Been a while - new project, T04 Gas Turbine.

Moderator: Mike Everman

Post Reply
Rossco
Posts: 589
Joined: Fri Nov 28, 2003 12:16 pm
Antipspambot question: 0
Location: Australia, Brisbane
Contact:

Post by Rossco » Thu May 24, 2007 3:06 am

Righto Ash, got it... still some issues tho.
Waiting on your word.

Rossco
Big, fast, broke, fix it, bigger, better, faster...
[url=callto://aussierossco]Image[/url]

larry cottrill
Posts: 4140
Joined: Sun Oct 05, 2003 1:17 am
Antipspambot question: 0
Location: Mingo, Iowa USA
Contact:

Visual Models

Post by larry cottrill » Thu May 24, 2007 12:01 pm

Johansson wrote:Those models that Rossco made are really nice, can´t wait to see one of the entire engine.
Rossco -

Yes, I intended to comment on these way before now. Man, that is some beautiful work!

What would be way cool is if you could take these "sectioned" drawings and break out all the individual parts into an "exploded view" - let me tell you, on complex mechanical systems, that is an art in itself.

Nice job, mate!

L Cottrill

Ash Powers
Posts: 176
Joined: Sat Sep 04, 2004 7:17 am
Antipspambot question: 0
Location: Edgewater, FL
Contact:

Post by Ash Powers » Tue May 29, 2007 7:21 am

For dessert: :)

Image

Image

Image

Rossco, I've gotta hand it to you my friend, when you sent these images to me and I saw them for the first time, there was almost an erotic response on my end. ;-) I'm really looking forward to doing more modeling and CFD in this project.

At the moment, Rossco and I are just *playing* around for the most part -- the model above is of the current combustor, but it lacks the evaporator tubes which obviously will have a significant impact on how all those gases flow. But if a CC were built identical to that with evaps running axially along either the inner or outer liner of the CC, it probably wouldn't look much different - at least, I'd guess. :) The models above were constructed based on some guesswork for certain variables - those were literally the "first shot" outputs of the CFD and although I'm no gas turbine expert, seeing those low velocity swirls in the primary zone are quite exciting. If that model is even somewhat accurate, I feel more assured of my understanding of how hole orientation between the outer and inner liners creates these vortex patterns. :)

There are still other aspects of the combustion chamber we are working towards modeling - static pressures around the chamber, massflow into the different zones, namely. The idea is to design a chamber in accord with the info we have available in the DIY community - namely the 30/20/50% flow rule.

I've done a lot of google searching on combustion modeling and found a lot of very interesting info....

http://www.fluent.com/solutions/examples/x211.htm
http://www.fluent.com/about/news/newsle ... i2/a21.htm

Seems it can be done - mind boggling stuff though. :)

I've dissected the current combustion chamber so I can get a better look at things and show the group.

Image

Image

Image

Image

These images aren't of the best - the colors are too saturated - not enough light.... But, I can clearly see with them in hand that the front plate hasn't seen much heat at all, the evaporators I would say a "2" (out of a scale from 1 to 10) and the outer liner an 8 through its mid-section. The inner liner and rear "cone" dilution zone haven't seen much heat either - around a 3-4 - the rear cone is barely even discolored on the inside.

It appears to me that the rear dilution ports are too large - preventing enough airflow into the primary and secondary zone to burn all the fuel at higher fuel flow rates, and causing the remainder of fuel to burn back near the entrance of the NGV. Additionally, all the wet sooty deposits on the front plate and about 3/4" back from the front section of the inner and outer liner tell me I'm not getting good vaporization either. The poor vaporization is probably not a factor of the vaporizers themselves, but with the distribution of air in the chamber not allowing the combustion to reside at the front of the CC - seeing how the midsection of the outer liner shows a lot of heat exposure tells me the fire is mostly burning in the middle/rear of the CC. Once we have the CFD models refined I believe we will see an overabundance of airflow entering the CC through the rear cone dilution zone - probably like 70% rather than the 50% we DIYers go by. Interestingly enough, the area of the holes in each zone are within 1% of the 30/20/50% rule, but just by observation it appears the geometries of the engine's internals as a whole are affecting how much air is going through those holes regardless of their size - the static pressures around the chamber at the different zones is not the same and although the hole sizes are correct, we dont have the same pressure conditions at each zone, so the air flowing through those holes isn't proportional to their size.. ....... enter CFD ....... :)

Although the CFD gives a representation of how the gases behave, that model begins to fall apart once the gases have entered the combustion chamber - the effects of combustion cause the gases to behave differently than they would if they are just passing through the chamber. Even though, my thoughts on this are to use CFD to design a CC that has even static pressures around the chamber, proper mass distribution (the 30/20/50% rule) for the different zones, and hole orientations that create good circulation zones - basically all the good stuff we already know we want. If we can achieve that, we should have a burner can that works really well without enduring the grind of "splitting hairs" trying to perfect every last variable to the n'th decimal.

