Been a while - new project, T04 Gas Turbine.

[marksteamnz]

Thanks for the info Ash, that is a nice furnace.

The very first casting I produced after building my refractory/furnace was an investment casting, but that is a stretch - it was just a piece of styrofoam cut out in the shape of a "Z", about 2" X 2" X 3/4" with two styrofoam gates - buried it in some silica sand and poured the metal. The part came out but as you can imagine, styrofoam and silica sand leaves a lot to be desired in the finish department.

Very nice work. Most impressed at the results
Is the fan on your furnace a heat gun or hair drier?

That's just a hairdryer. :)

[Fricke]

Hi Ash

How is work going on your nice engine?

One design question... (Yes I´ve had a lot of time thinking of it lately, while I´m traing for my license as a tigwleder)

Why are you extending the shaft of the turbine? Why not put the CC behind the turbine and have a shorter shaft? Like the PAL and HAWK turbines?

You are a inspiration!
And I hope that You´ll get the PWM working, if not... I have a working code for the PIC18F2620 on the controller for my jetkart. If you want it... let me know.

[Ash Powers]

Fricke,

Work is coming along nicely. A few setbacks, timewise, with getting the turbine shaft back from my turbo builder. The postal service apparently didnt' want to drop it off to him so he had to go pick it up himself. My order through Grizzly for some tooling also came incomplete which has stalled the completion of the fuel ring and bearing lubrication system.

I am extending the shaft because I'm not using a reverse-flow combustion chamber. Trying to minimize losses incurred from shuttling the flow of gases in so many directions as well as simplify the design. The design of the Phoenix 30.3 has been instrumental in the construction of my current GT - it too uses a radial flow turbine with the shaft extended to accomodate the layout. That engine uses a slightly smaller compressor and turbine than I am and it produced 25lbs of thrust comfortably and for long durations of time. I'm hoping to get somewhere in that ballpark figure, which will do just fine for the scooter I plan to put it on.

As of last night I finally got all of the PWM working properly. I found a different schematic for the SG3525 here:

http://forums.nasioc.com/forums/showthread.php?t=722841

and setup the IC on the protoboard with most of that layout. Took a good bit of time getting the software put together on the PDA to control the analog outputs of the DAQ device but as of late late last night she is up and running. :)

Not having everything here to work on the mechanical bits of the project isnt' so bad though - the electronics side of things needed some attention and I've been working on the schematics for the sensor/signal conditioner/amplifiers/power supply today. They are just about finished and I'll start in on designing the actual PCB shortly. I hope to have the board design done and submitted to the manufacfturer by tomorrow so they can get to producing a couple and have them to me by mid-week. I'm pretty excited about the DAQ side of this project as it can be used for far more applications than running gas turbines. Adapting this setup for use in an automobile for various monitoring/control wont be terribly difficult to do and likely quite popular amongst the tuner crowd.... :)

[Fricke]

Work is coming along nicely. A few setbacks, timewise, with getting the turbine shaft back from my turbo builder. The postal service apparently didnt' want to drop it off to him so he had to go pick it up himself. My order through Grizzly for some tooling also came incomplete which has stalled the completion of the fuel ring and bearing lubrication system.

Well it´s the same here to... the postal is postal! =)

I am extending the shaft because I'm not using a reverse-flow combustion chamber. Trying to minimize losses incurred from shuttling the flow of gases in so many directions as well as simplify the design. The design of the Phoenix 30.3 has been instrumental in the construction of my current GT - it too uses a radial flow turbine with the shaft extended to accomodate the layout. That engine uses a slightly smaller compressor and turbine than I am and it produced 25lbs of thrust comfortably and for long durations of time. I'm hoping to get somewhere in that ballpark figure, which will do just fine for the scooter I plan to put it on.

Ok... I understand! It makes sence to use a proven design... But check out the HAWK engine... It´s a nice piece of engineering... They ditched the automotive compressor and made their own... and acheve greater PRs that way...

