gear on the turbine shaft.

[greazy]

long time reader first time writer. an average guy obsessed with turbines and have a metal fabrication background.

so i was wondering if anyone has built a jet with a gear set on the turbine shaft to drive aux units like fuel and oil?? just like a commercial jet. making it a complete unit would be handy.

been toying with the idea in my head but i don't really have experience with gears. any thoughts on the idea appreciated

[gidge348]

Hi Greazy & welcome,

I am not an expert on this but my guess would be that the Rpm's that the smaller engines run at would be the stumbling block.

A large commercial jet engine may run at say 10,000 and has a reduction gearbox down to say 2-3000 rpm to run the ancillaries. On the other hand the smaller DIY engines may run at 50,000 to 100,000 so getting a reduction system from those kinds of Rpm's would be a challenge.

Also the turbine shafts at that speed have all kinds of weird harmonics going on so although it may be possible, it may not be worth he effort.

Interesting question though

cheers
Ian...

[finiteparts]

Hi Greazy,

There are several issues with gearing a turbocharger rotor that, though they are challenging, should be workable.

The first challenge is the pitchline velocity of the rotor mounted gear. You start to get all kinds of problems when you push spur gears over 10000 fpm pitchline velocity. So if we calculate the required pitch diameter to keep below 10000 fpm, for a turbocharger rotor at 100 krpms, we get around 0.38 inch. Cool, so the gear has to be around 3/8" pitchline diamter....but that is at the upper limit of 10000 fpm....so if we try to reduce that pitchline velocity to a more reasonable number, like 5000 fpm, we get a pitchline diameter of 0.191 inch. That's likely smaller than the shaft diameter. So the pitchline velocity drives a diameter limitation that can become problematic...but not insurmountable.

The second challenge is that at the high speeds that turbocharger rotors operate and due to their shaft sizes and bearing spans, they operate in super-critical bending modes. What does that mean? Well, when an object vibrates, it takes on a certain bending shape at a certain frequency, called it's natural mode at it's natural frequency. This natural frequency is modified by the system...the bearing stiffness, the center housing stiffness, the system damping, etc...When the rotor spins above the first critical speed (1st natural frequency) it is said to be super-critical. Usually the first mode is a bending mode were the rotor takes on a u shape, with zero shaft deflection at the bearings (like a kids jump-rope...with the kids hands at each end being the bearings). As long as the rotor doesn't dwell at or near that speed very long, the shaft will not experience harmful shaft bending amplitudes. If it was allowed to dwell there, the driving frequency (any shaft imbalance, etc) could lock-in to the natural frequency and drive resonant vibrations that can grow destructively in amplitude.

SO, what does this have to do with a gear? In the operating speeds of the turbo, the shaft may not be "straight" at the end of the shaft which is the likely spot for mounting the gear. It would be very hard to have a spur gear survive when the contact patch on the gear face is moving over the operating range of the rotor. The gear face contact stresses would change and if the gear alignment was highly skewed, it could lead to heavily overloaded contact stresses. Additionally, the shaft centerline moves around due to the journal bearings "adjusting" the oil film clearance as the load shifts. This shaft eccentricity is a natural and desirable feature of any hydrodynamic bearing, but it doesn't work out to well for a gear system.

Moving the gear closer to the bearing to reduce the radial movement in the gear mesh actually makes things worse, since any load that is imposed on the shaft due to the gearing would impact the journal bearings natural eccentricity. The use of rolling element bearings greatly reduces the shaft centerline eccentricity and this is how the commercial APUs, GTSs, and small turbojets can drive accessories from the shaft...but converting a turbocharger hydrodynamic bearing rotor to operate with rolling element bearings is a whole other challenge. Additionally, shafts supported on rolling element bearings, due to the higher bearing stiffness, operate farther from the rotors second mode, so the impact of additional mass (gear) and shaft bending can be smaller than on the hydrodynamic bearing rotors.

In addition to the above challenges, the gears extra mass would shift the rotors natural frequency. If the gear was placed out by the compressor nut, the mass has been added far from the bearing and will more greatly drive the bending due to the cantilevered compressor imbalance/natural frequency. If the rotor is supported on floating ring bearings, it is likely that normal operation is very close to the second bending mode and thus the gear could actually cause the second critical shaft speed to drop down into the normal operating rpm range.