... and on a similar note - I've finally figured out what my problem was with spinning stainless sheet. When I spun the aluminum cowling for the engine, I used the butt end of a hammer handle (wood) dipped in oil and used that to form the metal. It worked pretty well but my attempts to spin stainless was not so fruitful. Every try would cause the sheet to buckle and if you know anything about spinning metals, as soon as that happens, the part is junk and you start over.

Well, stubborn as I am and knowing it can be done, I did some more searching on the web about spinning metals. I found a few videos of the process and while amazed at seeing it done, was also somewhat disgusted at how easy *they* make it look, LOL. But the one thing I noted in all of these diagrams and videos was the actual tool that worked the part; they were all shaped like an inline roller skate wheel, but made of metal, supported by a ball bearing to allow them to spin with the part while forming. So, for kicks, I took an old live center and fixed it to my toolpost - put a new sheet of 316SS on my mandrel and started working it. I had the toolpost lock loose so I could rotate it and used the base of the center's taper (at its broadest point) to press against the sheet. In about 30 seconds I had formed the sheet around the mandrel without any problem.

Apparently trying to form stainless using the wood/oil approach creates too much friction and the metal folds up - using a roller is the "trick".

So I dug up an old ball bearing in good shape, ~1.25" o.d. and machined a roller to press it into - need to make up the handle and pin/grip to hold the bearing/roller, and I'll make up a different lever-action toolpost like what is used on spinning lathes, but it looks like I'm in at this point for making some nice pieces. :)

This just makes it look *all* too easy..... (CNC spinning)
http://www.mfgquote.com/process/media/M ... inning.wmv
http://staff.aist.go.jp/h.arai/video/spinning.avi

....... but then again, it is always a matter of having the right tools. 8)

racketmotorman
Posts: 503
Joined: Sat Jul 23, 2005 11:11 pm
Antipspambot question: 0
Location: Australia

Post by racketmotorman » Wed May 30, 2007 7:15 am

Hi Ash
That primary end of your flametube is still very "shiny" , not much colour at all .
Heres a pic of the karts F/T after only a tank full of fuel had been burnt in her
http://www.nickhaddock.co.uk/turbine_st ... 030502.JPG
Colour right up to the very top of the wall

I'd be pretty certain theres nothing much wrong with your evaporators , just no flames to heat them :cry:

Nice graphics Rossco :D

Cheers
John

Ash Powers
Posts: 176
Joined: Sat Sep 04, 2004 7:17 am
Antipspambot question: 0
Location: Edgewater, FL
Contact:

Post by Ash Powers » Thu May 31, 2007 2:21 pm

Yeah - static pressures around the CC are apparently "all jacked up" and messing with the flow.

My T25 turbojet engine's CC looks just like yours, although notably smaller:

Image

Image

Different animal with these axial configured annular chambers tho - lots more room for static pressure variations.

From what you've seen here, do you have the same thoughts/feeling I do about how the engine's performance can be improved with a new CC?

I've seen the P2 at 20psi for short bursts so I know the turbine end is capable of producing the power for the compressor - I would be inclined to say that if we can get the combustion in the right place the engine will be able to run at that P2/massflow within acceptable TOTs.

I've got all the materials here now to start on the new chamber as well as machine the rear NGV plate from 316SS - Rossco and I will spend some time modeling the new CC layout and doing what we can to get the static pressures around the chamber as equal as possible as well as get proper airmass distribution and good recirculation in the primary. I feel pretty confident that this next chamber design will make a world of difference in performance.

racketmotorman
Posts: 503
Joined: Sat Jul 23, 2005 11:11 pm
Antipspambot question: 0
Location: Australia

Post by racketmotorman » Thu May 31, 2007 5:54 pm

Hi Ash
Theres still that basic problem with your comp/turb flow matching that hasn't been adequately addressed yet :-((
You might be chasing an imposibility .
I spent two years trying to get more thrust from the TV84 without success , simply because I neglected to follow the maths , well didn't actually not follow it , just didn't know it back then :-))

Cheers
John

Ash Powers
Posts: 176
Joined: Sat Sep 04, 2004 7:17 am
Antipspambot question: 0
Location: Edgewater, FL
Contact:

Post by Ash Powers » Sat Jun 09, 2007 10:38 am

John,

The calcs you and I went through not long ago were based on the unclipped turbine wheel. I've shaved the blade's exducer back by a good amount as well as increased the throat area in the current NGV by way of making the vanes a little taller and spacing the turbine a little further away from the NGV plate.

