Been a while - new project, T04 Gas Turbine.
Moderator: Mike Everman
-
- Posts: 503
- Joined: Sat Jul 23, 2005 11:11 pm
- Antipspambot question: 0
- Location: Australia
-
- Posts: 5007
- Joined: Fri Oct 31, 2003 7:25 am
- Antipspambot question: 0
- Location: santa barbara, CA
- Contact:
Oops, of course this is directed to Ash...Mike Everman wrote:Ah, another installment in my favorite turbojet thread! Though I'm not responding just to say that,
I'm on the right coast for the week doing business, and will be driving by Edgewater sometime Thursday night, driving from Emelia island plantation and resort to Highland Beach. Can I come by and see your toys for a short visit? PM or email me if so! (I'm currently in MA, middle of whirlwind tour.)
Mike Often wrong, never unsure.
__________________________
__________________________
-
- Posts: 176
- Joined: Sat Sep 04, 2004 7:17 am
- Antipspambot question: 0
- Location: Edgewater, FL
- Contact:
-
- Posts: 5007
- Joined: Fri Oct 31, 2003 7:25 am
- Antipspambot question: 0
- Location: santa barbara, CA
- Contact:
Mike visits Ash
Hey, Ash, great visit. Tipping beers and talking jets, very good!
Oooh, I just looked at the photo. I'm looking quite sweaty. Us SoCal guys aren't used to the Florida humidity!
Thanks for your hospitality, and say hi to Andrew and Jim for me!
Oooh, I just looked at the photo. I'm looking quite sweaty. Us SoCal guys aren't used to the Florida humidity!
Thanks for your hospitality, and say hi to Andrew and Jim for me!
- Attachments
-
- ash and mike oct 07.JPG (46.18 KiB) Viewed 19743 times
Mike Often wrong, never unsure.
__________________________
__________________________
-
- Posts: 176
- Joined: Sat Sep 04, 2004 7:17 am
- Antipspambot question: 0
- Location: Edgewater, FL
- Contact:
Hi Mike!
It was a pleasure having you and Rick over - I'm still amazed at how small the world can appear at times. :) BTW, wonderful robotic products - quite the engineering marvel! Thanks for sharing that with us!
As you saw when you were here, I have finalized the turbine exducer plate in the 316SS. I have a few modifications I am pursuing given the new plate. Here's the scoop:
Pic of the new fuel manifold using the 0.050" O.D., 0.033" I.D. SS tubes for the injectors, supported by the 0.125" O.D. SS tubes. I imagine this setup will stand the test of time. :)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-FUELMANIFOLD01.JPG.jpg)
To address the issue of the NGV rear plate retaining bolts I have decided to weld the inner NGV plate to the outer NGV plate to make it a solid assembly. This will eliminate the need for the retaining bolts which suffer from exposure to the combustion gases and end up stretching out.
The idea here is to weld each of the vanes at their interface to the NGV outer plate at the back edge. The problem with doing this previously was that the center bore for the NGV outer plate was about 0.015" smaller in diameter than the turbine wheel itself. The NGV assembly had to be two pieces in order to assemble it. So the solution: machine out the NGV rear plate bore just enough such that the turbine will slip through it and re-machine a new exducer flange to mate the modified NGV plate.
I purchased several 316SS "pucks" of various sizes and the original mild steel exducer flange was showing serious signs of thermal exposure from surface flaking to blistering, so after opening the NGV rear plate bore ~0.020", I machined a new exducer flange to fit - 316SS this time too. :)
NGV inner and outer plates. (These will be welded together at the outer edge where the vanes meet the NGV outer plate.)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-NGVASSY01.JPG.jpg)
Here is the new 316SS exducer flange. I also increased the base flange thickness as the puck I used was thicker than the original mild steel plate.
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-TURBEXD01.JPG.jpg)
I am also eliminating the jetpipe flange setup. I originally made the exducer flange and jet pipe as two pieces so that I could easily modify the jet pipe. This time around I will have the jet pipe TIG welded to the exducer flange to integrate it as one piece and eliminate the extra set of bolts. (Still need to taper the jetpipe down)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-TURBEXD02.JPG.jpg)
Because the NGV inner and outer plate will be connected as one piece, I will also be able to hold a little tighter tolerances between the turbine wheel and exducer flange. Previously I was running at about 0.030" radial clearance and the new flange has cut that back to ~0.020" which should improve efficiency as well.
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-TURBEXD03.JPG.jpg)
I am also going to integrate a turbine hub "bullet" to aid the gasflow coming off the back of the turbine wheel. The bullet will be held in place by three SS plates which will also serve as a sort of flow straightener. The part shown was something I hacked up for kicks but the final part will be a linear cone. I need to get some SS bar stock to make the final part from..
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-TURBSWIRL01.JPG.jpg)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-TURBSWIRL02.JPG.jpg)
I also spent some time about a month ago machining a housing for the starter. I have done some more machining to integrate a support bearing at the front flange of the housing as well as began machining the starter shaft itself. I'm not going to use a bendix assembly for this - the idea is to allow the starter motor to slide about 1/4"; using a lever arm to engage it to the compressor nut and a return spring to disengage the motor once the engine has reached self-sustaining RPMs. I've alreaedy got the second PWM output on the GT Control board to operate the starter motor as well as the MOSFET bank for it, so once the hardparts are finished it will just be a matter of finishing the auto-start sequences within the programming.
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-STARTER01.JPG.jpg)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-STARTER02.JPG.jpg)
I've also been looking at the process of anodizing aluminum; it actually seems to be a relatively simple process too. I've found several resources for the materials needed (sulphuric acid, dyes, etc) and since I already have an electroplating system, the power supply and other materials are pretty much "on hand". Will be pretty cool to apply such a resilient surface treatment to the parts as well as add a cosmetic touch. :)
It was a pleasure having you and Rick over - I'm still amazed at how small the world can appear at times. :) BTW, wonderful robotic products - quite the engineering marvel! Thanks for sharing that with us!
