High Speed Motor Development/modification 1

[NIDinc]

We currently manufacture a specialized, compact 2 hp, 230 VAC three phase induction motor. It is designed for variable speeds up to 250 Hz, or about 15,000 rpm. It runs at about 2.5 amps, no load at this speed and requires only about 15psi air to operate under a working load of about 5.5 amps (full load amps: 6.4).

We modified the motor to run at about 24,000 rpm by changing the windings and wire appropriately and increasing the Hz to 400. Mechanically it can handle the speed.
The only problem seems to be that the efficiency has dropped off dramatically and the motor is generating too much heat at the windings and rotor.

We're not motor engineers and the consultant who designed our original motor years ago is no longer available. Any thoughts, tips and input for ways to improve the efficiency and lower the temps is much appreciated.
Thanks, JS

 

[ScottyUK]

Not a high speed machine designer, but a few initial thoughts / questions to get us going:

What is the construction of the rotor core - solid or laminated, silicon steel or something else, etc?
How hot is 'hot'? A Class H winding can run blistering hot more-or-less indefinitely.
Did you account for increases in friction and windage losses as the speed goes up?
How did you arrive at 'appropriate' changes to the winding design and in what way did it change?

 

[waross]

Try an unmodified motor at 400 Hz and 370 Volts to 400 Volts.
Let us know what that does to the losses.
Are you using a filter between the drive and the motor?
If the motor is internally Wye connected a change to a Delta connection will make it suitable for up to 26000 RPM at 230 Volts.
Your motor wants to see a V/Hz ratio of 230Hz/250Volts or 1.09 Volts per Hertz. It may be easier to change the voltage than to change the windings.
Alternately, cut each winding in the center into two equal windings and then place the windings in parallel. This will give a motor capable electrically of 30,000 RPM at 230 Volts. The HP at 24000 will drop to 80%.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[zlatkodo]

It is not clear what modifications have been made to winding?
- The number of poles is the same.
- Probably it is about a simple reducing the number of turns per coil in Hz-ratio ie. to 250/400 = 62,5 %. It can not be done just like that.
It is much more complex than it seems at first glance.

Motor Repair and Design

 

[NIDinc]

Thanks for the helpful replies. Yes, for modifications the turns per coil were reduced and the wire size increased. The rotor is laminated, silicon steel with double bars. FYI, the stator laminations are also bonded rather than welded for improved efficiency.

I like the idea of trying the unmodified motor with the increased voltage at 400 hz. The new application we are seeking does not require variable speeds either.

Here's what was done to the windings to modify this as mentioned above...


Original Motor Alteration New Motor
15,000 RPM 25,000 RPM

50 Turns/Coil 15/25 x Previous 30 Turns/Coil

1 Pair 22AWG 25/15 x Previous 17 AWG wire
in parallel, with with approx.
.001 sq inch in .00161 sq inch
Copper section in copper cross
Per turn Section Per Turn

 

[electricpete]

and requires only about 15psi air to operate under a working load of about 5.5 amps

can you explain the role of this air?

=====================================
(2B)+(2B)' ?

 

[NIDinc]

The role of the air is simply for cooling. The motor is a cylinder shape that fits into a housing with no possibility of fan cooling. Thus the air.

 

[NIDinc]

Zlatkodo, you wrote, "- Probably it is about a simple reducing the number of turns per coil in Hz-ratio ie. to 250/400 = 62,5 %. It can not be done just like that.
It is much more complex than it seems at first glance."
Can you elaborate?

What do you think of Waross' ideas of adjusting the voltage/hz. to increase rpms?

Thanks,
JS

 

[electricpete]

The role of the air is simply for cooling.

ok, so that is one factor to note as a difference from changing speed on traditional self-cooled machine:
[ul]
[li]For self-cooled machine increasing speed results in increased cooling air flow.[/li]
[li]You don't get that benefit if you don't change your externally-supplied air.[/li]
[/ul]

On the other hand, an adjustment to this air supply pressure or other changes to cooling configuration might be something to think about along with the ways you're looking to decrease the heat generation.


