DIY Gas Producer and Power Turbine Build

[cchange4us]

Gday everyone,

First up i have to say this site is fantastic, i have been reading a bunch of threads and have been able to answer so many of my own questions. This place has been a great resource. Now that has been said, on to my project:

I have been wanting to build a hobby gas-producer/power-turbine arrangement for a few years, however due to time constraints i have had to put it on hold. Beginning in a short while i will have a fair amount of free time and i am taking the opportunity to begin the build of my first diy gas turbine project. Looking forward it is my goal to create a combination which is compact, reliable and able to provide 40 hp at the power turbine shaft. I do however acknowledge that i should first be focusing on building the gas producer, optimising the combustion chamber and creating all of the correct supporting systems for such a producer. Below is a summary of the thoughts i have so far:

Which turbocharger to use - Admittedly, i have a bias towards holset diesel turbochargers as they are known for their reliability. I currently have a new and genuine holset HX35 sitting in my shed which from all of my reading appears to be an okay candidate for such a project. Straight away i can use the new HX35 to start building but for the first start up i will source a similar used HX35 and swap it over, just in case i mess things up! There are of course a few modifications that i will need to carrry out to the HX35 with one being the conversion to a single scroll turbine housing. I have read that the twin scroll inlet turbine housings have a tendency to soak up more heat than the single units which in my mind makes sense (more surface area as well as a large obstruction to flow which most likely causes turbulence). These conversion housings can be purchased on ebay (see below):

HX35 Turbocharger

Compressor:
54mm Inducer
78mm Exducer

Turbine:
60mm Exducer
70mm Inducer







Combustion Chamber and Flame Tube - I have toyed around with the JetSpecs calculator however as things progress i will be using some information i have read on this forum to calculate the desired dimensions, whilst of course also comparing it with Jetspcs. I want to treat it as an opportunity to learn more about the thermo and fluid dynamic aspect of gas turbines. For packaging sake, i am going to try and really optimize the efficiency of the combustor assembly and focus on controlling heat.

Electrical / Logic Control - I will use an Arduino Mega to take care of systems monitoring (sensors/readouts) as well as the starting/wind down process. I want to use a 24v DC system for powering everything. 24v because then it opens up the ability to adapt hobby brushless motors for various purposes. The brushless motor and controller packages appear to be a cost effective and compact way to provide power to various pumps and components, all whilst having the option to program some speed control into the motors if need be. 

Fuel Supply - Initially i will use LPG for simplicity sake though once i get the hang of things i am keen to try out a high pressure diesel injection system controlled by the arduino somehow. I am attracted to the idea of a high pressure injection system because i can see how it would benefit the burn qaulity, i can also imagine it would make start up times quicker with less time spent warming up with LPG. However first thing is first so for now a simple LPG injection system will be constructed. 

Oil Supply - I will use an electronic oil pump arrangement, either a modular unit that is completely off the shelf or a seperate pump/brushless motor combination so that i can use the Arduino to speed control the pump to account for viscosity changes/changes in turbocharger oil demand.

Starting / Shutdown - Of course in the short term a leaf blower will surfice however as things evolve i want to explore the idea of using a RC/hobby brushless ducted fan assembly. A system would need to be made that ducts the fans flow into the turbocharger compressor whilst starting and then opens the passage back up for running. What also excites me about this idea is that it can further be used to cool the engine after shutdown. A minute of getting fresh air shoved in its throat whilst still having the oil circuit flowing is ideal i think...



Power Turbine Section - Okay i am getting ahead of myself... 

After i get the gas producer schmicko and take the resultant readings of outlet pressure and temperature i will turn my attention to extracting shaft power. For that i am thinking either a c20 Rolls Royce 2nd or 3rd stage turbine blisk and respective NGV. I can fabricate my own NGV if needed though. Im not sure of which reduction method is best yet so this detail can be decided on down the track.  Its obvious that there are many, many things to be bought and built before i get to the power turbine section and hence i'll leave the discussion of it out for now.




I look forward to sharing my progress and by all means let me know what you think of my plan.

[racket]

Hi

Welcome to the Group :-)

I'll be looking forward to your Build Thread .

Cheers
John

[slittlewing]

All sounds good, personally I wouldn't worry about finding a "spare" turbo to start off with (or about the single scroll housing) because your money and time will be much better devoted to all the other parts above and the many other details.

I also think turbo damage likelihood is very low if you are vaguely sensible, stick to the jetspecs guidelines for combustor, make sure it has oil pressure and monitor the turbine outlet gas temperature (normally referred to as TOT).

I would recommend you base your combustor design around a central kerosene injector, with a propane injector to the side. This way, you wont have to re-do the combustor when you want to move form LPG to liquid fuel.