As you saw when you were here, I have finalized the turbine exducer plate in the 316SS. I have a few modifications I am pursuing given the new plate. Here's the scoop:
Pic of the new fuel manifold using the 0.050" O.D., 0.033" I.D. SS tubes for the injectors, supported by the 0.125" O.D. SS tubes. I imagine this setup will stand the test of time. :)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-FUELMANIFOLD01.JPG.jpg)
To address the issue of the NGV rear plate retaining bolts I have decided to weld the inner NGV plate to the outer NGV plate to make it a solid assembly. This will eliminate the need for the retaining bolts which suffer from exposure to the combustion gases and end up stretching out.
The idea here is to weld each of the vanes at their interface to the NGV outer plate at the back edge. The problem with doing this previously was that the center bore for the NGV outer plate was about 0.015" smaller in diameter than the turbine wheel itself. The NGV assembly had to be two pieces in order to assemble it. So the solution: machine out the NGV rear plate bore just enough such that the turbine will slip through it and re-machine a new exducer flange to mate the modified NGV plate.
I purchased several 316SS "pucks" of various sizes and the original mild steel exducer flange was showing serious signs of thermal exposure from surface flaking to blistering, so after opening the NGV rear plate bore ~0.020", I machined a new exducer flange to fit - 316SS this time too. :)
NGV inner and outer plates. (These will be welded together at the outer edge where the vanes meet the NGV outer plate.)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-NGVASSY01.JPG.jpg)
Here is the new 316SS exducer flange. I also increased the base flange thickness as the puck I used was thicker than the original mild steel plate.
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-TURBEXD01.JPG.jpg)
I am also eliminating the jetpipe flange setup. I originally made the exducer flange and jet pipe as two pieces so that I could easily modify the jet pipe. This time around I will have the jet pipe TIG welded to the exducer flange to integrate it as one piece and eliminate the extra set of bolts. (Still need to taper the jetpipe down)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-TURBEXD02.JPG.jpg)
Because the NGV inner and outer plate will be connected as one piece, I will also be able to hold a little tighter tolerances between the turbine wheel and exducer flange. Previously I was running at about 0.030" radial clearance and the new flange has cut that back to ~0.020" which should improve efficiency as well.
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-TURBEXD03.JPG.jpg)
I am also going to integrate a turbine hub "bullet" to aid the gasflow coming off the back of the turbine wheel. The bullet will be held in place by three SS plates which will also serve as a sort of flow straightener. The part shown was something I hacked up for kicks but the final part will be a linear cone. I need to get some SS bar stock to make the final part from..
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-TURBSWIRL01.JPG.jpg)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-TURBSWIRL02.JPG.jpg)
I also spent some time about a month ago machining a housing for the starter. I have done some more machining to integrate a support bearing at the front flange of the housing as well as began machining the starter shaft itself. I'm not going to use a bendix assembly for this - the idea is to allow the starter motor to slide about 1/4"; using a lever arm to engage it to the compressor nut and a return spring to disengage the motor once the engine has reached self-sustaining RPMs. I've alreaedy got the second PWM output on the GT Control board to operate the starter motor as well as the MOSFET bank for it, so once the hardparts are finished it will just be a matter of finishing the auto-start sequences within the programming.
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-STARTER01.JPG.jpg)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/ZZZZZ-STARTER02.JPG.jpg)
I've also been looking at the process of anodizing aluminum; it actually seems to be a relatively simple process too. I've found several resources for the materials needed (sulphuric acid, dyes, etc) and since I already have an electroplating system, the power supply and other materials are pretty much "on hand". Will be pretty cool to apply such a resilient surface treatment to the parts as well as add a cosmetic touch. :)
Excellent turbine work....
Hey Ash,
Love your work, really well explained and the pics are just great, keep it up, be sure to post a video of the engine running too!
I'm in the process of putting together the bits'n pieces to build a small "afterburning" turbine, the engine itself will be in the ~18lb thrust range which will be finally fitted to a 1/6th or 1/5th scale jet dragster, the engine, A/B setup and final testing "should" be finished within 8 weeks or so, then construction of the car itself will commence. Will keep everybody informed as I progress.
Jet dragsters are one of my passions, unfortunately I don't have the resources to own a real one so this is the next best thing!
Cheers from down under,
Smithy.
Love your work, really well explained and the pics are just great, keep it up, be sure to post a video of the engine running too!
I'm in the process of putting together the bits'n pieces to build a small "afterburning" turbine, the engine itself will be in the ~18lb thrust range which will be finally fitted to a 1/6th or 1/5th scale jet dragster, the engine, A/B setup and final testing "should" be finished within 8 weeks or so, then construction of the car itself will commence. Will keep everybody informed as I progress.
Jet dragsters are one of my passions, unfortunately I don't have the resources to own a real one so this is the next best thing!
Cheers from down under,
Smithy.
When I die, I want to go peacefully in my sleep like my Grandfather.....not screaming like the passengers in his car!!
-
- Posts: 176
- Joined: Sat Sep 04, 2004 7:17 am
- Antipspambot question: 0
- Location: Edgewater, FL
- Contact:
Thanks for the feedback! :)
I've always been fond of the turbojet with A/B dragsters - would really like to see a turboshaft engine put to use in one of those dragsters rather than using purejet. I would imagine producing at least what the top fuel engines produce wouldn't be that difficult. They are making appx 8000 horsepower from a piston engine.
I've been hammering away on the engine a bit more - doing some of the things I originally wanted to do but held off on for sake of getting her up and running and working out the bugs.
I have notched the rear NGV plate where the blades meet it in order to allow the weld to grab onto more material - had the vanes TIG welded this evening. Appears to have good penetration - the part is quite solid. I chucked the part into the lathe and cut back the welds to allow room for the CC outer liner to slip over.