Here's how the wheel originally looked:
Image

and here's where it is at now:
Image

I dont know what the flow capacity of the turbine section is at this point. We calculated some 35lbs/min with the previous NGV area and unclipped turbine so I would expect that value to have increased with the latest mods...

racketmotorman
Posts: 503
Joined: Sat Jul 23, 2005 11:11 pm
Antipspambot question: 0
Location: Australia

Post by racketmotorman » Sun Jun 10, 2007 12:02 am

Hi Ash
The problem we're encountering even at 35 lbs/m is a change in comp efficiency to a lower value as the PR increases , its flowing well outside the higher efficiency regions on the map
This puts added load on the turbine to produce power , power is produced at the turbine by gas deflection , clipping the turbine reduces deflection and power output , this is "normally" restored by using higher temperatures , but higher temps also reduce density and this makes it harder to get the gases into the turbine without the use of "steeper" NGV angles , but these steeper angles reduce gas deflection and turbine power .
Without turbine power the comp can't get to the higher rpm and P2s without excessive temps and those temps reduce flow etc etc .....

Its imperative to have a well matched comp/turb , and your turbine wheel is a little small in the inlet area department for the comp its trying to power .
In an IC engine situation it doesn't matter too much because the pistons will force the gases thru an "undersized" turbine . There are very few "smaller" sized turbochargers with correctly sized turbine wheels that can be made directly into a gas turbine, because the designers want fast spoolup to boost conditions regardless of the "backpressure" put on the engine .
Large diesel truck turbos will generally have larger turbines more "correctly??" proportioned for gas turbine use , because they are trying to get the absolutely best fuel efficiency from the truck engine , and any unnecessary backpressure , other than that necessary to produce the desired maximum boost level , wastes fuel .

You really need to do extensive data collection of engine, to determine all its running parameters before continuing too far with your development , without data like thrust level ,T2, P2 P4t , TOT, jetnozzle sizing , mass flow , etc etc . its going to be impossible to know what changes actually work and what don't , you're flying blind at the present time :(

Once you have hard data , you can start doing the maths .
Our engines are all maths and physics , so fairly simple to work things out with a pencil and paper :D

Cheers
John

Ash Powers
Posts: 176
Joined: Sat Sep 04, 2004 7:17 am
Antipspambot question: 0
Location: Edgewater, FL
Contact:

Post by Ash Powers » Thu Jul 05, 2007 9:52 am

Well guys, it has been real fun building this engine and sharing the project here with everyone. This T04 project has pretty much served its purpose at this point although I know there is still much to learn about gas turbine technology. When I set out on the build of this turbine engine I originally was drawing plans up for a larger unit using the Cummins ST-50 rotating group. I had both the ST50 and T04 bits here at that time and for sake of materials and machining/fabrication times, I opted to drop that original engine down in size and use the T04 bits as a "pre-build" to get my head around the design and operation before embarking on the larger version. Looking back, I'm really glad I did that for so many different reasons. I've come to realize that there are much better turbocharger components out there than the ST50 bits. I've also learned a wealth of knowledge about the thermodynamics involved in these devices but also realize I've only learned a drop's worth of an ocean of knowledge.

I really must give credit where it certainly is due as well: John Wallis (racketmotorman) has been extremely helpful throughout this project - even from just listening to my rambling hypotheses all the way to him taking time out of his day to mail gas turbine research documents half-way around the world to me. :) John, you have definately earned the title of "The MAN" in my book - thanks again for being a part of this project. :)

At this point, The T04 engine is back together and running well on the scooter. I remachined the outer NGV plate using 316 stainless as the original mild steel plate suffered from a lot of blistering, scaling, and warpage that rendered the component useless.

