Torque settings for HSMAdvisor

[Kitwn]

It isn't rerally CAD & CAM but I couldn't see where else this should fit:

I have HSMAdvisor for calculating feeds and speeds. Among the parameters it wants to know about the machine are max speed, max power and max torque. I have one of the ever popular HY 2.2KW water cooled spindles and matching VFD.

Basic maths says that if it can produce 2.2KW at 24000 rpm then the torque is 0.88Nm and at 12000 rpm it's 1.75Nm. But everyone says the torque falls off at lower speeds so is there a well used and trusted table of speed v torque for these spindles? HSMA can be given a table of torque v speed but only if you have one! I have the VFD set up in accordance with the advice given elswhere on the forum.

I currently have the max torque in HMSA set to 0.85Nm for safety but I'd like to be able to give more accurate values at lower speeds for use with larger diameter tools such as skimming hardwoods down to a given thickness with a 19mm end mill for example.

Kit

[Neale]

Kit - I think the problem with your sums is that you have assumed "constant power" with speed change. I doubt that this is true. Like internal combustion engines, torque tends to rise with speed and as power is the product of the two, max power occurs somewhere towards top RPM/top torque. Electric motors are different, especially with the fancy sensorless vector controls and what have you (and I really don't understand exactly how that works!) but I suspect that the truth is closer to "constant torque" across the usable speed range and then falls. Certainly mine drops off sharply below around 7K RPM, based on observation (i.e. not accurate measurement).

[Kitwn]

Thanks Neale,
That was my concern. If the 2.2KW power is constant over the speed range then the basic maths applies but it's not going to be that simple in fact. As you say, IC engines have a rising torque curve with speed whereas steam engines have maximum torque at zero revs. It all depends on the technology.

At the moment my fixed figure of 0.85Nm applies at all speeds but I'd be interested to know if anyone has some more detailed knowledge of how these beasts behave in practice.

[m_c]

In basic electric motor terms, they are constant torque sources, so the spindle only produces 2.2KW at the rated speed.
Half the speed, you half the power.

You can use methods to boost the torque at lower speeds, but you risk overheating the windings, and there is a point at which cogging becomes an issue, as AC motors rely on inertia to a certain extent to smooth out the acceleration/deceleration between phases.

[Muzzer]

The torque on an induction machine is simply proportional to the difference between the speed of rotation of the magnetic field and the speed of rotation of the rotor. That applies at any speed - positive, negative, zero or anywhere in between. The rotor doesn't care or know any different. As m_c says, there's a thermal issue if you apply high torque at low speed for any length of time, so machines intended for that sort of use tend to have separate cooling fans or water jackets.

The faster the motor spins, the greater the back emf is. But torque is proportional to the current in the stator. To drive current into the stator, you need a voltage difference but once the back emf reaches the DC bus voltage within the VFD, you run out of that driving voltage. This is the base speed and without clever control (field weakening), you can't go any faster. It's similar to when you connect an unloaded brushed motor across a battery - it reaches a speed where the back emf equals the battery voltage.

The back emf voltage constant is Kv and is proportional to the magnetic field strength, so field weakening extends the speed range by suppressing the magnetic field generated by the rotor in an induction machine and hence reducing Kv. In a PM machine, it does this by creating a field to partly cancel out the field from the magnets in the rotor. You don't get that in normal VFDs but it's used in EV drives to give an extended speed range.

If you were talking about permanent magnet machines such as servo motors, the same principles apply in terms of torque being proportional to current and back emf being proportional to speed but now the torque generated is proportional to the positional angle between the rotating field and the rotor. Without a closed loop position control, a permanent magnet machine driven with a static field feels like a torsion spring as you try to move away from the no load position. You see a similar angle error / torque relationship with open loop steppers. If you have a stationary field in an induction machine, there is no torque at zero speed but it increases as you turn the shaft faster. It's like stirring a bucket of treacle - rather weird.

[Kitwn]

It sounds as if what you REALLY want is an indication of the actual speed of rotation of the spindle to compare with the speed indicated on the VFD. Muzzer, you've reminded me of my old induction motor principles training. If I remember rightly a standard induction motor is designed to run at about 95% of the field rotation speed at it's rated power. Aren't these spindles permanent magnet rotors?

m_c, Sounds as if the fixed torque value in HSMA is not unreasonable but I should be setting the VFD to display motor current rather than rpm.

Thanks for the help.

Kit

[Muzzer]

If you are driving these spindles with conventional VFDs, they are almost certainly induction machines. While you can run PM machines sensorless without a position signal, you generally require an encoder if you are using a std VFD. That usually requires an expansion board, assuming your VFD supports it.

The slip frequency is easily deduced if you look at the motor plate. The rated power is usually specified with a rotor speed. With 50Hz, the unloaded speed for an IM would be 1500rpm (4 pole) or 3000rpm (2 pole) in the UK and 1800rpm or 3600rpm in the US. Typically you'll see ~1440rpm or so at rated load for a 1500rpm 50Hz machine, which is 96% of synchronous speed. For a PM, the speed will always be 1500rpm etc unless you exceed the breakover torque and it stops being a motor.

If you set the VFD display to show current, you will get an idea of the torque. However, there may be a torque signal available. I see from the manual for my Yaskawa V1000 that I could display the actual torque as a % of the motor rated torque. I've just changed mine to that setting, rather than show the spindle speed, which is already displayed at the controller. Some of the programs such as HSMA / FSWizard, Sandvik Machining Calculator etc give you an estimate of the required spindle power. I wish Fusion did this, even approximately, so I could dial up the load on the tool when doing the CAM to make best use of the motor. I have 3.3kW available apparently but it seems I rarely get close to using that when I have bothered to get an estimate of the spindle power required. I wonder how close to the limit people tend to push their machines?