Now I dont have to worry about the retaining bolts heating and stretching and the assembly will hold much tighter tolerances. The new exducer flange is snugging up the gap to the blade tips so I'm looking forward to better efficiencies.
The only question now is whether these welds will hold the forces at work on them. With a 5" dia casing (19.6 square inch area) and ~20psi of case pressure, there is effectively ~390 lbs of force trying to push the guts out the front. That load is retained through the bearing tube to the NGV inner plate, and then through the vanes to the outer NGV plate. With 18 vanes, each is carrying around 22lbs of force. Pretty amazing that the three small bolts I was previously using lasted as long as they did! Fortunately there is a bit of cooling air flowing along the inner wall of the outer CC liner and that should help keep the welds a little cooler - that area of the outer NGV plate is pretty thin and even considering, it doesn't appear to have been exposed to as much heat as you may think. Hopefully this approach will work - I'll have to do some test runs and disassemble the engine to check how the welds are holding up.
I also have an idea of welding some (3-4) ~0.25" thick plates on the outside of the main casing at the front that will stick forwards about 3/4" or so from the diffuser cover flange - these plates will allow me to bolt a small bulkhead to it for holding some relatively stiff springs against the compressor diffuser plate to assist in retaining the case pressure forces. *Shrug* - never thought I'd be contending with this problem. :)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/A-WELDEDNGV01.JPG.jpg)
I took a new jetpipe with me to the welder to have him weld the seams from the cutouts where the pipe tapers down but he was using far too much heat on it and the burn through was hideous. I made a new jetpipe and use my old trusty MIG loaded with some 316LSi wire and welded it up myself and ground them down for a nice finish.
I did take the turbine hub bullet with me and had the arms welded to it. I spent a good bit of time doing some sculpting of the welds freehanded with the Dremel and die grinder loaded with a carbide burr. The piece turned out pretty nicely and I have tacked it into the jetpipe - will go to the TIG welder tomorrow to have it final welded inside the jetpipe and have the jetpipe TIG'd to the exducer flange. Here's the shots:
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/B-NEWJETPIPE01.JPG.jpg)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/B-NEWJETPIPE02.JPG.jpg)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/B-NEWJETPIPE03.JPG.jpg)
Last night I spent a good bit of time on the starter. I had previously bored a seat for a ballbearing into the front of the electric motor housing to provide better support against the shaft wanting to whip. I drilled out the retaining screw holes, tapped them and secured the bearing.
I then proceeded to machine the shaft for the starter. I used some W2 tool steel I had laying around - was the only barstock I had that was large enough in dia. I machined the bore for the electric motor shaft to insert into - nice fitment too - very slight interference fit. Then machined the shaft down to fit into the bearing - nice fitment there too - no interference - the shaft slides nice and smoothly in the bearing's i.d. Then turned the part around in the chuck and machined her out with a nice o.d. taper and a shoulder for the bearing. The i.d. of the shaft was then machined out to ~0.5" dia about 3/4" deep.
I then put a drill bit into the lathe chuck facing backwards and put a 3/4" long piece of 5/16" i.d. 5/8" o.d. rubber fuel line hose onto the bit and spun up the lathe. I then used a dremel with a sanding drum to sand the fuel line down to around 0.550" o.d. I applied some 3M contact cement onto the rubber hose and into the bore of the shaft and inserted the hose into the shaft. I chucked the shaft into the lathe, rubber end out, and used my dremel with a round grinding stone to shape the rubber to an approximate fit with the aluminum compressor nut.
Set screw holes were drilled and tapped - two of them, 180-degrees apart. The main starter housing had a hole bored into it to provide access to the set screws and I assembled everything.
The electric motor has a flat ground into one side of its shaft for a set screw - by using two set screws, the set screw in the side that isn't ground sticks out a little further - this was intentional for the sake of the fact if the shaft decides to back off from the electric motor, it cannot come completely out of the starter housing as the set screw will interfere with the ball bearing. Last thing I want are pieces flying into the compressor!
I also had to machine a new cap for the main housing to provide room for the motor to slide back and forth. The design I have gone with doesn't use a bendix assembly on the shaft - instead, I am allowing the electric motor/shaft to slide inside the main housing. There is a return spring inside the housing at the front of the electric motor and the cap has a "button" that can be depressed to push the starter to the compressor nut. I made this button shaft a little longer than needed as I may end up winding my own solenoid and making a "rocker" for pushing the starter pin and making it electronically controlled - a new addition for a later time.
Everything turned out really well with the starter - the shaft has no (detectable) radial slop in it so there will be no chance for shaft whipping. Hopefully the rubber hose approach will suffice - I feel confident it will though. :)
(The starter shaft looks black in these images - that's just a photographic artifact)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/C-STARTER01.JPG.jpg)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/C-STARTER02.JPG.jpg)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/C-STARTER03.JPG.jpg)
Tomorrow morning I'll meet with the welder to finish up the parts and get to work on the coding for the software - I need to finish up the code to control the second speed controller for the starter motor and setup the sequencing for startup. I need to get my hands on some stainless bar stock so I can machine out a glow plug boss for ignition - I intend to have the electronics setup for a "pretty much" auto-start.
I bought a manual valve assembly for attaching to the small propane bottles used for brazing torches and will be making a mount to hold one of these tanks at the front of the scooter frame. It would be nice to have an electrically controlled variable-flow valve, but I'm sure something like that would go for a pretty penny so I'll just have to do some manual work for the startup - no big deal though.
I started on the startup sequencing when I originally wrote the program - the idea is as follows:
(Manual) Crack open the propane.
(Manual) Hold starter engaged to rotating group.
Initiate startup in program.
(program) Turn starter motor on - ramping up from 0% to ~10% over a few seconds.
(program) Turn glow plug on
(program) Wait for rise in EGT, once EGT rises THEN
(program) Turn off glowplug and ramp starter motor up to 100% over a "x" seconds duration.
