machining aluminium with 1.5kW spindle

[routercnc]

I've been machining more aluminium recently and noticed the limitations and problems associated with using a high speed spindle at lower rpms. It's a typical 1.5kW water cooled chinese spindle.

The spindle is rated up to 24000 rpm and ploughs through wood with no problems at these speeds. But if you ever try running at this speed for aluminium using a 2 flute carbide bit it just heats up the cutter and the aluminium sticks to the tool, often meaning pausing/rewinding the job to clear it.

The guide figures for cutting aluminium with a 6mm cutter shows 4000rpm, but dropping the speed this low just stalls the motor even with a 0.5mm DOC and feeds of 300-600mm/min. So I'm forced to run at 10,000 - 12,000rpm to avoid stalling. Coolant helps somewhat but there is still the occasional aluminium buildup on the tool, most likely due to the excessive rpm.

I'm getting by for now but it feels like I'm on the edge of successful machining.

So I've started to research other options such as:
1. Reading the VFD manual for torque boost (need to check it has this and suspect it will heat the spindle)

2. Making a seperate pulley driven slave spindle (e.g. ER20 collet and straight shaft) geared down to run at 3:1 or better from the existing spindle motor, therefore giving loads of torque but lower speed. Also would mean I could fit a wider range of cutters (ER11 currently), possibly even fly cut if I could reduce the ratio still further.

3. Upgrading to a 2.2kW spindle (more money!). Anyone have a 2.2kW spindle torque vs speed graph compared to 1.5kW? I notice lots of people successfully use the 2.2kW spindle for aluminium - are you able to run at the correct rpms for aluminium or are you also slightly over speeding?

Any comments on the above welcome . . .

[Detah]

I haven't tried milling aluminium yet but plan to very soon. I'll be very interested to see what you discover.

[EddyCurrent]

You could try a sensorless vector VFD that should give full torque at zero speed but people, myself included, seem to have problems getting them to work with high speed spindles.
AC Inverter Drives (230V) filtered by Input: 230Vac 3ph, Power: 1.5kW (Page 1 of 3) (http://www.inverterdrive.com/group/AC-Inverter-Drives-230V/?filter=Input|230Vac+3ph&filter=Power|1.5kW)

[routercnc]

Thanks Eddy. I'll look into that but I would need something which I knew definitely worked with a spindle before purchasing. Also if I ended up spending that much I probably go for a 2.2kW spindle + new VFD instead which is not too much more money.

[CharlesJenkinson]

The 'geared' belt driven idea sounds juicy. I thought you might get a few more replies on this, regarding how others do it successfully. There's not been any definitive answers yet, or at least a bigger sample of a variety of solutions that work for different people. It may be that there are so many parameters that need controlling or considering, and you just have the odd one not in your favour, but I hope you don't have to resort to too much empiricalism to sort it.

[JAZZCNC]

Can't help with torque guides etc but I can tell you that 2.2Kw cuts ali fine with speeds down to 5000rpm thou to be honest I mostly cut between 8-12K depending on cutter size.

Now if your getting Sticking and melting chips then I suspect your cutting to shallow so getting very thin chips which heat up quickly and melt. Try cutting deeper and you may well be pleasantly surprised what depth you can cut at. Set the VFD to amps and watch the load this will give you an idea of how the spindles coping with the depth.!

[routercnc]

Thanks for that Jazz. Gives me some hope for experimentation with the existing spindle. The machine is still under final construction so when I am set up properly I'll include higher DOC in the runs. I've only done a bit of quick machining to help build the last bits of the bed so used the 0.5mm DOC from the previous much weaker machine.

I'll also dig out the VFD manual and set up the display for amps. Seem to remember this being an 8A VFD but it was a while back so I'll confirm so I can see how it is coping.

[routercnc]

Hi Charles,

I think the biggest factor is not buying the 2.2kW spindle to start with! It's not much more than the 1.5kW spindle, takes the much larger ER20 collet, and although I suspect it is a little heavier, I can't see any real world downsides.

[Boyan Silyavski]

Generally speaking the best speed for aluminum is 8000 rpm. Thats what my tests show. 25 ipm feedrate is good feedrate to start with. No coolant or soap water or sprayed alcohol via airbrush.

I believe the problem is in the VFD, as with 0.8kw spindle i have no problem at that depth and feedrate / my machine is not strong/.

Torque boost is for very low speeds. The VFD should be set to Constant Torque.

Single flute bits work best if speeds are problem.

[JAZZCNC]

Generally speaking the best speed for aluminum is 8000 rpm. Thats what my tests show. 25 ipm feedrate is good feedrate to start with. No coolant or soap water or sprayed alcohol via airbrush.

That's far too general and you can't say "Best Speed" because there is no "one" best speed for cutting aluminium there are far to many factors come into play which determine the speed.!! . . . . Material grade, cutter size, No flutes, DOC, Cut type ie: slot or side cutting, step over, coolant type . . . .And thats before you get into Spindle power and machine strength.

