100-150 Hp turboshaft thoughts

[racket]

Hi Monty

I design for 10 lpm of oil, and 70 psi with cold oil , my fuel pump is connected to an oil pressure switch and disconnects power to the fuel pump if oil pressure drops below 45 psi , as for pumps , different guys use different pumps , I like my Subaru oilpump powered by a 24V motor running on 12V , I prefer the automotive pump as it has pressure relief valves as well as filter , it pumps ~ 10 cc/revolution.

Yep , know Barden , got one of their catalogues somewhere , but out of my budget range , I did experiment with home made hybid full complement bearings though , but prefer "brass" these days as its reliable and cheap .

Copying shamelessly is the way to go :-)

Cheers
John

[monty]

John,

Do you have a good reference for calculating the diffuser wedges and vanes? My inducer is 95mm exducer is 159.03mm with a 7.55mm tip ht. I'm planning to use wedges and then vanes on the outside.

I got my turbine in. The inducer measures 129X19mm The exducer is 111mm OD and about 34mm ID. Looks like about 4-5mm clip on the tips of the exducer.

Monty

[racket]

Hi Monty

Good Paper library.e.abb.com/public/032ab8398281326b85257da5007742ba/The%20Matchin%20of%20a%20Vaned%20Diffuser.pdf.

The turb wheel sounds like Andy M's ibb.co/eybJJU, the Holset and Garrett wheels are very similar just different shafting.

Could you please verify the Part Number of the turb wheel .

Just looking through some notes and I notice a reference to the tv94 F Trim Rotomaster wheel , it seems to have ~0.150 less clipping and tip height at ~20 mm , the less clipping provides ~5 degrees more deflection at the exducer tip, but less flow area

Cheers
John

[racket]

and another www.pcaeng.co.uk/library/Publications/TechnicalPublications/GT2012-69151.pdf

[monty]

Thanks John,

It is the Garrett 442208-003 wheel. It came in a Garrett Honeywell box, so I'm sure it's the real part. It looks like the one in the link you posted, except it has a hex nut on the exducer, not a 12 point.

Monty

[racket]

Hi Monty

Yep , Garrett have a hex "nut" , thanks for confirming the part number , .......... who did you get it through, was it at a reasonable price ??

Cheers
John

[monty]

John,

Here is where I purchased the wheel. Price was pretty good, with free shipping and prompt delivery. I had it quoted by a couple turbo shops and they all wanted almost $1000!!!!

What kind of efficiencies are you getting with the KTS wheels? I have a spreadsheet that I got from the GTBA that I used in the past to calculate diffuser and NGV geometry. I always had a map for the compressor I was using so I knew efficiencies, and mass flow. This is the first time I'm totally guesstimating....

Monty

[racket]

Hi Monty

Thats a reasonable price for a genuine Garrett wheel , talking to the turbo repair guy at a local earthmoving equipment shop www.expressway.com.au/ he said the Garrett wheels are "right" whereas the aftermarket ones can be a bit "rough" , in your situation you need the right wheel.

The KTS wheels should give 80% effic with a good vaned diffuser , thats what I'm currently seeing with the 12/118 engine and its a bit of an "unknown" quantity, I'll look around for some more Links to Papers for you that might help in determining what you can expect .

You'll need to measure the comp wheel when it arrives to get inducer tip angles, throat etc etc .............LOL, and please post a pic :-)

I use the maps on the Garrett Site as a rough guide to flows vs inducer areas ,roughly ~13 lbs/sq inch/min , so your 95 mm-3.74" inducer of ~11 sq ins should flow ~143 lbs/min or ~2.4 lbs/sec , this is roughly what our "old" TV94 comp wheels flowed at 98 mm inducer, the wheel that I used in my 9/94 and 10/98 engines and Anders in his bikes JU-01 engine

Cheers
John

[racket]

Hi Monty

If you can find it, this is a good Paper , Centrifugal Compressor Design by P M Came and C J Robinson , PCA Engineers . Lincoln , UK .