(Manual) Increase propane flow as starter ramps up to 100%
(program) At "x" RPM, ramp fuel pump up to "x" percent (idle flow rate) over "x" time.
(Manual) Back off propane feed to close while liquid fuel delivery comes up.
(program) At "x" RPM, MESSAGE "Started!" and hold idle fuel flow until throttle position sender is manally adjusted up to holding percent. At sender match, exit startup sequence.
All the while, the built-in thermal limits will be in effect to prevent overheating of the turbine. All of the "x" variables above are definable at the PDA level - they aren't hard-coded so I will be able to easily adjust the sequence parameters until I get it fine tuned.
It may not be necessary to have a variable flow valve on the propane - that is the only part that would make building this completely hands-free auto-start uneconomical. Does anyone know how this is achieved with the model aircraft turbines?
I've always been fond of the turbojet with A/B dragsters - would really like to see a turboshaft engine put to use in one of those dragsters rather than using purejet. I would imagine producing at least what the top fuel engines produce wouldn't be that difficult. They are making appx 8000 horsepower from a piston engine.
I've been hammering away on the engine a bit more - doing some of the things I originally wanted to do but held off on for sake of getting her up and running and working out the bugs.
I have notched the rear NGV plate where the blades meet it in order to allow the weld to grab onto more material - had the vanes TIG welded this evening. Appears to have good penetration - the part is quite solid. I chucked the part into the lathe and cut back the welds to allow room for the CC outer liner to slip over.
Now I dont have to worry about the retaining bolts heating and stretching and the assembly will hold much tighter tolerances. The new exducer flange is snugging up the gap to the blade tips so I'm looking forward to better efficiencies.
The only question now is whether these welds will hold the forces at work on them. With a 5" dia casing (19.6 square inch area) and ~20psi of case pressure, there is effectively ~390 lbs of force trying to push the guts out the front. That load is retained through the bearing tube to the NGV inner plate, and then through the vanes to the outer NGV plate. With 18 vanes, each is carrying around 22lbs of force. Pretty amazing that the three small bolts I was previously using lasted as long as they did! Fortunately there is a bit of cooling air flowing along the inner wall of the outer CC liner and that should help keep the welds a little cooler - that area of the outer NGV plate is pretty thin and even considering, it doesn't appear to have been exposed to as much heat as you may think. Hopefully this approach will work - I'll have to do some test runs and disassemble the engine to check how the welds are holding up.
I also have an idea of welding some (3-4) ~0.25" thick plates on the outside of the main casing at the front that will stick forwards about 3/4" or so from the diffuser cover flange - these plates will allow me to bolt a small bulkhead to it for holding some relatively stiff springs against the compressor diffuser plate to assist in retaining the case pressure forces. *Shrug* - never thought I'd be contending with this problem. :)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/A-WELDEDNGV01.JPG.jpg)
I took a new jetpipe with me to the welder to have him weld the seams from the cutouts where the pipe tapers down but he was using far too much heat on it and the burn through was hideous. I made a new jetpipe and use my old trusty MIG loaded with some 316LSi wire and welded it up myself and ground them down for a nice finish.
I did take the turbine hub bullet with me and had the arms welded to it. I spent a good bit of time doing some sculpting of the welds freehanded with the Dremel and die grinder loaded with a carbide burr. The piece turned out pretty nicely and I have tacked it into the jetpipe - will go to the TIG welder tomorrow to have it final welded inside the jetpipe and have the jetpipe TIG'd to the exducer flange. Here's the shots:
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/B-NEWJETPIPE01.JPG.jpg)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/B-NEWJETPIPE02.JPG.jpg)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/B-NEWJETPIPE03.JPG.jpg)
Last night I spent a good bit of time on the starter. I had previously bored a seat for a ballbearing into the front of the electric motor housing to provide better support against the shaft wanting to whip. I drilled out the retaining screw holes, tapped them and secured the bearing.
I then proceeded to machine the shaft for the starter. I used some W2 tool steel I had laying around - was the only barstock I had that was large enough in dia. I machined the bore for the electric motor shaft to insert into - nice fitment too - very slight interference fit. Then machined the shaft down to fit into the bearing - nice fitment there too - no interference - the shaft slides nice and smoothly in the bearing's i.d. Then turned the part around in the chuck and machined her out with a nice o.d. taper and a shoulder for the bearing. The i.d. of the shaft was then machined out to ~0.5" dia about 3/4" deep.
I then put a drill bit into the lathe chuck facing backwards and put a 3/4" long piece of 5/16" i.d. 5/8" o.d. rubber fuel line hose onto the bit and spun up the lathe. I then used a dremel with a sanding drum to sand the fuel line down to around 0.550" o.d. I applied some 3M contact cement onto the rubber hose and into the bore of the shaft and inserted the hose into the shaft. I chucked the shaft into the lathe, rubber end out, and used my dremel with a round grinding stone to shape the rubber to an approximate fit with the aluminum compressor nut.
Set screw holes were drilled and tapped - two of them, 180-degrees apart. The main starter housing had a hole bored into it to provide access to the set screws and I assembled everything.
The electric motor has a flat ground into one side of its shaft for a set screw - by using two set screws, the set screw in the side that isn't ground sticks out a little further - this was intentional for the sake of the fact if the shaft decides to back off from the electric motor, it cannot come completely out of the starter housing as the set screw will interfere with the ball bearing. Last thing I want are pieces flying into the compressor!
I also had to machine a new cap for the main housing to provide room for the motor to slide back and forth. The design I have gone with doesn't use a bendix assembly on the shaft - instead, I am allowing the electric motor/shaft to slide inside the main housing. There is a return spring inside the housing at the front of the electric motor and the cap has a "button" that can be depressed to push the starter to the compressor nut. I made this button shaft a little longer than needed as I may end up winding my own solenoid and making a "rocker" for pushing the starter pin and making it electronically controlled - a new addition for a later time.