As of last night I finally got all of the PWM working properly. I found a different schematic for the SG3525 here:

http://forums.nasioc.com/forums/showthread.php?t=722841

and setup the IC on the protoboard with most of that layout. Took a good bit of time getting the software put together on the PDA to control the analog outputs of the DAQ device but as of late late last night she is up and running. :)

Thats good news... I´ve been looking for a simple PWM control option for a project that I have... I´m trying to mate a electomechanical brake to my side car... =) And dont feel to use a PIC controller for that... need foolproof^2 for that... =)

Not having everything here to work on the mechanical bits of the project isnt' so bad though - the electronics side of things needed some attention and I've been working on the schematics for the sensor/signal conditioner/amplifiers/power supply today. They are just about finished and I'll start in on designing the actual PCB shortly. I hope to have the board design done and submitted to the manufacfturer by tomorrow so they can get to producing a couple and have them to me by mid-week. I'm pretty excited about the DAQ side of this project as it can be used for far more applications than running gas turbines. Adapting this setup for use in an automobile for various monitoring/control wont be terribly difficult to do and likely quite popular amongst the tuner crowd.... :)

That DAC seems to be a nice little gadget! I´ll have to look deeaper into it later this spring... I´m going to turbo convert my CJ 750 side valve bikeengine... currently I´m looking for two small IHI RB3 turbos... and later it´s going to have EFI... but for now I´ll use two MC CV carbs... and see if the engine survive the turbo transplant =)

//Fredrik

[Ash Powers]

Finally finished up with the PCB layout and am pretty happy with how it turned out. I also integrated a small power supply section to the board to produce a regulated 12VDC and 5VDC for the ICs, the DAQ device, and the bluetooth RS-232 adapter. There is a screwdown terminal block for connecting all of the IO that isn't integral to the board and the two thermocouple CJC and amplifiers have their own screwdown terminals local to the CJC/amplifier to reduce the difference of localized thermal conditions between the compensator and the cold junction itself. I'm going to have ten of these made and should have them by next weekend.



This board has the following:

12VDC Regulated Power, 1A
5VDC Regulated Power, 1A

(2) 0-100psi self-amplified, temperature compensated pressure sensors
(2) 0-100% duty cycle PWM generators with POT adjustable base frequency (Speed Controllers for fuel pump and starter motor)
(1) Frequency to Voltage Converter with input pulse divide by ten (660Hz/V - 198,000RPM MAX) (For optical tacho input)
(1) Frequency to Voltage Converter (66Hz/V - 19,800RPM MAX) (For speed sensor input)
(2) Cold Junction Compensated and Amplified Thermocouple Inputs w/integral POT for calibration (5mV/C - 1000C MAX) (TIT and TOT)
(1) Digital Output (Igniter Circuit)
(2) Digital Inputs (Switched Control - possible afterburner triggering)
(1) Analog to Digital input 0-5V (will be used in the GT for throttle position input)
(2) Digital to Analog outputs (0-4.3V output)

All in a convenient footprint of 5.6" X 3.1" (142mm X 79mm)
:)

[Fricke]

Finally finished up with the PCB layout and am pretty happy with how it turned out.

Nice work!

All in a convenient footprint of 5.6" X 3.1" (142mm X 79mm)
:)

That´s quite a small footprint! Don´t forget to make room for mounting holes! =)

//Fredrik

[pezman]

These TI microcontrollers are incredibly cool and it's hard to beat the prove of the development kit.

May not be quite up to the task that the DAQ will need to handle, but I'm sure that someone here can find some kind of use for them.

http://focus.ti.com/mcu/docs/mcuprodoverview.tsp?
sectionId=95&tabId=140&familyId=342

[Ash Powers]

Finally received the turbine/shaft back from my turbo builder with the bearing journals precision ground to fit the 10mm ceramic bearings - fitment is good and I also received the tooling to make the shaft extension.

