sizing servos for a cnc lathe

I need the spindle to be used as a C axis, I plan to add live tooling in the near future. It is a DC servo however. It is a slant bed lathe with a 6 station turret with 2 rings of tools. I will be doing some measuring work tomorrow, my spindle also has a brake on it, handy for such endeavors as cutting splines, live tool milling etc. 

I think the newer DC drives on the existing servo motors is the best way forward. obviously this will require me to verify the electrical specifications of the dc servo motors, but pretty much any brushed DC motor can be a servo with sufficient cogging and a encoder. 

Sounds ambitious but should be pretty sweet when you get it sorted out.

FYI - there's a good thread under the "Turning" category where a member built a C-axis lathe from scratch.  I believe he managed to work out some of the issues surrounding switching between turning mode and C-axis mode for polar interpolation milling.  Might want to have a look.

I saw that, I am trying to find documentation on these drives, Yes they look over sized., 15.5kw, but better that than undersized. I think what I might do is use output analog out 4 for spindle and output 5 for c axis and just swap it with a relay if required
www.ebay.com/itm/165534824457?hash=item2...9:g:ShkAAOSwUuFWz0lf

YASKAWA SERVO DRIVE CPCR-MR152GD MR152 GD MR152G

Need manuals for those is anyone has one ^^^

I found a 30kw yaskawa 505 spindle drive to that is PERFECT for me, using the brake and switching to low gear, should work well

This thread lacks photos. Please post photos of the lathe. Everyone likes photos of lathe.

This is too much focus on rated power in the selection of the motors. That tends to be a pretty unimport secondary parameter, in that once you have sized the motor for the inertial loads, and necessary torque, the resulting motor will have some rated power (typically way way more powerful than you would expect).

I did a course on this at the TU, and summarised the process they teach here:


The expectation of a manufacturer would be for a very rigid, highly dynamic drive system. For your purpose, you are prepared to compromise in the machine dynamic performance, and also the cutting power, but when you run through the inertial calculations, you may find you need a pretty beefy AC motor, to get it to work at all, especially if it has a long,  large diameter Z ball screw, direct driven. On a lathe, the inertial load of the Z ball screw and motion of the saddle can be significant, with the sort of light cuts you are willing to accept maybe buried in uncertainty of those loads.

the machine already has a 5hp and a 7.5hp X and Z axis motor, I am just trying to match the those motors for power output. Cincinatti thought to put 10hp drives on these motors, kinda wild. I will post pics of lathe in the next day or so

Did a bunch of parts bin shopping

cpcr mr055   servo drive 5.5kw
cpcr mr084  servo drive   8.4kw
jusp 60B       DC power supply  17kw

and a replace DC drive for the spindle motor at 30kw, cdmr30kw 22kw continuous 30kw peak spindle drive

the reliance and cicinatti stuff is being used as a trade in, so it save some $. but I like the yaskawa stuff, it's available and really reliable. 
 

The rated power is not that relevant. You can get 7.5HP motors in a wide range of moment of inertia, and peak torque.  Those are what need matching to the axis.

1500rpm = x torque around 90nm otomh, to create 90nm of tq requires x watts of power. this means that they will have roughly the same geometery, physics of materials etc. the motors are already on the machine. I am replacing the drives. the motors are staying. I expect to have to do some tuning.

IF I was going to swap out for smaller AC motor with less HP, the rpm rating of the motor and it's power characteristics come into play.

HP = tq x rpm/5250. I care about Torque, that's the bottom line,

It is possible for 5hp rated motors to have vastly different TQ output, and it doesn't really matter because you can gear ratio your way back to the same shaft Torque values. so long as the operating range of the lower tq motor is equal to after reduction the large higher tq motor.

for instance a 5hp 1500 rpm motor will have nearly double the TQ of a 5 hp 3000rpm motor at any specific RPM

Now, what might blow your noodle, a 3hp DC motor with a narrow power band of 3000 rpm, can have the same shaft Torque after reduction as a 6000rpm AC motor with half the power running at 2x the rpm. the apparent load at the motor shaft will be virtually identical.

I don't think I would be to terribly concerned with rotor inertia characteristic's in a significant way unless that rotor was massively oversize for a given application. a really large diameter pancake motor would be one such motor. The load the rotor is reacting to and against has a significantly greater mass than the motor rotor. IE the bed carriage and turret have multiples more mass than the rotor ever will, and those loads are inertial in a linear way, which means they are reacting in a angular fashion against the lead screw relative to the rotor, and this means that in a mass kicking contest, the linear load is going to massively overcome the rotational load every time. reciprocating inertia will almost always be higher than polar rotational inertia at transients. IE connecting rods loads are WAY high than crankshaft loads in a combustion engine.

now, where things get dicey, accuracy of motion can and will be effected by the motor output and it's power curve, what this means is that I will have to do tuning on my control pid, acceleration rates etc to ensure that I get highly accurate following. I was going to be stuck doing that anyway, so no great loss there.

smc.collins wrote:

I don't think I would be to terribly concerned with rotor inertia characteristic's in a significant way unless that rotor was massively oversize for a given application.

As I understand it, it's the other way around: you don't want the inertia ratio (load inertia/rotor inertia) to be too large. E.g. Omron lists maximum load inertia for their servos. So does Yaskawa.

smc.collins wrote:

The load the rotor is reacting to and against has a significantly greater mass than the motor rotor. IE the bed carriage and turret have multiples more mass than the rotor ever will, and those loads are inertial in a linear way, which means they are reacting in a angular fashion against the lead screw relative to the rotor, and this means that in a mass kicking contest, the linear load is going to massively overcome the rotational load every time.

Have you actually done the calculations on this? In my (admittedly very limited) experience the rotational inertia from lead screws and/or gearing tends to be larger than the linear inertia from the driven part. (Connecting rods in an ICE act on the crankshaft via a moment arm which is quite different from how the linear axial load affects a ball screw.)

I posted a spreadsheet that calculates inertias here . Feel free to try it out if you want to. (And please notify me of any errors in the spreadsheet.)

unless rotor speeds are extremely high, 10k rpm, I am going to go ahead and say that the reacting load in the carriage, is drastically higher, calculate the deceleration G's and acting mass from that in a direction change. Rotors inertia is a fraction of it. The carriage on my lathe weighs nearly 4,000 lbs, that tiny little rotor in that 7.5 hp motor has nothing in way of mass compared to accel/decel of 2 tons of carriage at say 4-5gs, that's we we have to use gear reduction of 5:1 with the lead screw just to attempt to control it.

You may very well be right, and you're certainly right in that if you use a 5:1 reduction then rotor inertia won't be a problem since inertia is transformed as the square of the reduction - so load inertia as seen by the motor will be reduced 25 times in that case. (Well, minus the inertia of the gearing itself.)

Just out of curiosity, what's your ballscrew lead, diameter and length? And what's your target top velocity for the axis?