I've also modified the fuel ring and no longer using hypo tubes; I made a new ring and drilled precision 0.025" holes in the ring. 1/8" diameter SS tubing pieces were cut 1" long and brazed to the fuel ring over the holes; these tubes drop down into the vaporizer inlets to deliver fuel - no more worries about the hypo needles breaking off and dumping fuel all around the chamber. I also blocked off three of the dilution ports at the back of the chamber to help promote a bit more airflow into the primary zone to aid in more complete combustion in that area. I ran her again yesterday coming up to 15psi P2 at 680C TOT which is a small improvement over previous runs. In about 600 feet of distance, I can get her up to around 40MPH or so at this power level. Plenty fast enough to make it a fun ride :)

I could easily start splitting hairs on this engine and spend countless more hours trying to fully refine the design but I'm going to leave her as is and start working towards getting everything together for the final part of this project: A TV94 rotating group in the gas producer with an axial free-power turbine and gearbox to drive a ducted fan unit: a "TurboFan". :)

My goals are to produce a GT capable of generating around 350 lbs of thrust for use in a VTOL airframe capable of lifting itself and me. :) I've been kicking a few ideas around for the airframe design and haven't committed to any of them at this point. I would love to build a "jet pack" type of device - something to strap to my back and joy-fly, but there are some inherent risks in that design of which could easily be fatal. I'm considering something more along the lines of a flying platform where there is a more comfortable distance between the ballistic bits and my body - envisioning something like a slightly oversized go-cart frame; engine behind the seat with the gas producer's rotating group axis parallel to the front-rear axis of the frame. The ducted fan assembly will sit horizontal with plenty of scatter shielding and feeding a plenum that distributes airflow to the four corners of the airframe. Gimballed nozzles at each of the four corners will be used to vector the thrust for yaw control and butterfly valves in each duct will allow control of pitch and roll. This layout will likely be too complex to employ a mechanical control system; not for sake of the mechanical bits/control mixing,etc but more for stability. A fly-by-wire system, gyroscopically stabilized. Should make for a challenging (read: FUN) electronics project in and of itself.

All of the electronics aspect isn't really the "IF" part of the project - I know a system like that is well within my grasp but the "IF" lies completely in the turbofan engine's performance. I'm not sure if it is possible to construct a gas producer from these turbocharger components that will be capable of driving a ducted fan to produce this kind of lift. The ducted fan unit will likely be quite a challenge as well - a ~22" ducted fan will be under significant stress and I dont want to be around it when something lets loose so it will have to be overbuilt with safety in mind.

So there it is, for what it's worth. :) I know something like this is an enormous undertaking; I wont discount that fact by any measure. But it will be a challenge that undeniably will keep me entertained for quite some time and offer plenty of opportinuties to learn new things. Whether it actually comes to fruition or not I cannot say but that idea never stopped me before. :)

Thanks to everyone who took part in this thread - I really appreciate all of the positive feedback you had to offer during this project and I am looking forward to bringing this next project to the forum later this year when I start building. :)

Johansson
Posts: 1161
Joined: Sun Nov 21, 2004 9:42 pm
Antipspambot question: 0
Location: Northern Sweden

Post by Johansson » Thu Jul 05, 2007 7:45 pm

Thanks for sharing your project, it has been both interesting and entertaining to read. How many hours have you run the engine up to now?
My goals are to produce a GT capable of generating around 350 lbs of thrust for use in a VTOL airframe (...)
Holy sh*t! Now THAT is a project, I have always wondered if it is easier to build a microturbine-style engine with a large core and soon I will get the answer. :-)

I assume that you have found this page: http://www.technologie-entwicklung.de/G ... opter.html

//Anders

Ash Powers
Posts: 176
Joined: Sat Sep 04, 2004 7:17 am
Antipspambot question: 0
Location: Edgewater, FL
Contact:

Post by Ash Powers » Thu Jul 05, 2007 10:55 pm

Anders,

Thanks! Its been a lot of fun for sure. The engine probably has about 4 hours runtime total since it was first run in mid February but I have rebuilt many of the parts since then to get her running well.

To produce some 350lbs thrust from a turbojet configuration will require quite an engine and will guzzle a lot of fuel whereas a smaller gas producer coupled to a FPT/ducted fan can produce the same static thrust and likely use less fuel. I'm really interested in the turbofan configuration - I know it will be a challenge to build. :)

I've seen Thomas Baumgart's website before - spent a good bit of time going through everything on that site - quite an extensive lineup of gas turbine technology and applications he has there. :) That monocopter is an inspiration to say the least! I just dont have a lot of faith in homebuilt bits spinning at ballistic velocities in the same plane as my skull.

It is going to be an awesome project though - I'm really excited about getting to work on it later this year.

