T04 Gas Turbine - RESURRECTED!

[ashpowers]

Patty,

I am just using compressed air. I have a 50-gallon, 5HP air compressor running at 135psi and she spools up without any problems with that. The ball bearings have a lot lower resistance so startups go easy.

[ashpowers]

OK guys, having some pretty serious issues with the engine. Hoping I can get some ideas on which direction to go.

I can get the engine fired with propane and she spools up nice to around 12KRPM but as soon as I start feeding the JetA I get a burst of flames out of the jetpipe and she spools up to higher speed but she wont register any more than about 1-2psi of pressure at around 25KRPM and still with pretty decent size flames coming out of the rear. No engine is going to survive in this condition so I've aborted the startup.

I'm not sure if I need to just hang in there and add more fuel to get the revs up or if this condition is just going to get worse. This is nothing like how she ran before - startup was always pretty straightforward and easy without all these huge flames coming out of the pipe.

So A couple of things I am thinking may be playing into this.

1) The opening for the cooling air to the bearing tube is just too big and resulting in a lot of air bleed out behind the turbine wheel and preventing the engine from producing proper pressure. That opening is about 1/4" in diameter.

AND/OR

2) The new combustion chamber almost seems like its getting too much air into the primary zone and not allowing slow moving air to handle the burn.

I went by the numbers on the combustion chamber liner holes, cross sectional area distribution, etc - maybe a litte bit large on the holes in the top "Scoop" plate out around the periphery as when I indented them to form injection nozzles, the holes opened up a little more than I anticipated - perhaps by an additional 0.030" per hole in a total of 24 holes.

Another possibility is the fuel/oil mixture I am using. I went pretty high with the oil content - 16:1 ratio as per the mixing chart on the back of the 2-stroke oil (2.1oz per 0.333 gallon of JetA). Not sure if this is affecting the burn or not.

My injector needles on the fuel rail are 6 X 0.027" i.d. which is what I used before.

I'll be opening up the engine later this evening to take a look at the bits to see how the fuel is moving through the system. Last time I took her apart the front wall of the combustion chamber was ripe with 2-stoke oil and I'm not sure why that was the case. The fuel rail has no leaks but it possible that there may be reversion in the vaporizers OR some fuel/oil may have run back out of the vaporizers after shutdown.

Quite disheartening after all of this effort to get her back up and running to have an issue like this. When I first built this engine I didn't go by any numbers at all - literally built it entirely off the cuff and although she wasn't able to produce any notable power in her original form, she did at least run on her own, LOL.

So, here's the possible candidates:

1) Too much bleed air to the bearings?
2) Too much air into the primary zone?
3) Too much oil quantity in the fuel?
4) Possible air reversion in the vaporizers?

I may just shotgun an approach, reduce the bleed air port by half, close up the holes in the top plate of the combustor, get rid of the oil in the fuel and inject straight 2-stoke oil into the bearing tunnel using P2 pressure into an oil vessel and a needle valve to adjust the oil flow to the bearing tunnel.

[racket]

Hi Ash

Sounds like a hung start situation , more starter power required to get her "over the hump" .

Looking at my notes from when we were sorting the engine all those years ago , you're running a TO4 B 60-1 comp with a 2.324" ind and a 3" exducer , turb was a TO4 P trim with a 2.92" inducer and 2.54" exducer , turb exducer is only ~20% bigger than the comp inducer, a little marginal in the flow area so it could be easy to thermally choke the turbine stage during a spoolup , you might need to be careful about how quickly you add kero.

Idle rpm will probably be in the 40,000 - 50,000 rpm range, so some ways to go from the current 25K .

I wouldn't be looking at changing any of the flametube at this point in time, just get those rpm up with a little less fuel to keep the flames inside , adding more fuel will only exacerbate the thermal choking to the point where the comp could be surging .

The 1/4" bleed hole shouldn't be a problem at such low air pressures.

Yep , hang in there :-)

Cheers
John

[racket]

Hi Ash

Just been checking out the pics of your evaporators , they appear to be rather close to the end cap and not projecting down into the "flame area" .

Also , the opening of the evaporators appear to be very close to the endwall ( might just be the angle that the pic was taken from ) its generally advisable to have ~1.5 times the ID of the evaporator between end of evap and wall , there could be some evaporation problems that are causing the "oiling" of the front wall as well as flames thru the turb wheel .

Cheers
John

[ashpowers]

Hi John,

Well, I closed the cooling air port down to about 1/8" diameter and put straight JetA in the tank. I hooked up a syringe to the oil feed port and attempted to start her up. I was able to get her up to higher spool and making around 6-7psi of case pressure but the throttle was super finicky. I seem to think I had put a small jet in the fuel line previously to help smooth out the delivery from the pump. I did get her up to about 70KRPM but I was still seeing small flames coming out of the jetpipe and the TOT was in the 850C range, spiked to around 900 a few times but I couldn't get her to stabilize. I Dont think closing up the air port or running straight JetA had anything to do with it - I had read your responses before I tried starting her up this last time and tried easing into the fuel and did manage to get her up to some decent case pressure - she was running without needing any assistance from pressurized air but again, there was still visible flames coming out of the tailpipe. Combustion sounded smooth, no oscillations of fluttering but I dont think the vaporization system is working well enough. With the previous combustion chamber she would throttle quite well and no flames coming out of the tail so at this point I'm thinking the vaporizers just aren't working well enough. They are slightly shorter than they were in the previous combustor and yes, the outlets are pretty close to the front wall. I did cut the outlets off at a 45-degree angle with the more "open" side of that cut directing outward to the outer liner but I just dont think they are getting hot enough to vaporize all of the fuel.