I also plan to use an Arduino Mega 2650 to control my turbine, having got it running initially with cheap ebay chinese PWM modules to control fuel pump speed (and therefore throttle).

Look forward to seeing your build,

Cheers

Scott

[cchange4us]

John - Thanks mate

Scott - Thanks for the comment. Perhaps you have a good point. I do notice that a lot of folks will begin one of these projects and get bogged down in the details to the point which actually stops them entirely (life gets in the way, we all have responsibilities... i understand). I will put the purchase of another turbo and housing on hold for this reason. This way i can imediately use the funds to buy build materials and then use my upcoming time to melt the materials together.

The central/side injection layout you describe makes sense and leaves room to grow, i will add that into the Solidworks model so we can all comment and tweak it if needed. I did some research last night into possibly using a piezo electric diesel injector from a common rail, controlled by the Arduino. Im skeptical though, even if i were to figure out a way to first pressurize the diesel and control the piezo electric injector (a project in itself), the spray would still be susbtaintially "pulsed" which is something that goes against the fundemental characterstic of "steady flow" for a gas turbine... perhaps i should be keeping things more simple.

In regards to the TOT - I thought that setting the probe up was going to cost an arm and a leg (thats what i am used to with the car scene) however, i have just found that i can use a standard "K Type" thermocouple sensor paired with a small interface chip to send a simple signal to the Arduino. This setup seems to be a very cost effective solution.



Arduino's are the bomb. They ofcourse have limits, but for DIY guys you really cant go past em!

Blair

[slittlewing]

Sounds like a good plan! Yeah the availability of shields appears to make everything quite simple and low cost, I found the same one when researching K type connectivity!

In terms of injector, I learned after joining this forum that the trusted method is to use an oil burner nozzle (danfoss or monarch) with a semi solid spray cone. I chose a PLP 80 degree monarch nozzle. They are inexpensive (£10) and made for continuous injection and also available in a huge range of flow sizes. You can run them with low pressure (between 2 and 10 bar) or ideally higher pressure is better (finer atomomisation). All you then need to do is choose a pump and control speed or delivery pressure.

Cheers

Scott

[racket]

Hi Scott

Preferable if the spray nozzle is Hollow Cone rather than Semi Solid

Cheers
John

[cchange4us]

I have ordered two sets of K-type thermocouples as well as the sensor interface chipset so that i can wire the thermcouple to my arduino.



Oil Supply



I want to get the parts on order for the oil supply, to do this i need to consider the turbocharger requirements as well as provide myself with some rough guesses of the equipment i want to use (hydraulic pump, electric motor and electric motor controller).



Here is a snippet of the HX35/40 manual:






So an oil pressure of between 30-72 PSI and a flow rate of 2-3 lpm is required.
On top of this i would like to leave some system capacity in reserve for down the track if i start to play with power turbines and gearboxes etc.


I would like to directly mate a hydraulic gear pump to a speed controlled DC motor. I have chosen to use 24V for this project. The gear pump would be very small in displacment (think 2cc/revolution or less). The system would be closed loop, with the arduino controlling the motor speed controller and a pressure tranducer feeding the arduino with the pressure.



This looks like a nice and cheap candidate for the pump:







not accounting viscous for losses, i can get a rough idea of what power and speed a 2cc pump will run at for 2 to 8 lpm at 50 PSI:







Assuming that the turbo consumes 2 lpm and that i do not have any other components drawing from the same circuit, I need to find a drive motor that doesnt mind sitting at 1000 rpm. There is no fundemental issue with driving the system at this speed however i do have a concern that the brushless motor would overheat because it is not moving enough air. Does anyone think this may be an issue?

Here is the motor i am thinking of purchasing:



If i understand these brushless motors correctly, 1000 RPM equates to (1000 / 170 ) = 6 Volts and (4000 / 170 ) = 24 Volts . This means that at my lowest calculated flow rate, i may be pulling a lot of current through the ESC and motor. I guess these motors have a maximum current rating of 80 A so the motor may manage just fine.


Im ready to pull the trigger on the pump, motor and ESC controller. What is everyones thoughts?

[slittlewing]

I like your measured approach and calculations, the only thing I would add is that I am using an off the shelf Marco UP9 gear pump. They are rated at 12v and have plenty of power for providing upto 4 bar (or a lot more depending on oil viscosity). If you went down this route, it might save you having to buy an ESC, doing a linkage between motor and Hydraulic pump etc.

If you choose to carry on with your current plan however, I will be following with interest! Would you mind sharing your calculation method/formulas for motor torque etc for a given hydraulic pressure and volume pump?