Can tell thou 100% There is no ONE speed does all and you'll fluctuate between 7-15K when cutting ALi with 4-10mm tooling with varying no of flutes and DOC & S/F etc.

Experiment is the best way to learn speeds & feeds and dont be affraid to cut deep, just be prepared and accept that you will break tools while finding the best cutting conditions for your machine and material.

[Boyan Silyavski]

Thats what i meant, in general. I remember doing a lot of experiments and always there was some smallish problem. And a lot of broken bits and melted aluminum :-). Then i read somewhere, /i don't know where exactly, but it was a person doing alu jobs everyday/ that speed 8000RPM is the key. And so it was, at least for me.
Again as you say- quite generally speaking.

[EddyCurrent]

1. Reading the VFD manual for torque boost (need to check it has this and suspect it will heat the spindle)

This is not applicable to your issue, Torque Boost is to produce initial torque from standstill, for example to get a heavy load moving, it usually cuts out at low frequency or when the speed you selected has been reached.

[routercnc]

Thanks Eddy. I misunderstood what that was for and this afternoon (before reading your reply) I experimented with boosting the voltage by 10% in the region of 4500rpm [default was 5% at 720rpm] and also had the DOC set to about 1.5mm. It happily cut a slot 50mm long (6mm 2F Carbide bit) without problems.

Now I'm confused. If the boost is momentary until the speed is reached it wouldn't have affected the torque delivered during the cut since I allowed the tool to reach the required speed before starting the cut. Perhaps everything was cold (tool and workpiece) and I would need to machine for longer to see if this was actually better.

Anyway before I could experiment further I had to take the machine out of action to send one of the ballscrews away for straightening so I'll have another play when it returns.

Thanks for the advice - I don't think things are so drastic as I first thought.

[Jonathan]

The low speeds you're suggesting imply you're trying to use HSS tools, but from latter posts maybe not? If so switch to carbide, then you can use around 12-13krpm with a 6mm cutter. A splash of something wet can make a huge difference...



You could try a sensorless vector VFD that should give full torque at zero speed but people, myself included, seem to have problems getting them to work with high speed spindles.


A vector drive will get a bit higher torque at low speeds, but we want more than that. In general we're still after a high power at low speed, so even when you have rated torque available across the full speed range, power is still torque times angular velocity, so the power will be decrease proportionately with speed. e.g. If you have a motor rated for 3000rpm and 3kw, then with a vector drive the best you can do without exceeding the ratings is get the same torque at say 300rpm, so you'd only have 0.3kW available at this speed. With a V/f drive you'd get a bit less than 0.3kW, but in the whole scheme of things that's not really a significant difference compared to the difference in output power compared to rated. It's only at very low speeds (like a few Hz) that the difference in torque output is very large.

I expect the issue with getting the vector drives to work with these spindles is that the spindle motor's parameters (time constants etc) are going to be quite different from your average 'low speed' induction motor. To control an induction motor with vector control, a model of the motor is used within the drive to calculate the rotor flux orientation, which is necessary to ensure constant torque and stable operation. This model requires certain parameters, so the manufacturers tend to ship the VFDs with the PI constants for the current control set to values which work for typical motors within the power rating of the drive, as the parameters over a small range of motor power ratings usually don't vary by much and you could damage the drive if these were set incorrectly.


Having said that, there is one thing you can do to get more torque from the spindle at low speed, if you're feeling adventurous. I accept no responsibility if you try it...I posted it a long time ago and nobody noticed, maybe for the better:



On a related topic, if you've got one of the Chinese water cooled spindles it's possible to get a bit more torque out of it at low speed. All you have to do is alter the V/f curve parameters in the VFD to raise the voltage slightly at low speed. This increases the flux and thus the torque output available, however you can only gain a little due to saturation. The no-load losses at low speed will increase dramatically (e.g. 10x more) so only raise the voltage a little bit at a time, and keep an eye on the no load current. Clearly this means your water cooling system must be good. You can gain a useful amount of torque from doing this, but do be very careful as it's probably an easy way to burn out the spindle if you're not careful.


I've left my VFD set that way and it doesn't seem to have caused problems. You can get away with it to an extent since a lot of the extra heat is generated on the rotor and although that is poorly cooled, it's essentially just a few lumps of metal so it doesn't matter so much (compared to the stator) if it gets very hot. There rotor resistance is a function of temperature and the former is often inferred as part of the vector control algorithm anyway (to maintain accurate rotor flux orientation), so I intend to use the vector drive I've made to do some further testing in this area to more accurately determine the temperature (as I clearly have full control over my drive's control structure and parameters).



3. Upgrading to a 2.2kW spindle (more money!). Anyone have a 2.2kW spindle torque vs speed graph compared to 1.5kW? I notice lots of people successfully use the 2.2kW spindle for aluminium - are you able to run at the correct rpms for aluminium or are you also slightly over speeding?


Multiply the torque-speed graph for the 1.5kW spindle by 2.2/1.5=1.47 and you'll be pretty close.