's been stress relieved, you'd have to ask the shop about it. I don't think they stress relief extrusions which is probably why it deformed after machining.

If you have it machined at a company the result depends on the machine they used. I've seen many builds where even after machining the result was poor and not to spec as ordered, in the end the person had to use steel epoxy. To ensure proper surfaces you'd probably need to pay for grinding too.

Think it's better to just get the DWH instead of trying to reinvent it, it's not that expensive anyway. It's a ready made product and it's been proved and tested, you're unlikely to match it's performance if you try to make it yourself.

[suspension]

What are your current parameters and I'll run a one page calc for you. Are you using steppers or servos? What torque rating are they? Have you used the spreadsheet I posted? Peter

I tried, but not entirely sure what values I must put for certain input fields such as acceleration and even velocity. My current CNC is a small one and the values I have used there may not work for the mill. Better to get them from some one who has more experience with a mill of this size.

I really appreciate you help as I am trying to use ballscrews available from the ebay seller to reduce cost.
My machine params are:

Motors - Serve, 750W, if needed 1000 is also ok. Rated torque - 2.39, max torque - 7.16, rated speed 3000, max speed 5000. For Z I will use braking.
Travels -> 600 X 300 X 400
Spindle -> BT30 cartridge. 12000 RPM. Driven by servo or AC motor at 12000 RPM max.
Z Counter weight -> Air cylinder
Linear rails -> HIWIN. If possible 15mm for X and 25mm for Y, Z. Otherwise all 25mm.
Ballscrew accuracy -> C5, medium pre-loaded.
Frame epoxy granite.
X -> Not sure yet. If possible epoxy granite, otherwise welded steel I guess.
I want to direct drive ballscrews from motors.

I hope above prams are enough?

Thanks
Sus