Everything turned out really well with the starter - the shaft has no (detectable) radial slop in it so there will be no chance for shaft whipping. Hopefully the rubber hose approach will suffice - I feel confident it will though. :)
(The starter shaft looks black in these images - that's just a photographic artifact)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/C-STARTER01.JPG.jpg)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/C-STARTER02.JPG.jpg)
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/Updates/images/C-STARTER03.JPG.jpg)
Tomorrow morning I'll meet with the welder to finish up the parts and get to work on the coding for the software - I need to finish up the code to control the second speed controller for the starter motor and setup the sequencing for startup. I need to get my hands on some stainless bar stock so I can machine out a glow plug boss for ignition - I intend to have the electronics setup for a "pretty much" auto-start.
I bought a manual valve assembly for attaching to the small propane bottles used for brazing torches and will be making a mount to hold one of these tanks at the front of the scooter frame. It would be nice to have an electrically controlled variable-flow valve, but I'm sure something like that would go for a pretty penny so I'll just have to do some manual work for the startup - no big deal though.
I started on the startup sequencing when I originally wrote the program - the idea is as follows:
(Manual) Crack open the propane.
(Manual) Hold starter engaged to rotating group.
Initiate startup in program.
(program) Turn starter motor on - ramping up from 0% to ~10% over a few seconds.
(program) Turn glow plug on
(program) Wait for rise in EGT, once EGT rises THEN
(program) Turn off glowplug and ramp starter motor up to 100% over a "x" seconds duration.
(Manual) Increase propane flow as starter ramps up to 100%
(program) At "x" RPM, ramp fuel pump up to "x" percent (idle flow rate) over "x" time.
(Manual) Back off propane feed to close while liquid fuel delivery comes up.
(program) At "x" RPM, MESSAGE "Started!" and hold idle fuel flow until throttle position sender is manally adjusted up to holding percent. At sender match, exit startup sequence.
All the while, the built-in thermal limits will be in effect to prevent overheating of the turbine. All of the "x" variables above are definable at the PDA level - they aren't hard-coded so I will be able to easily adjust the sequence parameters until I get it fine tuned.
It may not be necessary to have a variable flow valve on the propane - that is the only part that would make building this completely hands-free auto-start uneconomical. Does anyone know how this is achieved with the model aircraft turbines?
-
- Posts: 176
- Joined: Sat Sep 04, 2004 7:17 am
- Antipspambot question: 0
- Location: Edgewater, FL
- Contact:
Thanks Mag! Nice to hear feedback like that. :) When I look at all these parts or the whole thing put together and whaling at some 120+ dB, it looks just like, ..... a jet engine on a scooter. :) I guess having it here and seeing it all the time kindof takes a bit of that "impact" visual impression that everyone else seems to get from these posts. You gotta understand, there are hundreds of hours in this one project and I've built every bit and piece of it myself - my focus at this point is on trying to make it even better and most of that effort goes on within the grey matter between my ears. :) I "see" the pieces, hear the sounds, smell the jet fuel, and feel the thrust pushing me onward - but working with the parts is like doing math in your head - more thinking, visualizing, imagining to it than anything else. I build it in my mind and then build it with my hands - A Mind--into--Matter affair. :)
I got the jetpipe TIG welded earlier today and it turned out very well. I got an offer from my welder for a TIG box he has had for some time that has been used very very little - he's hard up for cash and willing to part with it for around $800 - it was a $1400 machine new. I'm really considering buying it for these small projects as well as for my business.
I will be putting the engine back together this evening and mounting back to the scooter. I put some time into tidy-ing all the electronics bits up yesterday - housings mounted on the RS232 connectors, bluetooth reciever mounted ot the electronics casing.
I've been constructing the program layout in my mind for the startup sequence and have the logistics nailed down - will be working on writing the code this evening and continuing to hammer away on it. So I've got a healthy dose on my plate for the time being - I'll report back once I have more to share.
I got the jetpipe TIG welded earlier today and it turned out very well. I got an offer from my welder for a TIG box he has had for some time that has been used very very little - he's hard up for cash and willing to part with it for around $800 - it was a $1400 machine new. I'm really considering buying it for these small projects as well as for my business.
I will be putting the engine back together this evening and mounting back to the scooter. I put some time into tidy-ing all the electronics bits up yesterday - housings mounted on the RS232 connectors, bluetooth reciever mounted ot the electronics casing.
I've been constructing the program layout in my mind for the startup sequence and have the logistics nailed down - will be working on writing the code this evening and continuing to hammer away on it. So I've got a healthy dose on my plate for the time being - I'll report back once I have more to share.
Ash,
You sound just like me, planning, thinking, doing, all in your head, then putting it all into material items. I often spend hours laying in bed thinking of better ways to do things, keeps me awake at night too often. I keep a note pad next to my bed to write things down before I forget them, my missus thinks I'm a "nutcase" and totally crazy!! Methinks she's right...!!
I'm very interested in your programing skills, unfortunately I'm an "electronic idiot" but I can usually nut things out if given enough time. I'd love to see some details of your "start-up" sequencing logic and I'd love to learn how to do it myself, can I pick your brains!!??
BTW....love the starter motor, ingenious, well done sir!
I spent some time with "Racketmotorman" (aka John) and his lovely lady on Sunday, he lives not too far from me, he's also completely crazy, but a genius at the same time, he has some great mental and practical skills, his new project engine is a real gem and should run very nicely indeed, can't wait to see it run in person. Had a good look at his free turbine powered motorcycle too, very impressed with his ingenuity and logical thinking. There are some very clever people among these pages.
Cheers,
Smithy.
You sound just like me, planning, thinking, doing, all in your head, then putting it all into material items. I often spend hours laying in bed thinking of better ways to do things, keeps me awake at night too often. I keep a note pad next to my bed to write things down before I forget them, my missus thinks I'm a "nutcase" and totally crazy!! Methinks she's right...!!