I'm seeking some advice on what mateiral I should use for the turbine shaft though - I'm not sure if CRS can be used for this application or if there is some other material that is specifically recommended for this app.

Any advice is appreciated!

[Zippiot]

I thought plain tool steel was most common.

[Ash Powers]

Do you know what grade of tool steel to use? There are a ton of different grades out there.

41L40 and 4340 also have excellent properties that seem suitable for producing a shaft from.. Any thoughts on that?

[jthompso]

M2 High Speed Tool Steel is the way to go--it's hard to machine (very difficult to get a fine surface finish) but it has the material properties that you need in a turbine shaft. 41L40 and 4340 would actually be poor choices--they are tough metals, but for a turbine shaft you don't want a tough material. Toughness is a measure of the amount of elastic deformation that can occur in a material (meaning that it will temporarily deform when a load is applied). It is absolutely essential that your shaft doesn't bend because it may set up an oscillation that will end up destroying the shaft. M2 is not a "tough" material, it will not bend much before it hits its tensile strength and snaps.

It's counterintuitive but M2 High Speed Steel is the better choice for your shaft, also the material keeps its strength at high temperatures making it an even better choice.

If you need any machining tips let me know, I used the material to make my turbine shaft. I learned by trial and error, after two shafts that I screwed up my third one (which I don't have pictures of in my build thread yet) turned out perfectly.

Sorry for being so verbose but it's nice to be able to apply what I'm learning

[Zippiot]

So we are looking for tensile, not yeild strength. Correct me but ductile iron has a higher tensile strength than most steels, surely not more than tool or high speed steels but...

What about quartz? :) just kidding

[jthompso]

Not exactly, you can't just look for one property when designing something, you have to look at the general trend. But for a turbine shaft you want a material with a high modulus of elasticity (stiffness) and a high yield strength.

The modulus of elasticity (E) determines the ductility of a material-lower values are more ductile.

The yield strength is a point marking the end of the region of plastic deformation in a material--in other words, after the yield strength is exceeded the part will be permanently deformed.

Also, the reason I mentioned the tensile strength in the first place is that in general with a stiff material there isn't much of a difference between the yield and (ultimate) tensile strength. When you hit the yield strength the material will break shortly afterwards.

[multispool]

Guys,

Using the right grade of steel for the shaft is important for safety reasons.
In the past, High tensile steel has failed, resulting in the turbine being ejected at very high velocity. Just like defective wheels, failures happen suddenly at any time and not necesarily on the first test run or at maximum RPM.

Turbine shafts in our engines are now made only from EN24T. This is the
result of much testing (and failures when using other types) and it is advised that you don’t use any alternatives.
The British standard classification of EN24T is:
[quote]BS. 970 1991 817M40 Heat treated.
A high quality alloy steel giving good wear resistance combined with ductility
and shock resistance properties. Supplied hardened and tempered to ‘T’
condition – 55-65 T.S.I.
International name 817M40T. Identification colour “Brownâ€

[Ash Powers]

Given that one of the primary factors of consideration is the geometry of the shaft (length, diameter) as well as distances from bearing to center of mass of the turbine wheel and compressor wheel, it would make sense to depart from the conventional 2-bearing setup and employ additional bearing(s) for a total of three or more depending on design. You wouldn't need to use such a thick shaft either (which adds a bit of extra inertia) and although it will introduce a couple more parts to provide sprung bearing load, it certainly is feasable.

Not many people have built axial compressor gas turbines in our scale but I would be inclined to think it would be absolutely necessary to employ more than two bearings to properly support the rotating group. I'm starting to lean in this direction for my GT build given that trying to make a shaft extension from tool steel is going to prove to be considerably more challenging than just building a conventional shaft as is used in the commercial engines available (as well as homebuilts).

Last thing I want on my testbench is a grenade....