=====================================
(2B)+(2B)' ?

 

[waross]

A note on the attempt to rewind the motor;
This is an induction motor and induction and inductive reactance plays a big part in the design.
The induction of a coil is largely dependant on the SQUARE of the turns.
The relationship is not a simple linear ratio.
In a motor there may be further complications.
HOWEVER.
Dual voltage motors are commonly wound with two identical coils per phase. These coils may be connected in parallel for operation on a lower voltage or connected in series for operation at twice the lower voltage.
Consider rewinding your motor as before but breaking each winding into two identical windings. For example, instead of 30 turns per coil, wind two coils with 15 turns each. They may then be connected in series for the full 30 turns or connected in parallel for operation at 115 Volts or for operation at 230 Volts and 200% frequency.
You may consider rewinding the motor for use as a dual voltage motor. You would then be able to connect the motor for 115% Volts. This would allow you to maintain the V/Hz ratio and increase the voltage and frequency together. You would be able to apply double frequency at 230 Volts and get full torque at twice the speed, or double the HP. Losses will probably increase and you may not get the full 200% HP, but you should get most of it.

zlatkodo may be willing to advise as to how to physically split the winding into two parts so as to allow all the poles to remain magnetized.

Another option depends on the present internal connections. If the motor is presently connected internally in a wye configuration, a reconnection as a delta will allow the voltage to be increased to 173% of 230 Volts or 398 Volts.
Either of these solutions may be preferable to requiring your customers to supply a higher voltage or preferable to supplying a step-up transformer.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[electricpete]

The relationship is not a simple linear ratio.

Sorry if I'm misunderstanding but I think it is fairly close to a simple linear relationship in terms of maintaining airgap flux density.
V = N * d/dt(Phi) = N * B * A * f
where N is turns
If f changes by 40/25 and N changes by 25/40 while V and A are constant, then then fundamental airgap flux density B stays is roughly the same. As I understand there is no change in poles, coil span, coil interconnection etc so it should be more straightforward from electrical standpoint than most redesigns.

There may be some change in reactances. Certainly X2 increases with the increase in frequency. X1 and Xm a little tougher to judge considering both effects of frequency and turns. A designer should be able to estimate these parameters and predict resultling changes in characteristics (torque speed characteristic, starting current, etc). In general the reactances don't play a huge role in the flux densities or running efficiency.

From a distance it seems like a reasonable redisgn attempt to me. Feel free to point out if I'm misunderstanding or misrepresenting.

======

op - We don't know whether something went wrong with the redesign (I don't see any indication) vs whether the expectations were simply unrealistic for performance after redesign. While the redesign may have attempted to control stator copper losses, we know there will be increase in friction/windage, an increase in stator core hysteresis losses (~f) an increase in stator core eddy losses (~f^2) and probably an increase in stray losses. But no increase in cooling. We on the forum don't really know anything about before and after performance from what you gave. We have before currents but no after. There is mention change of efficiency but no numbers and no mention of the associated loading level. Did you run to the same current, or the same mechanical load, or maybe 40/25 higher load based on constant torque, or did you just let the driven equipment find its own load level at new speed (we don't know what this drives). Did you happen to measure slip before/after. Sorry if this came across as judgemental but it doesn't seem there is much to work with. EDIT: I'm not guaranteeing we can provide a performance analysis if you provide all these details but we can't begin to judge performance without these details.

=====================================
(2B)+(2B)' ?

 

[zlatkodo]

We currently manufacture a specialized, compact 2 hp, 230 VAC three phase induction motor. It is designed for variable speeds up to 250 Hz, or about 15,000 rpm. It runs at about 2.5 amps, no load at this speed and requires only about 15psi air to operate under a working load of about 5.5 amps (full load amps: 6.4)

But, we know nothing about:
- Amps after modification,
- kind of motor application,
- impact of increased speed to load ie. whether the increase of speed will cause the increase of load ?
- internal winding connection,
- external connection ( wye or delta)...