I'm very interested in your programing skills, unfortunately I'm an "electronic idiot" but I can usually nut things out if given enough time. I'd love to see some details of your "start-up" sequencing logic and I'd love to learn how to do it myself, can I pick your brains!!??
BTW....love the starter motor, ingenious, well done sir!
I spent some time with "Racketmotorman" (aka John) and his lovely lady on Sunday, he lives not too far from me, he's also completely crazy, but a genius at the same time, he has some great mental and practical skills, his new project engine is a real gem and should run very nicely indeed, can't wait to see it run in person. Had a good look at his free turbine powered motorcycle too, very impressed with his ingenuity and logical thinking. There are some very clever people among these pages.
Cheers,
Smithy.
When I die, I want to go peacefully in my sleep like my Grandfather.....not screaming like the passengers in his car!!
-
- Posts: 176
- Joined: Sat Sep 04, 2004 7:17 am
- Antipspambot question: 0
- Location: Edgewater, FL
- Contact:
OK - bit of an update here....
Finished writing and debugging the startup sequencing - pretty much have it nailed down but I have encountered a few other issues in the process which I'll elaborate on in a moment.
As for the coding, here's what I came up with:
In the Startup parameters configuration page there are four variables for the user to set.
- Initial Starter Duty
- Starter Ramp Speed
- Fuel Ramp Speed
- Idle Fuel Duty
At the initiation of the startup sequence the system will check that the throttle potentiometer is at zero - if so, it will proceed to take a turbine outlet temperature measurement and store it, turn on the starter motor, and turn on the glowplug. The system disables the throttle potentiometer during the startup sequence as well but continues to obtain and monitor the TOT throughout the whole process to prevent overheating - it will shut everything down if TOTs exceed the 850C parameter in the emergency shutdown Settings configuration page.
The initial starter duty specifies what the starter motor is run at, % duty, at the initiation of the startup sequence. It will hold this speed until ignition is detected, which is by way of monitoring the turbine outlet temperature. I have hard coded the program to watch for a 100C increase in TOT - once achieved, the system turns off the glow plug and begins the starter ramp sequence.
The Starter Ramp Speed variable controls how quickly the starter motor ramps up in speed from the initial start duty speed once ignition is detected. The value is a percentage of duty increase to the starter speed at 1 second intervals. i.e. If the start duty is 20% and the ramp speed is 10, it will ramp up to full speed in 8 seconds. During the ramp subroutine, the system is monitoring the TOT to make sure there isn't a flameout. Once the starter has achieved full speed the system takes a TOT reading and monitors it for 5 seconds - if the TOT falls more than 100C in this time, the startup sequence halts and reports a flameout condition. If the temp doesn't fall below that value, it proceeds to the fuel ramp subroutine and leaves the starter motor running at full speed.
Once in the fuel ramp routine, the system will increase the fuel pump duty from zero percent up to the idle fuel duty parameter at a rate specified by the fuel ramp parameter. The fuel ramp parameter is percentage of duty increase per second - just like the starter ramp value. Once the idle fuel duty has been achieved it turns off the starter motor and will hold this fuel rate until the throttle potentiometer has been increased up to the idle fuel rate to match. At this point the startup routine exits and hands fuel control over to the throttle potentiometer.
The startup sequence works rather well - it goes by really quick and since I've got one hand on the starter push-pin to engage it and currently having to put a torch at the jetpipe for ignition AND adjusting the propane feed, it is quite a "hands on" procedure. Not to mention, once the sequence reaches the idle fuel flow, I have to reach over and turn up the throttle potentiometer, then shut down the propane feed.
I also discovered a problem with the starter itself. Something I overlooked in the design of it was the fact that the electric motor has to support a rather big axial load when pushing it in to engage the starter shaft to the compressor nut. I found a small ball bearing and machined a seat for it into the starter pin. This bearing just happened to have an inner diameter just slightly smaller than the electric motor's shaft. So when you press the starter pin inwards, it engages over the armature shaft - thereby eliminating the axial load placed on the electric motor's brass bushings, which were never designed to handle any significant axial load to begin with.
The other issue I have come across with the setup is the sheer amount of current required to run the starter motor with this kind of load on it. This load is causing the MOSFETs driving it to get pretty darn hot. I am using two NTE2985 MOSFETs which are rated up to 30A continuous but it looks like I need to bump up to around 6 MOSFETs and a larger heat sink to spread the load across.
During the startup routine a few times I experienced a flameout due to too little propane feed but I did discover that if I turned the propane feed up high enough to start with I didn't have a flameout nor did I have to change the propane feed rate during the rest of the startup routine. This is nice to know - this means I could get a high pressure solenoid valve for the propane and hook it in after the propane flow valve. The startup sequence could trigger the propane feed automatically and turn off the propane at the end of the sequence without me having to touch any of it.
Since I have spent the time back with all of the programming I think I will go further to make some additions to the code that I wanted to do previously. One of the things I dont have is a means to control how rapidly the fuel pump speed can be changed. As it stands, I have to be rather careful when adjusting the throttle - I can't make any quick drastic changes to it else it will either shoot a huge flame out the tailpipe and kick into the emergency shutdown or it will throttle back too quickly and flameout. I want to integrate a routine to handle the rate of change allowed for the fuel pump speed. Should be an interesting section of code given the nature of the throttle input (it can change at any time) and the goal-seeking nature of the routine. I've pretty much got it nailed down in my head but as I came to find when programming the startup sequencing, it will probably be a challenge in implementation. :)
Finished writing and debugging the startup sequencing - pretty much have it nailed down but I have encountered a few other issues in the process which I'll elaborate on in a moment.
As for the coding, here's what I came up with:
In the Startup parameters configuration page there are four variables for the user to set.
