DIY Turboshaft Kart build for the University of Kansas

[Feathers]

Thanks guys,

We're using electronic oil pressure regulation so we don't waste electrical power with bypass regulation. With our fans, pumps, control system, and 150 amp starter system, we need to preserve every amp-hour we can.

Electronic oil pressure regulation shaves 10 amps off our operating current draw.

[racket]

Hi Feathers

Your oilpump power requirements seem a tad high , originally I had a 10 amp fuse fitted to my pumps 12 V power supply , but it blew on first startup , a 20 amp fitted and no problems despite it pumping 10 litres/min with cold oil ............theres some data here jetandturbineowners.proboards.com/thread/612/fat-boy-118-thrust-engine?page=11

Cheers
John

[Feathers]

Hey Racket,

Yeah, running oil pumps via PWM can have some interesting effects. Regarding output power vs current, they're much happier when they're spinning a bit faster than our UP-12 was on cold oil.

We're replacing the plastic end cap of our pump with a billet piece, and I'd like to find a heatsink to place around the pump as well, to prevent heatsoak from loosening magnets from the housing, or any similar crisis.

Also going to need a heatshield or blanket for the turbo, right next to our driver

- Feathers

[Feathers]

Here's an infrared video of some runtime.

I don't know if you guys have seen anything like this before, it's certainly helpful when working on things like this.



Sorry for the double-post.

-Feathers

[jetjeff]

Hi Austin,

Pretty interesting vid, thanks for posting.  Were you using propane as fuel?

Regards

Jeff

[Johansson]

That is a great idea to use a heat camera Feathers!

I have a FLIR infrared camera at work so I will use it to check the bike out when running stationary to search for hot spots around the interstage duct.

[racket]

Hi Feathers

Interesting vid .

I did notice on the vid that your "exhaust" pipe appears to have a fairly steep angle between the clamp and straight pipe section , unless the straight pipe goes through to the turb wheel exducer you could end up with some massive restriction problems due to the turbulence created by the sudden increase then decrease in flow area downstream of the exducer.

Cheers
John

[Feathers]

Hi Austin,

Pretty interesting vid, thanks for posting. Were you using propane as fuel?

Regards

Jeff

Hi Jeff, we're burning diesel, and lots of it... 1.5 gallons for 10 minutes of runtime.

Hi Feathers

Interesting vid .

I did notice on the vid that your "exhaust" pipe appears to have a fairly steep angle between the clamp and straight pipe section , unless the straight pipe goes through to the turb wheel exducer you could end up with some massive restriction problems due to the turbulence created by the sudden increase then decrease in flow area downstream of the exducer.

Cheers
John

Our exhaust is a little weird right now We've got a nozzle on it to simulate the restriction of some NGVs, and it contains flow-straighteners so we can make measurements with a pitot tube and and determine mass flow.
The final exhaust will still be pretty messy, it's got to make a turn down into a power turbine plenum, but it won't be as "weird".

John, in your opinion, is >1100°F a reasonable temperature for a turbine housing? It seems to us that our engine is getting a little hot, and the turbine runs a very dull cherry at full speed.


We've also reassembled the starter system with a new shaft, new engagement, and an extra mechanical solenoid to hold the motor really "in there". We're tracking down some vibrations now:







- Feathers

[jetjeff]

Hi Austin,

593 degrees C is well within the norm for model airplane turbines.

Regards

Jeff

[racket]

Hi Feathers

Temps sound OK if a tad modest .............. , unless we are certain theres an Inco wheel inside and a new one at that , then temps exiting the flametube and entering the scroll need to be kept to a 1450 F maximum ,......... if we then subtract roughly the temp rise in compression , that should be our TOT .

Your scroll temperature will vary depending on how much of a Pressure Ratio there is across its throat as well as any frictional component thats being "subtracted" from the total temperature and is going back into the "static" scroll , you may be very likely measuring close to your T I T temp if measuring scroll temps .............not a reliable way to measure your very important temps .

You haven't indicated whether or not your exhaust tube ID matches your exducer shroud diameter and if the tube projects back through the "nozzle" to close by the exducer ,theres no point continuing with testing unless you have such an "exhaust" , the steep angle of the scroll casting downstream of the exducer is designed to create turbulent flow into a largish diameter exhaust pipe on the truck , whereas we are preferably needing lamina flow in our jetpipes.

