why vary pitch in a generator 7

[eleclew]

What is the affect of varying the "pitch" in the windings of a generator - is it to reduce eddy current losses and harmonic currents - we have a generator with a 7/9 pitch (so we're told).

 

[Motorwinder]

I recently wound a 400KW Cat generator that had a 12,10,8,6 pitch. Since I don't have basket heads that big, I wanted to convert to lap.

I converted using an 11 pitch, and ran it by EASA.

They said I needed a 1-9 pitch.

This was their response.

The 1-9 span must be used not only to closely match the original flux densities, it is necessary because it duplicates the effective 2/3 pitch of the original winding. The 2/3 pitch (8 slots of 12 per a pole) eliminates the third harmonic in the generator output; and may be needed for paralleling to other 2/3 pitch generators.

 

[electricpete]

Thanks Muthu. So there is in fact a fairly wide variety of pitches in used. I would presume this comes from a balance of many different factors relevant to the application. If I'm understanding Wolf correctly copper doesn't belong in that discussion....(not sure I understood why that would be or whether that's really what he's saying).

For 2-pole I suspect it is construction and cooling considerations for the congested wendwinding moreso than the copper considerations I mentioned that drives it toward low pitches.

If we restrict the discussion to hydro, I gather 5/6 pitch is practically universal. That's an interesting thing and it's got me wondering why that appears the case. I have 2 ideas (open to comment):
1 – There is a famous curve of harmonic leakge reactance that shows a minimum at around 0.83 = 5/6. It is shown here. Whether and why it might be more of a relevant factor for hydros as opposed to others I have no idea (maybe it's not the reason hydros stick exclusively with 5/6).

http://books.google.com/books?id=th...TOR DIFFERENTIAL LEAKAGE COEFFICIENT"&f=false

2 - Hydros tend to have very few number of slots per group, so that they cannot take advantage of low distribution factors (in addition to pitch factors) to drive the harmonics down to the extent other applications can. For example if q=5 slots per group, then distribution factor for 7th harmomic is 0.15. We may be able to choose 7/9 pitch even though it doesn't drop the 7th harmonic down as low as 5/6 would.



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

 

[zlatkodo]

Hi Motorwinder,

I recently wound a 400KW Cat generator that had a 12,10,8,6 pitch. Since I don't have basket heads that big, I wanted to convert to lap.
I converted using an 11 pitch, and ran it by EASA.
They said I needed a 1-9 pitch.

The correct replacement for the step 12,10,8,6 (concentric) is lap 1 - 9.
If you are using 1-10, you have changed the fundamental winding factor, and therefore turns / phase should be re-calculated. Such change in the number of turns / phase is not always negligible.
Zlatkodo

 

[wolf39]

Hi everyone:

My feeling is the hydro generator field might be too special for this forum and because I currently am engaged in the preparation of two seminars there is little time left for other activities.

To finish the pitch discussion from my side with respect to hydro generators, I'd like to repeat my main points:

1. A pitch in the "region" of 5/6 is most effective to reduce 5th and 7th harmonics. The term "region" is important, because a design with 3 slots per pole and phase, for instance, can only be pitched 6/9 or 7/9 or 8/9. Both, 7/9 and 8/9, are equally close to 5/6 but I would select 8/9 for better flux utilization.

2. All posts sofar were referring to multiple turn coil windings, i.e. lap windings. This type of winding has the disadvantage to require lots of jumpers, adding to the copper weight (the number of jumpers can be calculated to 3 * number of poles). Because of this, large low-speed hydro generators nowadays will be designed with Roebel bar wave windings. There is no pitch related copper saving here but the vast reduction of jumpers (saving material and labor) does outweigh this easily.

3. Turbogenerators always have lap windings and a winding pitch as close to 5/6 is especially important for such units because the 5th and 7th harmonics would be responsible for very high extra losses in the massive rotor surfaces. Field poles of hydro generators usually are laminated. Therefore these harmonics here are of lesser importance loss-wise.

Wolf

http://www.hydropower-consult.com

 

[electricpete]

I am very glad you'll be sticking around at least a little bit. I will study your post because I'm sure there's a lot I can learn from it

As you pointed out, I was wrong about suspecting increase in core losses on the stator for the full pitched coil due to harmonics. For my own benefit (and maybe bystanders will find it helpful), I will recall the reason for that: The harmonics we are talking about are spatial harmonic produced by stationary coils. In the stationary frame, they have the same time frequency as the fundamental, but their wavelength lambda if smaller by a factor h (where h is the harmonic number). Using the wave speed equation c = f*lambda, their speed c must be a factor of h lower than the fundamental and we know the direction is reversed for the 5th and 11th. So they impose no different time frequency for the stator core, but can impose dramatically higher time frequency for the rotor core, which can increase rotor iron losses (also depending on other factors mentioned: small airgap couples the higher harmonics across the airgap much more effectively).


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

 

[electricpete]

but can impose dramatically higher time frequency for the rotor core,

maybe I shouldn't have said dramatically higher... those harmonics are moving slow so they look approximately stationary, so the time frequency as seen by the rotor is approximately 50hz or 60hz, which is higher than the main wave frequency seen by the rotor (0 for sync machine and slip frequency for induction machine), but dramatically higher may have been misleading. This is in contrast to negative sequence fundamental where the frequency seen by the rotor can be 100hz or 120hz. I'll shut up now.

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