- Initial Starter Duty
- Starter Ramp Speed
- Fuel Ramp Speed
- Idle Fuel Duty
At the initiation of the startup sequence the system will check that the throttle potentiometer is at zero - if so, it will proceed to take a turbine outlet temperature measurement and store it, turn on the starter motor, and turn on the glowplug. The system disables the throttle potentiometer during the startup sequence as well but continues to obtain and monitor the TOT throughout the whole process to prevent overheating - it will shut everything down if TOTs exceed the 850C parameter in the emergency shutdown Settings configuration page.
The initial starter duty specifies what the starter motor is run at, % duty, at the initiation of the startup sequence. It will hold this speed until ignition is detected, which is by way of monitoring the turbine outlet temperature. I have hard coded the program to watch for a 100C increase in TOT - once achieved, the system turns off the glow plug and begins the starter ramp sequence.
The Starter Ramp Speed variable controls how quickly the starter motor ramps up in speed from the initial start duty speed once ignition is detected. The value is a percentage of duty increase to the starter speed at 1 second intervals. i.e. If the start duty is 20% and the ramp speed is 10, it will ramp up to full speed in 8 seconds. During the ramp subroutine, the system is monitoring the TOT to make sure there isn't a flameout. Once the starter has achieved full speed the system takes a TOT reading and monitors it for 5 seconds - if the TOT falls more than 100C in this time, the startup sequence halts and reports a flameout condition. If the temp doesn't fall below that value, it proceeds to the fuel ramp subroutine and leaves the starter motor running at full speed.
Once in the fuel ramp routine, the system will increase the fuel pump duty from zero percent up to the idle fuel duty parameter at a rate specified by the fuel ramp parameter. The fuel ramp parameter is percentage of duty increase per second - just like the starter ramp value. Once the idle fuel duty has been achieved it turns off the starter motor and will hold this fuel rate until the throttle potentiometer has been increased up to the idle fuel rate to match. At this point the startup routine exits and hands fuel control over to the throttle potentiometer.
The startup sequence works rather well - it goes by really quick and since I've got one hand on the starter push-pin to engage it and currently having to put a torch at the jetpipe for ignition AND adjusting the propane feed, it is quite a "hands on" procedure. Not to mention, once the sequence reaches the idle fuel flow, I have to reach over and turn up the throttle potentiometer, then shut down the propane feed.
I also discovered a problem with the starter itself. Something I overlooked in the design of it was the fact that the electric motor has to support a rather big axial load when pushing it in to engage the starter shaft to the compressor nut. I found a small ball bearing and machined a seat for it into the starter pin. This bearing just happened to have an inner diameter just slightly smaller than the electric motor's shaft. So when you press the starter pin inwards, it engages over the armature shaft - thereby eliminating the axial load placed on the electric motor's brass bushings, which were never designed to handle any significant axial load to begin with.
The other issue I have come across with the setup is the sheer amount of current required to run the starter motor with this kind of load on it. This load is causing the MOSFETs driving it to get pretty darn hot. I am using two NTE2985 MOSFETs which are rated up to 30A continuous but it looks like I need to bump up to around 6 MOSFETs and a larger heat sink to spread the load across.
During the startup routine a few times I experienced a flameout due to too little propane feed but I did discover that if I turned the propane feed up high enough to start with I didn't have a flameout nor did I have to change the propane feed rate during the rest of the startup routine. This is nice to know - this means I could get a high pressure solenoid valve for the propane and hook it in after the propane flow valve. The startup sequence could trigger the propane feed automatically and turn off the propane at the end of the sequence without me having to touch any of it.
Since I have spent the time back with all of the programming I think I will go further to make some additions to the code that I wanted to do previously. One of the things I dont have is a means to control how rapidly the fuel pump speed can be changed. As it stands, I have to be rather careful when adjusting the throttle - I can't make any quick drastic changes to it else it will either shoot a huge flame out the tailpipe and kick into the emergency shutdown or it will throttle back too quickly and flameout. I want to integrate a routine to handle the rate of change allowed for the fuel pump speed. Should be an interesting section of code given the nature of the throttle input (it can change at any time) and the goal-seeking nature of the routine. I've pretty much got it nailed down in my head but as I came to find when programming the startup sequencing, it will probably be a challenge in implementation. :)
-
- Posts: 176
- Joined: Sat Sep 04, 2004 7:17 am
- Antipspambot question: 0
- Location: Edgewater, FL
- Contact:
I've setup a throttle control routine in the code that allows me to configure the maximum throttle change rate per second. It is working pretty nicely and prevents me from kicking in too much fuel too rapidly as well as prevents flameout from reducing fuel flow too rapidly.
I took the scooter out for some test runs today - she is working very well and I feel/see a bit of an improvement in performance. I am able to run a few more percent fuel pump duty with higher P2 and the TOTs are staying the same as previous.
I ran the scooter for about 15 minutes around the neighborhood - stopped by a buddy's shop where I tune my customer's vehicles just to let him see it in action - kept it running the whole time and made a fly-by for him on the road in front of his shop and proceeded home.
She ran really well and I filled the tank all the way up to go out for another run. I took her down one of the main roads near our neighborhood with my buddy following in my pickup truck - I was running in the bike lane and cruising a solid 40MPH (the speed limit on that street). About 4 miles down the road from out of nowhere the engine siezed up - flames bellowing out of the jetpipe, LOL. I pulled off into the grass and let the flames go out - fortunately there was no damage due to the fire. We loaded the engine into the back of the truck and proceeded back home.
Once she had cooled down I removed the engine from teh scooer and dismantled it. The first thing I noticed before I took it apart was that the compressor diffuser cover plate was sticking out a little further than where it should be and I suspected that the welded NGV/rear NGV plate had broken free or something stretched out because of heat. Once I got the guts out of the housing the first thing I noticed was that the whole diffuser section was loose! I removed the diffuser cover and removed the compressor wheel to find that the 6 bolts that hold the diffuser to the bearing tube had come loose and jammed up into the back of the compressor wheel. :-( After further disassembly I also found that the shaft extension was also bent pretty good as a result of what happened.
