Air gap and starting current 3

[eemotor]

I recently ran into an issue of how does the air gap of a 3 phase induction motor relate to the motor's starting current. Even though many other performance characteristics of the motor change as well, the starting/locked rotor current was my focus. It appears that with a larger air gap the starting current increases. What is your take on this?

 

[electricpete]

Increasing airgap should:
1 - Decrease exciting reactance,
2 - Increase leakage reactance

These two effects work in opposite direction in terms of their effect on starting current.

My gut feel is #2 wins and starting current decreases.

The extreme case of increasing airgap might be energizing the stator with rotor completely removed. I don't recall what kind of currents are seen in this scenario but I think they are on the order of full load current... much less than starting current. This would also tend to suggest that starting current decreases.

Then again, I'm just talking off the top of my head. Might be way off base.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

The transformer analogy would be energizing a transformer with secondary short-circuited (and neglecting saturation effects for simplicity since they are not as severe on a motor).

If we are looking only at the linear sinusoidal steady state model as discussed above, it is clear that #2 wins (suggesting LRC decreases). The impedance for locked rotor current is L1 plus parallel combination of Xmag and L2+X2. Since Xmag is much larger L2 plays a larger rols and the effect of change in leakage reactances should be much larger.

I don't think saturation plays a big role in determining LRC so I think we can believe the linear model.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

L1 and L2 should be X1 and X2 (primary and secondary leakage reactances)

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[jraef]

It depends upon what you are actually looking at. By looking at energy efficient motor designs you can tell. One of the main design changes was to decrease the air gap to increase overall efficiency. The consequence of that was an increase in Inrush Current; true inrush current, not starting current. In other word, the magnetization current went up, that is why the NEC had to allow for 2000% instantaneous trip settings on circuit breakers feeding EE motors. "Starting current" however is another thing. In most cases, starting torque was actually reduced. So although starting CURRENT (after the inrush current) was reduced, starting ENERGY was actually slightly increased because the reduced torque meant a longer start-up time, and the total slip losses in starting were slightly increased. There is information on this at the DOE "Motor Challenge" website, or at least there used to be.

"Venditori de oleum-vipera non vigere excordis populi"

 

[electricpete]

First I repeat my comments above come just from trying to analyse things... not positive I'm right.

Some questions for jraef:

First looking only at LRC -you said energy efficient motors we see LRC decrease.....since gap decreased, as well, we might assume cause and effect relationship (increasing gap increases LRC). My two questions: #1 - t a fact that energy efficient motors have lower LRC? #2 - Aren't there quite a few other changes involved in making an energy efficient motor beyond adjusting the air gap which could affect LRC?

Second - regarding increased inrush on high-efficiency motors - My opinion is that from a linear model, the peak instantaneous current theoretically cannot exceed twice the peak of LRC. If the new motors behave differently than the old, I can only see two explanations: A - it may be either that the circuit L/R increases such that there is less decay before the first quarter-cycle peak.... resulting in peak instantaneous coming closer to the theoretical maximum of 2.0. B - It may be that there are saturation effects which increase current beyond that predicted by the linear model.

If the true explanation is A - I can't see that decreasing the airgap creates this effect... decreasing airgap would decrease X1 and X2 which would decrease L/R and make the current decay more before the first quarter cycle..... leading to the suspicion that there are other factors beyond airgap responsible for the changes (decreased L/R) that we see in high efficiency motors.

If the true explanation is B... the impact on inrush would be consistent with the expected impact of decreasing airgap. If you change nothing else in the motor, decreasing the airgap should move you closer to saturation in the iron which could result in increased inrush current that we see in high efficiency motors.

Regardless of the impact on this discussion (how does the airgap play into it), I would be interested to hear opinions on whether explanation A or B is the reason for perceived increased inrush of new energy efficient motors.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

I think the first few letters of my most important question got blocked out:

"1 - Is it a fact that energy efficient motors have lower LRC?"