I imagine that you are probably wondering why I keep saying spur gears...why not put a 90 degree gearset in there? The reason is that other types of gears introduce additional side forces. So if I put a 90 deg helical gearset in there, it would actually introduce a bending moment into the shaft at the gear mounting. This highly complicates the rotordynamic loads and that is the reason that I limited the discussion to spur gears, which only introduce in plane loads.

Also, as Ian said above, the introduction of the gearing now supplies the shaft with more driving frequencies with which it could couple and generate high resonant vibrations. This "can" be anticipated in the design phase by checking to make sure that none of the driving frequencies (due to the number of teeth in the gearing, the ball frequencies in the bearings or the gear box mounting natural frequencies) are close to the natural frequencies of the rotor. There are all kinds of dampers out there for gearing to help reduce these, but they could be tough on a small gearbox and warrants some investigation.

These are just the big challenges, there are more including oil churn and heat generation, gear stresses, losses, etc, but this gets the discussion going. I am sure that someone could make this work out...most of the above design challenges could be overcome by using a properly designed quill shaft. It would allow some small shaft misalignment, reduce the transmitted gear noise back into the shaft, it could be "floated" on it's own bearing to potentially minimize it's impact on the shaft natural frequency (especially if it replaced the compressor lock nut), etc. But of course, designing and building a good quality quill shaft is no small undertaking, either. Hopefully, I haven't discouraged you with this information, but instead, gave you some things to research and work around in your design of a successful rotor gearing system.

I seem to recall that NASA had a really good high speed bearing design guide...I will look around to see if I can find it and post it so that anyone interested can find it.

Good luck!

Chris

[pitciblackscotland]

G'day Greazy,
As Ian and Chris said it would be a challenge. I did some experiments on my first turbojet, bleeding air from the combustor to spin up a alternator winch work but as soon as i put load on it it stop.
www.youtube.com/watch?v=IOGBhX4HpjU Check out RC Don vid www.youtube.com/watch?v=EoztC5-8E6s

Cheers,
Mark.

[greazy]

very extensive chris the quill shaft sounds like a promising avenew ideally I'd like the unit not to protude from the front of the housing but rather around it. thoughts on the use of a belt drive set as close behind the compressor side bearing as possible?? is it a good way to reduce the vibration problems that gears may cause?

I'll have to do a bit of research on dampers i have no idea what the are haha.

ryan

[Iron_Knight]

Simple Solution Starter/Generator... We Have Seen Them Build Starters Diy On TurboChargers Up To 20,000 rpms.. Leave It Connected And Start Pulling Her Down Till You Get Required Voltage.. To Run AUX Pumps. Never Know Till You Try.

[gidge348]

Yep, along that line maybe copy the idea of the Mercedes F1 turbo system.



It uses an inductive motor/generator (part of the KERS system) in the shaft between the compressor & turbine. In this application the motor/generator acts as motor to keep the compressor at optimal speed and as a generator to control boost pressure rather than dumping it to a waste gate as well as charging batteries.

In a small turbine it could be a starter and supply power to run pumps....

Just a thought.

Cheers
Ian...

[racket]

Hi

Checkout these guys www.jakadofsky.com/htm_en/company.htm somehow they manage to get high rpm gearing to work

Cheers
John

[greazy]

the jakadofsky engine is totally **top secret** i cant find any info on it anywhere apart from their vauge website (cool that they managed to do it though). I have had a while to ponder and research the idea after the info you guys have shared and decided unless i make an engine that rotates less than 40,000 its just too difficult and not worth the effort/moneys anyone know a compressor wheel that maxes out at 40,000??? I do like the idea that you can load up the starter to charge the battery ill keep it in mind for the next build. if anyone comes up with any feasible design or idea hit me up and ill have a crack at fabbing it up!


off topic but this kamps book people mention i'd appreciate a good link to buy the book for a fair price or even a download version.

cheers ryan

[racket]

Hi Ryan

Not many turbos only run to 40K , though some of our larger ones will be producing some decent power at that rpm, as we only go to ~60 -70K at 100%.

From what I can gather on the Jakadofsky engine it appears as though they have a gear on the shaft behind the comp wheel which then drives another one or two gears to get the output shaft through one of the diffuser vanes, the lube for the gears appears to come from the usual fuel/oil mix, probably before it then goes through the bearings.

This setup would be OK with ball bearings on the main shaft , but with our "brass bushes" it could be problematic .