Fortunately that was the only damage to the engine, which isn't too bad. I'll get a new compressor wheel for ~$70 and machine a new shaft extension. The compressor wheel is the easy part - making the shaft extension usually requires several attempts to finally get one with acceptable tolerances, so I'll be looking at several hours of machining to get her back up and going.
As for the NGV section - everything looks perfect. :) Even with those tight tolerances between the turbine wheel and exducer there are no signs of interference and all of the welds holding the NGVs to the outer NGV plate are intact and literally look no different than they did before I ran the engine after the modification. The 316SS parts are holding up incredibly well - the metal looks really really good - no scaling, no blistering - just a nice brown coloration to it.
I'm really glad I have put a little more time into the T04 project - the results of using the 316SS on it have pretty much convinced me to use this alloy in the TV94 engine's NGV section. The TV94 probably wont run as hot as this engine does either so in that case this alloy should do even better.
I'll post later once I get the new wheel and machine up a new shaft extension. :)
I took the scooter out for some test runs today - she is working very well and I feel/see a bit of an improvement in performance. I am able to run a few more percent fuel pump duty with higher P2 and the TOTs are staying the same as previous.
I ran the scooter for about 15 minutes around the neighborhood - stopped by a buddy's shop where I tune my customer's vehicles just to let him see it in action - kept it running the whole time and made a fly-by for him on the road in front of his shop and proceeded home.
She ran really well and I filled the tank all the way up to go out for another run. I took her down one of the main roads near our neighborhood with my buddy following in my pickup truck - I was running in the bike lane and cruising a solid 40MPH (the speed limit on that street). About 4 miles down the road from out of nowhere the engine siezed up - flames bellowing out of the jetpipe, LOL. I pulled off into the grass and let the flames go out - fortunately there was no damage due to the fire. We loaded the engine into the back of the truck and proceeded back home.
Once she had cooled down I removed the engine from teh scooer and dismantled it. The first thing I noticed before I took it apart was that the compressor diffuser cover plate was sticking out a little further than where it should be and I suspected that the welded NGV/rear NGV plate had broken free or something stretched out because of heat. Once I got the guts out of the housing the first thing I noticed was that the whole diffuser section was loose! I removed the diffuser cover and removed the compressor wheel to find that the 6 bolts that hold the diffuser to the bearing tube had come loose and jammed up into the back of the compressor wheel. :-( After further disassembly I also found that the shaft extension was also bent pretty good as a result of what happened.
Fortunately that was the only damage to the engine, which isn't too bad. I'll get a new compressor wheel for ~$70 and machine a new shaft extension. The compressor wheel is the easy part - making the shaft extension usually requires several attempts to finally get one with acceptable tolerances, so I'll be looking at several hours of machining to get her back up and going.
As for the NGV section - everything looks perfect. :) Even with those tight tolerances between the turbine wheel and exducer there are no signs of interference and all of the welds holding the NGVs to the outer NGV plate are intact and literally look no different than they did before I ran the engine after the modification. The 316SS parts are holding up incredibly well - the metal looks really really good - no scaling, no blistering - just a nice brown coloration to it.
I'm really glad I have put a little more time into the T04 project - the results of using the 316SS on it have pretty much convinced me to use this alloy in the TV94 engine's NGV section. The TV94 probably wont run as hot as this engine does either so in that case this alloy should do even better.
I'll post later once I get the new wheel and machine up a new shaft extension. :)
-
- Posts: 503
- Joined: Sat Jul 23, 2005 11:11 pm
- Antipspambot question: 0
- Location: Australia
Hi Ash
Looks like you might need to use lockwire on those bolts , she's certainly going well though , really clocking up the miles :-))
Pulled FM-1 apart after its second spoolup yesterday , took her to ~35,000 rpm , and found the inner flametube wall had been running considerably hotter than either the outer or front/back walls , so came in to have a look at the remedy you applied to fix the same problem when you had it
Congratulations on the development work you're doing , nice :-))
Cheers
John
Looks like you might need to use lockwire on those bolts , she's certainly going well though , really clocking up the miles :-))
Pulled FM-1 apart after its second spoolup yesterday , took her to ~35,000 rpm , and found the inner flametube wall had been running considerably hotter than either the outer or front/back walls , so came in to have a look at the remedy you applied to fix the same problem when you had it
Congratulations on the development work you're doing , nice :-))
Cheers
John
-
- Posts: 176
- Joined: Sat Sep 04, 2004 7:17 am
- Antipspambot question: 0
- Location: Edgewater, FL
- Contact:
Hi Jon!
Unfortunately those screws have countersunk heads so there isn't really any way to retain them with wire. :-(
This is the first and only time the engine has suffered this problem - I knew it was always a possibility that this could occur but the main reason for using the countersunk screws was because the shape of the head tends to help lock the screws in place to keep them from backing out.
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/J-DIFFUSER5.jpg)
What I will be doing from here on out is using blue loc-tite on these screws to prevent them from backing out. That will keep them held down really well but not so tight that I can't get them back out (as would be the case with red loc-tite). It is a shame that the engine was damaged so much just by simple bolts backing out - I'll see to it that this doesn't occur again though. :)
Unfortunately those screws have countersunk heads so there isn't really any way to retain them with wire. :-(
This is the first and only time the engine has suffered this problem - I knew it was always a possibility that this could occur but the main reason for using the countersunk screws was because the shape of the head tends to help lock the screws in place to keep them from backing out.
![Image](http://ashspecz.com/Image%20Gallery/GasTurbineT4/images/J-DIFFUSER5.jpg)
What I will be doing from here on out is using blue loc-tite on these screws to prevent them from backing out. That will keep them held down really well but not so tight that I can't get them back out (as would be the case with red loc-tite). It is a shame that the engine was damaged so much just by simple bolts backing out - I'll see to it that this doesn't occur again though. :)