=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

jraef - I'm pretty sure high efficiency motors generally have higher LRC.

Compare Reliance's 2-pole 460v, 405T frame, TEFC motors, identical except for minimum efficiency

high-efficiency

http://www.reliance.com/pdf/pdf/aced/W06386-A-A002.pdf

LRC = 695A

E-pact standard efficiency

http://www.reliance.com/pdf/pdf/aced/W06377-B-E001.pdf

LRC = 642A

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[jraef]

Electricpete,
To be quite honest, I am not a motor design expert, I was making an observation of some data I saw from the Motor Challenge program. My comment about reduced LRC was based upon the info they provided, warning users that EE motors often have reduced LRT, and I was ASSuming that the two conditions go hand-in-hand. My bad I suppose, not really good engineering practice, but I hope I have made it clear that this was an obsevation not an actual fact known to me. Now that Itakethe time to look at it again, I realize I was wrong. The LRT is in fact lower, but the LRC is actually higher! The info I was looking at does not appear to be published on the web, I have in in a binder, but this article seems to be based on the same study.

http://www.ecmweb.com/mag/electric_dont_let_new/

I totally agree BTW that there are several factors that go in to increasing the efficiency of the motors. The permeability of the steel used in the laminations is another one, so it is possible that some of these effects are not attributable to air gap. Again, my bad. The topic wasn't EE motors to start with, so I didn't want to go too far off on that tangent. Would that I could withdraw my statement, it was too vague and misleading. Mia culpa.



"Venditori de oleum-vipera non vigere excordis populi"

 

[electricpete]

Sorry, I didn't intend to make you feel the need to apologize. There is a lot more to sort out when we bring in the issue of the high efficiency motors and I have still never fully understood what the deal was. At this moment in time, it seems the LRC may be higher which in itself might explain the need to increase magnetic trip settings (and is also consistent with the theory that increasing gap causes decresing LRC), but I am under the impression I have heard folks say on this board that energy efficient motors behave differently with respect to the factor by which peak inrush exceeds LRC ... never quite figured out why that would be.

Any more comments on the original posters question?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[Marke]

Hi Jeff and Pete

The reduction in the airgap was only one of the changes to increase the efficiency of the motors.
The rotor bars were also changed to reduce the effective resistance of the rotor bars under low slip conditions. This has the effect of increasing LRC and reducing LRT, so I am not sure that the change in LRC and LRT would have necessarily been due to the change in airgap.

I would expect that an increase in airgap without changing the stator design would increase the magnetising current and the leakage reactance. This would inturn lead to a change depending on which was dominant as Pete suggested above. I don't know that there is a hard and fast rule that could be applied here.
A curious question, lets see what pops out!!

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[aolalde]

I ran my motor performance program for a motor that we rewound. 1250 HP, 3ph, 4160 V, Ifl 166.8 Amps, 893 rpm.

The rotor OD (outside diameter) is 23.622” and the stator bore diameter 23.750 for a radial air gap of 0.64”.

I ran the program for 125% (0.080”) and 150% (0.96’) air gaps (equivalent to machining the rotor OD) , some of the main motor parameters results are:

0.064 air-gap 0.080” air-gap 0.096” air-gap

Locked rotor Current (amps) 1163 1134 1109
Locked rotor torque (%) 182.8 172.5 163.6
Prim. Leakage react (pu) .077 .076 .075
Sec. Leakage react (pu) .165 .164 .163
No load amps 63.9 71.0 79.5
Full load amps 166.8 171.9 177.2
EFF @ FL (%) 95.41 95.44 95.43
P.F. @ FL (%) 81.24 78.82 76.49


These results do not mean necessarily the same behavior for different motors, since saturation, distance from the cage to the rotor bars, etc could be totally different.

 

[electricpete]

I'm jealous. Wish I had a program like that.