[Kitwn]

Muzzer,
The spindle has 24000rpm 400Hz on it and has no sensor wires. Having looked on YouTube at a video of one being dissmantled the rotor does look like it's made from laminations so it is probably an induction design as you say.

I use HSMA and do take note of the estimated power. As most of my cutting is done with tools of 6mm diameter or less it's the deflection warning that goes red before the power climbs very high. I've only been getting torque warnings with a 19mm tool I want to use for 'thicknessing' pieces of hardwood but was interested to get the benefit of other peoples experience and knowledge.

It seems like sticking with my fixed torque setting of .85Nm is going to be sensible. I've also set the minimum speed allowed in HSAM to 6000rpm which I've read elswhere on the forum is a speed below which the torque really drops off.

Thanks for all your input.

Kit

[Muzzer]

I'd forgotten about HSMA / FSWizard, although I can't quite bring myself to cough up for the paid version. Perhaps I'm being a tightwad there but the free browser and iOS/Android versions do most of what I need. https://app.fswizard.com/#

[Kitwn]

I'd forgotten about HSMA / FSWizard, although I can't quite bring myself to cough up for the paid version. Perhaps I'm being a tightwad there but the free browser and iOS/Android versions do most of what I need. https://app.fswizard.com/#

It's not like me to spend hard-earned dosh when not required but I decided HSMA would probably pay for itself in the long run. I've been experimenting with 'High Speed Machining' techniques lately using a nice plug-in for CamBam which works a bit like adaptive clearing in F360. HSMA has been very informative in comparing the effects of cutting depth and tool engagement. I plan to put some info on the forum once I find time to do proper tests and come up with something useful to say.

[Muzzer]

True HSM is supposed to involve some truly frantic speeds and feeds, where the spindle power actually comes down from the normal level, for the same MRR. It's probably also a fire hazard! What sort of numbers and materials are you trying out?

[m_c]

I always think HSM isn't really the right term, but it does seem to be what's used to describe trochoidal tool paths.

Actual cutting speeds aren't really any higher than any other machining methods, as you're still limited by how hard you can push the cutter/machine/workpiece, but it does benefit from high acceleration/rapids, due to the continual direction changes and amount of time the cutter is moving between cuts.

I personally find it quite therapeutic watching mills use a trochoidal path to mill out big sections, seeing big long flakes of metal come off, instead of the olde going around in circles/squares/lines taking shallow depth cuts.


I do have a full license for HSMadvisor/FSwizard, and think it's a fantastic program, but I bought mine years ago when you could get 3 full lifetime licenses for $110.
I do see they've got a new 3 year 3 hp license option for $95, which is probably still a bit much for a hobby user.

Off course the other option is G-Wizard, but I never really got on with it, and the whole CNC Cookbook marketing nonsense/endless plugs/popups annoy me. Which is a shame, as they did have some good articles buried in amongst the nonsense.

[Kitwn]

Muzzer,
I've had very little time to play recently but the most punishing test to date is using a 6mm 2-flute end mill on Tasmanian Oak. Depth of cut 22mm, Stepover 0.25 (1.5mm), speed 1200rpm, feed 2500mm/min.
One of my main reasons for wanting to use this kind of cutting is for wooden gear wheels. Shallow depths of cut tend to chip the teeth where the grain is tangential to the disc of the wheel. Large depths of cut using the trochoidal plug-in for CamBam fixes thing. No smoke as yet!

m_c,
The trocho tool paths are quite mezmerising. I have the 'smooth chords' option set true which gives smoother movements between cuts which reduces jerks and accel/decel delays.

[Muzzer]

What I was referring to was machining at (very) high speeds, rather than trochoidal or adaptive toolpaths. It seems that if you keep winding up the surface speeds way above where we normally work, the input power starts to come down quite noticeably. Very few of our machines would be capable of operating in that region but I wondered if that was what Kitwn was talking about, as he has a router that presumably can operate at high spindle speeds. We seem to be talking surface speeds in the region of 1000m/min https://www.ctemag.com/news/articles/what-high-speed-machining

m_c, as you say, the use of "HSM" is very fuzzy. Companies like Autodesk / HSMWorks merely(!) mean adaptive or trochoidal. Even so, they have been a revelation, particularly when applied to older machines which can now achieve impressive MRRs the manufacturers never would have expected to see.

I might be wrong but my understanding is that the adaptive clearing algorithm was devised by a couple of guys in Liverpool. They seem to have missed out on the rewards from such a widely adapted and successful concept. https://www.freesteel.co.uk/wpblog/frontpage/

[Kitwn]

I hadn't come across the high surface speed version of 'HSM' only the adaptive clearing of Fusion 360 and the similar looking trochoidal plug-in for CamBam.

It's the ability to use high depth (that sounds like it should be a contradiction!) of cut that is most appealing, partly because of the gear-cutting issue mentioned above and also because of the increased tool life to be expected. Shorter cutting times are also a bonus as it reduces the time for which I'm producing noise in a residential street.

My spindle is the ever-popular Huan Yang 2.2KW water-cooled beastie capable of 24000 rpm.

CamBam users may not be aware of the range of plugins available from the following link. The main site is in French but the plug-ins page is in English. The list of free goodies available is amazing and greatly extends the basic program to include features you would otherwise have to buy other, much more expensive software to obtain. V-Engrave, Arc Text, Text on Polyline and Trochoidal Pocket being just some of the ones I've already used. There is also a very active support community for all the plugins on the relevant forum page.

http://www.atelier-des-fougeres.fr/Cambam/Aide/Plugins/plugin_menu.html