I can't remember where I got it from though :-(

Cheers
John

[monty]

John,

I think that first reference will get me there.

I ordered the gears today. I've been thinking about the theory regarding high speed gears, and discussing with a colleague who has done a bit of gear design. We think a lot of the theory is bunk. They are worried about tenths of a thou...but no mention of at what temperature. What about case expansion? What about bearing expansion? Simple arithmetic puts the theory in bunk category. You can grind gears to the tenth, but the instant they mesh and run under load all that is gone! Somehow Chinese air tools function with gears that look like they were chiseled from raw stock by drunkards with bad eyesight....

My plan is to make sure the bending is OK. Harden the gears and rough lap them together at 180F. Then I'm going to send them out to be nitrided, and polish them together at 180F. Surface durability is THE issue at high speed, that and error in positioning/machining...hopefully the surface treatment and lapping/polishing take care of that. Cooling is also an issue: oil sprayed between the opening gear mesh should take care of that. It's worth a punt...

I'll take some pics and start a build thread soon.

[racket]

Hi Monty

Yep , the difference in expansion coefficients between steel gears and an alloy casing soon changes meshing when the temps change .

The gearbox we made for this Project jetandturbineowners.proboards.com/thread/34/motorcycle-project suffers from a "tight" mesh when temps are low , but it soon gets some backlash once the alloy cases warm a bit .

I provided oil jet lube into the mesh of ~1 lpm.

The new kart gearbox I'm making wooden patterns for at the moment will only have rough cut spur gears , "durability" will be provided by the "chunky"size of the gears , duty cycles will be <1 minute .

Cheers
John

[monty]

John,

Sounds good to me. I'm currently planning to have the ring and pinion on rolling element bearings and the planet gears on hydrodynamic bearings. The thinking is this will allow the planets to self position and give some amount of "give" for thermal expansion. Being planets, they can move radially around the pinion, and still have the offset necessary with a pressure fed sleeve bearing while being in position. The radial tolerance on the sleeve should allow compensation for case expansion....hopefully. That's My current thinking anyway.

What size lines are you using for your fuel manifold inside the engine? I'm thinking of using -3 or -4 AN stainless for that.

Monty

[racket]

Hi Monty

LOL......AN lines , what are they , I use steel brake line, bundy tube , either 3/16", 1/4" or 5/16" for the manifold depending on the configuration/size, 1/4" should be big enough for your engine.

Cheers
John

[monty]

John,

It does help that I know where there is a drawer full of surplus stainless fittings, and I have the flare tool and coils of tube...so it's cheaper than dirt for me . Bulkhead fittings will work nicely to get the fuel and oil through the front cover. I'm going to tap the diffuser and thread them into it. The O-ring can seal the front cover, and a nut on the fitting in the back. Makes a nice assembly.

I found this....unfortunately after I've already shelled out the bucks for the Garrett wheel.

And also this!

And to make it worse THIS!

I could have just ordered all this stuff off of Ebay....and not had to wait so long or deal with the dealership crap...

Well live and learn I guess. Nice thing about the Holeset is the thrust bearing design. I could adapt it to the Garrett wheel since the stem sizes are the same if I decide to go that route.

what are your thoughts about matching to that wheel for the Garrett, vs the Holeset? I'm thinking I might get more thrust through the nozzle with the Garret and more shaft power with the Holeset since it isn't clipped.

[racket]

Hi Monty

That comp wheel is probably the X853 KTS wheel , other suppliers often incorporate the KTS #number onto their own .

As for the turb wheel , clipped vs unclipped, it'd be a matter of doing the full cycle calcs to determine the required flows at the exducer , if you run a diffuser exhaust to maximise HP production then its possible to have sub atmospheric static pressure at the exducer with reduced densities requiring more flow area, lotsa variable for you to work out .

Cheers
John