I would never have guessed leakage reactances go down with increasing gap. Certainly if I increase the e primary and secondary winding cyclinders in a tranformer the leaakge reactance increases. I guess motors are more complicated with their leakage... toothtop, zig-zag, harmonic... all that stuff I never understood.

Also I struggle to understand why IF leakage reactances decrease and magnetizing reactance goes down goes up (based on no-load current going up), THEN the program still predicts locked rotor current decrease.

I'm sure the program is right but I just don't understand it. If anyone can explain it, please do.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

One change is clear, the flux will be lower with increased airgap and the core will be further away from saturation. But I don't see how that leads to the other stuff.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[aolalde]

Epete :

The leakage reactance’s (specially the rotor or sec. leakage reactance) change with the slip and saturation.
I assume that the reactance values shown are between no load and the breakdown cusp of the speed torque curve, but not at stand still.

The magnetic flux per pole (Phi) is constant, because it is function of the line voltage, frequency, turns per coil, connection, span, etc, all in the stator winding.

Phi = E / (4.44 * f*N*Kd*Kp*Ks)
E = electromotive force per phase
f = frequency
N = series conductors per phase
Kd = winding distribution factor
Kp = coil pitch factor.
Kd = slots skew factor.

the following are other parameters for reference:

0.064 air-gap 0.080” air-gap 0.096” air-gap

Flux per pole (kLines) 7839 7839 7839
Air gap flux density(kL/sqin) 64.5 62.9 62.6
Saturation factor 1.463 1.350 1.300
stator tooth density 117.1 117.1 117.1
rotor tooth density 124.6 121.9 121.6
stator zz factor .307 .281 .254
stator end factor .356 .356 .356
Magnetizing reactance (p.u.) 1.948 1.747 1.553
R2 hot (p.u) .00588 .00588 .00588
R1 hot (p.u.) .00890 .00890 .00890
R equiv to iron loss (p.u.) .06024 .04295 .03193
Locked rotor P.F. (%) 27.7 26.9 26.2
kVA/HP @ L.R. 6.705 6.538 6.392
X/R ratio @ L.R. 12.502 12.822 13.114

The last parameter X/R shows that the total reactance increases with the air gap increase at Locked Rotor condition.
Since the resistance does not change.

 

[electricpete]

Thanks aolalde, that is definitely good info (starworthy).

It makes sense that maybe there are different values of reactance for starting and running which explains how LRC goes down even though Xm X1 and X2 all appeared to go down.
I still have not come to an understanding of why X1 and X2 would go down rather than up even in running conditions. Will have to read up on the various components of leakage reactance.


=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[aolalde]

E-pete.

Another thought, based only in the flux linked by the rotor bars. With increased air gap the flux catched by the rotor bars is lower and the rate of lines cut is the line frequency, then the induced voltage in the rotor is lower and so the rotor current component at locked rotor,which has more influnce than the increasing magnetizing current.

 

[electricpete]

Yes, I agree with that. I was thinking the same as you are thinking that upon increasing airgap, there is an increased fraction of stator flux which does not couple to the rotor and increased fraction of rotor flux which does not couple to the stator. From my understanding of a power transformer model, Xm is associated with mutually coupled flux and X1 and X2 are associated with flux that does not couple both windings. So the increased airgap should by my thinking result in X1 increase, X2 increase, Xm decreases.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[bentov]

Very interesting, and mostly over my head. I can tell electripete from experience on the repair shop floor that if you apply rated voltage to a rotorless stator, it will draw more than FLC (like, no time to measure, what the hell are you doing? Shut it off! Shut it off!). I'm surprised at the program results though - I can follow (sort of) how increased air gap would result in lower LRC, LRT, and PF, and higher NLC and FLC, but I would think that higher hysteresis losses in the larger air gap would drive efficiency down significantly (appears nearly constant in the table) . . .

 

[jraef]

I learn so much in this forum. Thanks for all the great posts to this thread!