Cheers

Scott

[racket]

Hi

That brushless motor is a bit of overkill , you'll only be needing <10% of its power potential , I generally use 500W 36V motors which turn ~3,000 rpm , but run them on 12V for ~1,000 rpm and hook them directly to a Subaru oil pump for my 10 lpm , pics of one of them half way down this page jetandturbineowners.proboards.com/thread/78/garrett-gt6041-powered-kart?page=4 and a piggy backed one on this page , just hit the icon for the pics jetandturbineowners.proboards.com/thread/612/fat-boy-118-thrust-engine?page=10

You might need to look here for motors www.motiondynamics.com.au/ as I get the impression you're based here in Oz , if so, whereabouts ?

Cheers
John

[miuge]

I believe Andym was using 6455 220kv motor in his direct drive oil pump, pump was a 3cc/rev hydraulic unit what I read from my notes. Commonly these low-speed motors are rated high power, also quite cheap alone without the controls etc. I'm also planning to use a similar setup for my upcoming build.

[cchange4us]

Jan 12, 2020 11:21:59 GMT -5 slittlewing said:

I like your measured approach and calculations, the only thing I would add is that I am using an off the shelf Marco UP9 gear pump. They are rated at 12v and have plenty of power for providing upto 4 bar (or a lot more depending on oil viscosity). If you went down this route, it might save you having to buy an ESC, doing a linkage between motor and Hydraulic pump etc.

If you choose to carry on with your current plan however, I will be following with interest! Would you mind sharing your calculation method/formulas for motor torque etc for a given hydraulic pressure and volume pump?

Cheers

Scott

Hi Scott, Thanks. Your comment led me to have a squiz on google and find that style of pump. Those pumps look great, perhaps a good pairing would be a 10lpm 24v marco up9 pump with a PWM controller. The controller could lower the pump speed to moderate the pressure at low flow rates. I'll say though, the only reason im not actually going down this route is that by the time i read your comment i had already placed the order for the 2cc hydraulic chinese-made unit... i guess im committed now.

The calculations are based around "pump work", which is an "energy in = energy out" method. I am busy with work this week however on the weekend i will compile a sexy looking spreadsheet and reference it back to some example calculations, i'll then share them on my google drive. I'll try be propper about things and add in pump efficiencies etc. No promises on how accurate it will be, though sometimes it is good just to have a rough estimate of the system.

[cchange4us]

Jan 12, 2020 15:39:18 GMT -5 racket said:

Hi

That brushless motor is a bit of overkill , you'll only be needing <10% of its power potential , I generally use 500W 36V motors which turn ~3,000 rpm , but run them on 12V for ~1,000 rpm and hook them directly to a Subaru oil pump for my 10 lpm , pics of one of them half way down this page jetandturbineowners.proboards.com/thread/78/garrett-gt6041-powered-kart?page=4 and a piggy backed one on this page , just hit the icon for the pics jetandturbineowners.proboards.com/thread/612/fat-boy-118-thrust-engine?page=10

You might need to look here for motors www.motiondynamics.com.au/ as I get the impression you're based here in Oz , if so, whereabouts ?

Cheers
John

Hi John. Yes, probably way overkill. The thing is though, the only brushless motors that are close to the KV rating (from AlienPower) i need are all quite high power output. With that said the brushless motors still appear to be relatively small in size when compared to brushed alternatives of the same power rating so having a bit of overkill does not impact the scheme of things in that regard. I was not aware of motiondynamics, their website looks really good.

The contender to the 6374S mentioned earlier would be the MY9764-450W im guessing, however what made me decide against that was the physical size. I am trying to package everything really tight.

I hang around Perth and Bunbury in W.A

[cchange4us]

Jan 12, 2020 17:33:40 GMT -5 miuge said:

I believe Andym was using 6455 220kv motor in his direct drive oil pump, pump was a 3cc/rev hydraulic unit what I read from my notes. Commonly these low-speed motors are rated high power, also quite cheap alone without the controls etc. I'm also planning to use a similar setup for my upcoming build.

www.youtube.com/watch?v=KZB-WAG9lSY

I stumbled across this video, it couldnt be more applicable to the subject. Hopefully this proves the idea a bit. I really like the idea of not having a pressure relief valve and relying on the arduino to do the pressure control instead. From my understanding this is the way to create an efficient system, albeit possibly hard to initally program.

[slittlewing]

Excellent, look forward to the excel sheet. I agree PWM is the way to go. I did manual PWM control (with a cheap Chinese one) and it worked nicely on my previous fire ups (no regulator or anything). Now am closing the loop with arduino!

Efficient as you say, allows for automated low pressure starting, I intend to just keep the target oil pressure about 1 bar above P2.

Cheers

Scott

[racket]

Hi

Bunbury is nice :-)

Plenty of good turbo suppliers/repairers in Perth , 25 years ago, whilst living in Perth for a while, I scavenged the Cummins W shaft for my 2 shaft kart build jetandturbineowners.proboards.com/thread/40/2-shaft-turbine-kart-build out of one of their scrap bins .

Cheers
John