Thanks!

racketmotorman
Posts: 503
Joined: Sat Jul 23, 2005 11:11 pm
Antipspambot question: 0
Location: Australia

Post by racketmotorman » Fri Jul 06, 2007 6:08 am

Hi Ash

I can't wait to see the new thread start , along with a lot of "lurkers" out there who read , but don't say too much because of "second language" difficulties .

They appreciate your pics, developmental "difficulties??" and your solutions , coupled with your good machining and standard of finish , its a valuable educational experience to a number of people who'll now feel more confident at starting a similar build .

Roll on the TV94 engine to bigger and better results :D

Cheers
John

Ash Powers
Posts: 176
Joined: Sat Sep 04, 2004 7:17 am
Antipspambot question: 0
Location: Edgewater, FL
Contact:

Post by Ash Powers » Sun Oct 14, 2007 10:15 pm

Been a while since I posted regarding this build but I have made some recent changes to the setup and have seen good results. :)

About two months ago I was running the scooter to get some new videos and had a somewhat catastrophic failure. The three bolts that hold the rear NGV plate to the NGV assembly overheated and stretched. This allowed the entire internal assembly to be pushed out of the front of the main casing due to the pressure within the engine. This allowed the clearances between the turbine and the exducer flange to grow and it burned up one of the vanes - this spit little bits of stainless steel into the turbine wheel, chipping the inducer blade tips pretty badly. Upon disassembly, I found that a few of the injector needles had also broken off, which is undoubtedly what led to this overheating as the fuel was basically spitting out all around the combustion chamber rather than being guided into the vaporizer tubes.

I ordered a new turbine from my turbo builder, ground for the bearings, clipped back by 10 degrees and final balanced. I machined the threads into the shaft quill and cut it to length as the previous turbine was.

I ordered new stainless tubing to make the injector needles from - this time I went with 0.050" o.d. with a 0.033" i.d. which has a wall thickness of 0.0085" - considerably thicker than the previous needles I was using. I silver brazed these into the fuel rail and to add additional support for them, I slipped 1/8" o.d. stainless tube over each of them and brazed them into place. I crimped down the end of the 1/8" tubes so that they are "holding" the injector needles firmly.

I also cut the cooling air slots for the bearing tube in the diffuser plate a little deeper to get more cool air into them as it appeared that the rear bearing and the back end of the shaft extension could use a little more cooling.

This afternoon I assembled the engine and mounted her on the scooter. I filled the 2.75 gallon tank up with Jet-A with ~20oz of 2-stroke lube and fired her up for the test run. I had a friend of mine follow me in my pickup truck and I ran her until all the fuel was gone. We covered ~9 miles at a cruising speed of 30MPH (speed limit in my neighborhood) with a TOT of 735C. Total runtime was around 22 minutes. We returned back to my place to inspect things and she couldn't have run any better. The new turbine wheel doesn't even show signs of discoloration - still has that nice inconel shine to her. :)

At this point I am pretty satisfied with her performance and expect that I should be able to run her like this for quite some time before any servicing is needed. The only thing I can think of doing is to replace the three bolts holding the rear NGV plate with an inconel counterpart - something that can take the heat without stretching out, although the current bolts seem to be just fine.. The overheating from the injector needles breaking off was the cause of that problem but even though that problem is fixed, it would probably be a good idea to follow up on it. I'm going to leave the engine together and see what kind of longevity I can get out of her before something needs to be replaced. This week is "Bike Week" in Daytona and the surrounding are so there will be lots of harley davidson events around here. I'm strongly considering to drop in with the jet scooter - for entertainment's sake at least.. If I do, I'll take pics and share them here - I know there will be a lot of priceless faces. :)

Johansson
Posts: 1161
Joined: Sun Nov 21, 2004 9:42 pm
Antipspambot question: 0
Location: Northern Sweden

Post by Johansson » Mon Oct 15, 2007 7:10 pm

He he, I think that your turbine bike would steal the show at the bike meet... 8)

Do you have professional balancing equipment at home or do you balance it the old fashioned way by rolling it over a flat surface? Since you are into car tuning I guess that you have access to a balancing rig.

Mike Everman
Posts: 4906
Joined: Fri Oct 31, 2003 7:25 am
Antipspambot question: 0
Location: santa barbara, CA
Contact:

Post by Mike Everman » Tue Oct 16, 2007 3:37 am

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
__________________________
Follow my technical science blog at: http://mikeeverman.com/
Get alerts for the above on twitter at: http://twitter.com/mikeeverman

Post Reply