Gas turbine build - 100+ shp

[dieselworks]

Hi everyone,



After completing my first turbojet recently, I have come to the conclusion that it is time to start over. Not that it didn't turn out okay, just that there are too many components which I would like to improve for it to make sense continuing work on that project.



As such, I am working on starting a new project. I am going to build a turboshaft engine with the goal of producing at least 100 shp. I would like the gas generator to be as simple as possible and fortunately there has been a lot of good advice given on this topic. 

About the gas generator:

I am looking at using an On3Performance turbocharger for the gas turbine. These are chinese turbos with some nice features, like properly sized turbines for jet engines and a cast stainless steel exhaust housing with a large V-band inlet.

The one I am looking at has these specifications:

Compressor Wheel Details:

Inducer 94mm

Exducer 138mm


Turbine Wheel Details:
Inducer 112mm

Exducer 102mm (17.7% larger area than compressor inducer)

on3performance.com/products/on-3-performance-94mm-cnc-billet-wheel-turbocharger-t6-stainless-v-band-exhaust-housing?_pos=6&_fid=1f55cb7fc&_ss=c

I think this is probably a good candidate for this build, but I would be curious to hear the thoughts of some experienced members.

Based on John's (Racket) estimate of mass flow potential for my 62mm inducer engine of 0.83 lbm/s, I am guessing that an 0.177 lbm/s per 1 sq in of inducer area is probably reasonable. Based on an area of 10.75 sq in of inducer, would 1.9 lbm/s of air flow be in the ballpark? Maybe a bit more due to larger volume/surface area?

About the freepower stage:

I have found an APS3200 turbine and NGV. The NGV doesn't exactly match but I think it will work with some light modification. The APS3200 has a rated fuel mass flow of 327 lbm/hr gives a fuel flow of 0.090 lbm/s and multiplying by an estimated AFR of 60 gives 5.5 lbm/s air mass flow rate. The design speed is 50,000 rpm. This gives me confidence that the APS3200 would be sufficiently sized for my application. I would need to make the bearing section and housing. I think it would be quite a challenge, but doable. 

The wheel in question:



Another option, which would definitely be preferable to me in some ways would be to use an existing turbocharger turbine and housing instead. I may be able to get an turbine from On3Performance with 118mm inducer and 112mm exducer with a lightweight cast stainless steel housing to match. Then all I would need to make is a suitable bearing section, which would be much easier than for the axial style unit. Would this turbine setup be suitably sized for this application? It has an exit area that is 20.5% larger than the gas generator turbine. Would it be possible to use the same size 112/102mm turbine from the gas generator for the freepower stage? If so, this would definitely save some cost. 

About the transmission:

I am planning on using a CVT from a Polaris RZR ATV. These can handle 150+ hp and relatively high input speeds, so a single stage reduction with gears or high speed minipitch timing belts would likely work. I would be curious about people's thoughts on this point, since decoupling is not needed with freepower turbines. I am considering this route to improve off the line acceleration since the CVT won't start transmitting torque until the turbine has reached decent speeds and therefore power level.

CVT in question:



The extras:

This is going to go into a cut down Fiat 500. I think Fiat was silly for building such a large car so I am going to take the wheels, spindles, brakes, VIN and title and build a proper street legal go kart. That's the main reason why I want the CVT. I will consider the project a failure if it has a curb weight over 750lbs. Maybe ambitious I don't know. 

Fiat 500, actually small. 


I can't wait to hear your thoughts and suggestions. 

Best,
Brandon

[noshell]

I'll be following this for sure. My goal is something similar on a far smaller scale. I like the idea of using an existing turbo housing for the freepower, and I also want to use a cvt and centrifugal clutch.
Ill be interested to see what the more experienced guys have to say.

[racket]

Hi Brandon

That turbo in the Link might flow considerably more than 1.9 lbs/sec , best to develop it as a pure jet first so that you can collect data on jetpipe temps and pressures which along with your thrust level will provide the necessary info for sizing the freepower wheel as well as potential power output .

I don't think you'll find a large enough radial wheel for the freepower , a 112 mm exducer will be a bit small , the axial freepower I used in my bike had 10 sq ins of throat area but only had an 88 mm comp inducer feeding it, that 94mm billet is a modern design compared to the TV84 so will undoubtedly have better flow capacity .

Have a look at Anders build jetandturbineowners.proboards.com/thread/53/landspeed-bike-diy-gas-turbine it ran a freepower with 15 sq ins of flow area at the throat

Cheers
John

[noshell]

I saw a 6cyl train engine on the engine dyno at work.
That had a very big turbo on it. I'm fairly sure it actually had an axial turbine. Might be worth looking into.
Edit: Seems what I saw is pretty unusual . Most locomotive turbos look like the size of a dining table haha.

[dieselguy86]

On3 is only 1.5 hours from me, I wouldn't trust their turbos... they work ok for cars that don't run turbos very hard.

What you linked to is a 94mm gt55 turbo, my concern with that one is that alot of times the Chinese will copy a turbo but change a couple small things to get around any potential lawsuit. This can leave you with only one variation of turbo. Meaning, if you find out later that the compressor is to big, or a/r is to small, your screwed because nothing else out there fits without machining something.

I "rebuilt" one of their 78/75 turbos last summer and it was a hodgepodge of parts, something from t4 turbos, some from gt37 turbos, and stuff I couldn't locate and had to re-use. I was able to gain 200hp for the customer by doing some turbine side customization, all because a bigger a/r housing wouldn't fit over the custom sized turbine wheel.

You could use that turbo as the freepower and get a gt42 turbo for your gas producer. The gt42 and gt55 is a well proven match, it's what cat/ garrett used for compounds on their 15L acert engines. Done right I think you could still get 100shp.