With the correct diameter jetpipe in place , with no step between exducer shroud ID and pipe ID , you'll have a more accurate indication of TOTs if a couple of thermocouples are fitted a few inches downstream of the exducer, once you have those accurate TOT temps then a bit of maths will tell you if your most important T I Ts are within the 1450 F limits

Fitting any sort of restrictive jetnozzle can only be done once you have your datum data collected from that open pipe .

You mention "vibration" , was it there before you fitted the "jet pipe" ??

Cheers
John

[Feathers]

Hi Racket,

Watching the turbine while running, it was clear that we didn't have any temperature problems that threatened the engine, such as extreme inlet temperature.

We now have our communication working again, so we will begin to log data from the kart during runs It will collect TOT (via integrated thermocouple) and engine speed, in addition to practical concerns such as oil temperature, oil pump speed, starter speed, P2, and some others.

We'll have real TOT numbers for you soon.

The hot turbine scroll isn't great for us, as it means a hot core and hot oil, and greater requirements for our oil cooling system. Depending on the maximum oil temperatures we reach, we might have to make changes to our cooling system to avoid affecting the properties of the oil. We also need a "turbo blanket" to insulate the driver from the hot scroll, which will allow the scroll to get even hotter...

We will send the scroll off to be ceramic coated, which (if applied internally) should decrease heat transfer to the scroll, and decrease the equilibrium temperature of the scroll.

The vibration we mentioned was with reference to the starter assembly. If slightly misaligned, the vibration appears. We've since modified the engagement to make it less susceptible to misalignment. No more vibrating starter.

All is going well

Thanks!

- Feathers

[Feathers]

So with some further tweaking of the mechanicals and a lot of programming, it starts itself!

The control firmware basically has five states:

Idle - Just hanging out waiting for a start command, default state if engine is cold. If any other state reaches an out-of-bounds condition and cannot correct it, and the engine is still cold, we return to idle.

Heating - If a start is requested, but EGT is less than 800*F and oil temperature is less than 100*F, engage the starter motor and burn propane. A minimum heating period of 30 seconds is enforced to ensure ignition, starter engagement, and evaporator heating.

Starting - After the heating period is finished, begin stepping up fuel flow and starter speed by conditional "milestones" of engine speed. Disengage starter when above 22000 RPM, and abort if EGT is too high or any step takes too long.

Running - Pretty simple. Just monitor for out-of-bounds conditions and respond to requested throttle position.

Cooling - If EGT is greater than 250*F or oil temperature is greater than 130*F, engage the starter and run cold air through. Otherwise, return to idle. If any state is aborted or stopped, and the engine is hot, it will return to cooling before idle.

We use a raspberry pi with a touchscreen and an Xbee wireless module to communicate with the FADEC on the kart. Just type "start" and it starts . Type anything else, and it returns to idle or cooling. The rest of the time we use this "telemetry station" to monitor telemetry packets transmitted by the FADEC in real-time. Its very handy.

Here is a video of the automated start:



Thanks for the help guys.

- Feathers

[racket]

Hi Feathers

Very impressive :-)

Now the freepower and some shaft horsepower .

Cheers
John

[Feathers]

Indeed John

We've been working on the freepower section. The shaft tunnel is finished:



We've ordered some 6.5" ID, 8.5" OD mild steel tubing from which we can machine our turbine shroud. It should be here toward the middle of next week.

Also finished the design of the power turbine plenum, and have sent the pieces off in flat patterns to be laser cut:









In the meantime, we're tweaking the autostart feature of our engine. Everything is working well, except even in sub-zero weather, our oil system is lacking capacity.

Additionally, changes in oil viscosity and the associated fluid drag changes our idle engine speed significantly, for a given idle fuel flow. I think I will write in some closed-loop control to adjust fuel flow and maintain idle at exactly 30,000 RPM, right where we want it.

We've begun saving and plotting our telemetry, as I suppose it might mean something now.

Here is a recent test, and the associated telemetry plotted (sorry for the odd aspect ratio of the chart):





- Feathers

[racket]

Hi Feathers

You've got a well controlled engine there :-)

LOL,...............it looks a little cold there with all that snow on the ground , yep , your idling rpm will be affected by the oil temp .

Can't wait to see your next additions .

Cheers
John