What I have often though about would be a gear centred between the brass bushes similar to the industrial Centac centrif compressors , which also use "brass"
www.caps.com.au/wp-content/uploads/2012/06/Ingersoll-Rand-Oil-Free-Centac-Compressor.pdf

Kamps book , buy here au.trapletshop.com/model-jet-engines-3rd-edition-by-thomas-kamps those grossly overpriced ones on Amazon are a joke , buy direct from Traplet the distributor, there should be an US Traplet shop which could save on postage .

Cheers
John

[enginewhisperer]

what about a planetary setup with the shaft through the middle? It should at least have balanced forces on the shaft

[racket]

Hi Andrew

Planetary would be better , but impossible to fit on a turbo :-(

I was thinking to use a standard turbo , with the takeoff gears fitted under the turbo in the oil outlet area so that the draining oil lubricates things , its just something I've had in my head for a long time :-)

Cheers
John

[finiteparts]

Hi Ryan,

Here is the patent image of the Jakadofsky engine (colored so that it is more clear)...it should be quite clear how the gearing works...though the aero impact of a giant gear in the compressor discharge path, is sort of bothersome...



As for belt drives, you probably couldn't find a belt that wanted to stay on and even if you did, you would have to tension it so much that the side load on the bearings would be large...

Going with rolling element bearings means that the rotor shaft's eccentricity is much smaller and more controlled as compared to hydrodynamic bearings...in fact, Honeywell has a patent on using ball bearings as a means to increase turbocharger efficiency due to the reduced rotor clearances required. But there is a lot that goes with getting a rolling element rotor to live while being tortured with high speeds and heats....

If you used hydrodynamic bearings, then the trick is to go to larger rotors such that the journal clearance is relatively small compared with the gear meshing tolerance. Often, the industrial units only use hydrodynamic bearings on the rotor shaft itself, the gearing is supported on rolling element bearings...the tolerance stack-up is tough with multiple hydrodynamic bearing clearances on AGMA Q-13 or higher tolerance gears.

The planetary gear set-up might not be too far out there...you maybe could steal a gear set from the newer Kawasaki supercharged street bikes.







The shaft gear would be challenging, but the other gears are lower speed and easier to manage the rotational and contact stresses.

There is no fundamental physics that is telling you that you cannot do any of this. There are some challenges, but don't give up on it...I think it is achievable for our purposes.

As for a damper...it is a device that converts vibrational energy to something else, like heat or noise...like the shocks on your car. Take off the shocks and when you hit bumps, you car will bounce for a while...the shock absorbers damp out some of the vibrational energy so that you don't bounce...this is called the over-damped case. Under-damped would mean that you need new shocks because your car keeps bouncing after hitting a bump.

Now don't get confused with a dampener...that makes things wet! Ha!

Good luck!

Chris

[finiteparts]

I finally dug up the gear design document that I was referring to in an earlier post. I totally forgot that it was specifically referring to turbopumps, so I had to actually go through the files and find the actual document.

      NASA SP-8100, "Liquid Rocket Engine Turbopump Gears" ( ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19750002094.pdf )

So if you go by their standards, 15000 ft/min is the upper range of the "Normal" class of gearing (see page 11). Dennis Townsend (of NASA, Lewis Research Center) has the 10000 ft/min limit stated in another paper, based mainly on noise, trapped air/oil issues, etc causing wear...it appears from SP-8100 that the basis in this document is the mechanical loads and the work of Darle Dudley...which happens to be a leading expert on gearing and of course he is the writer of the gear design "bible", "Dudley's Gear Handbook" and the practical implementation version:

    "Handbook of Practical Gear Design" ( books.google.com/books?id=NOqBnpN7ElIC&printsec=frontcover&dq=Dudley#v=onepage&q=Dudley&f=false )

....both are excellent resources for your design book collection...highly recommended!

While I was poking around the old documents, I came across this one too...

     NASA TM-87096, "Lubrication and Cooling for High Speed Gears" ( ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19850026094.pdf )

It does an excellent job describing the challenges of getting the oil into the meshing gear teeth. At high rotational speeds, the oil jet can just be thrown out of the meshing gears due to the high centrifugal force and then a portion of the gear can actually be operating in a lubrication starved condition...this of course leads to high wear rates and early gear tooth failure.

Enjoy!

Chris

[racket]

Hi Chris

Thanks for the pic of the Jakadofsky , I've wondered what was inside :-)

So the gear is simply cut into the shaft and then a two stage redux , pretty straight forward.

LOL, but that interruption to the airflow certainly is a concern , not the nicest solution especially as theres going to be a need for some sort of "casing" around the gears ...........a freepower is much cleaner.

Good Paper on lubrication .

Cheers
John