I'm kindof leaning towards using a combustor design like Phil Heward's Phoenix 30.3 engine. Our engines are very similar and I still have a good bit of sheet left to make another chamber. His vaporizers run from rear to front along the inner liner of the combustor - they have a good bit of length to them and at least it will promote a bit more airflow down through the inner liner and along the bearing housing.

I connected my propane to the main fuel rail and started her up on that and she ran significantly better without any flames coming out of the rear so the only thing that makes any sense is that the vaporizers just aren't working well enough.

I just disassembled the engine and unfortunately the rear bearing cages are done. Fortunately I had two bearings in there so even without control of the balls, they kept each other in check since the only way it would come apart is if all of the balls were located exactly on the same side of the races at the same time. So I guess in this case having two bearings back there did save the rotating group. The front bearings look like they just came out of the packaging. There is a good bit of heat transfer through the shaft and the shaft looks like it has seen a fair amount of heat. The bearings looked to have gotten up in temp too although I only saw the temp in the bearing tunnel get up to about 375F. I did go through a long cool-down period but there was a lot of heat at the back end during this last run.

Looks like I'll have to get my hands on some better bearings for the rear and make a new combustor.

[finiteparts]

Ash,

The thermocouples that I have for measuring the rear bearing temps are 0.040 inconel sheathed type-K, capped and grounded that I picked up for a steal! The are small enough that they can be routed in small grooves to just about anywhere (I've seen some amazing instrumentation in my day and that is why I searched these ones out.) Omega sells even smaller diameter stuff if need be. But the J-44 has the thermocouple embedded into a brass (I think) sort of "head" that is spring loaded to stay in contact with the bearing outer race. I have a picture of it somewhere...I'll try to find it.

I think it might be tough to say exactly what the cause would be without testing, since there are so many factors at play there. I would tend to agree with you on the vaporizers not doing the trick since the gaseous fuel worked better, but that still leaves the question why the vaporizers aren't working well? Your comment "but as soon as I start feeding the JetA I get a burst of flames out of the jetpipe..." sure sounds like you are not getting enough heat in the vaporizer tubes to vaporize the fuel...ie. liquid slugs of fuel are passing through the vaporizers and burning "late". Also, since the flame in the combustion chamber heats the gas, which reduces the gas density and thus creates a back pressure on the system since it has to push a larger volume flow out of a fixed area, maybe the low pressure is an indication that your flame is not burning upstream of the throat section of the turbine and thus the "cooler" low density gas is more easily passing through...just a thought.

I think John is right, your vaporizer tubes look like they might be discharging too close to the head end...they weren't that close in your CFD were they? A tube discharging really close to a wall has a reduced flow as compared to one discharging into a plenum...If your previous engine had a better discharge coefficient on the vaporizers, you might have had a higher airflow in the vaporizer tubes...higher flow would help break up the liquid jet into smaller portions that would more easily vaporize...plus, a rich primary zones would better fit the idea of the flame burning "late", once it exits the primary zone and picks up sufficient air in the secondary, dilution zones or even outside the engine (assuming that there is a stable place for the flame to anchor in the primary zone, which I think the gaseous fuel test tells us that there is).

Combustion design is really tricky. When reacting flows are involved, they can change the flow field in the combustor from what the cold CFD predicts. When you took the engine apart, does it look like the flame is washing over the vaporizer tubes? (oxidation staining, etc) If not, then maybe you could play with the primary hole placement to get a better swirl in the primary zone.

Good luck! ~ Chris

[racket]

Hi Ash

Yep, sounding more like evaporator troubles

I normally run ~10% of the air thru the evaporators , so for you that'll be 0.42 sq inches of total vaporiser bore area , with 6 of them its down to 0.07sq ins/evap , or ~0.3" ID .

Evaporator surface area ............~6 times inducer area , so ~4.2 sq ins/evap for you , assuming 0.4" OD for each evap , that'll need a length of 3.38", depending on how/where the evaps are situated with respect to the hot combustion gases you might to increase the length of the main arm to get the two elbows into the flames .

Both Anders engine and my micro engines have straight evaporators only ~4 inches long , the TV94 has ~11.7sq ins of inducer, roughly 3 times the TO4 , and I use 18 evaporators with a 3/8" OD , the ID flow area is reduced a bit because of the "D'ing" of the tube and this should bring flow area back somewhere close to the 10% mark.

With your bearings , was there any difference in the cage damage between the "front" and the "rear" cage closest to the turbine wheel , a difference in damage possibly caused by hot gases contacting the one closest to the wheel ?

Cheers
John

[ashpowers]

Well, since the consensus seems to be vaporizers then I'll get to work on building one similar to the Phoenix 30.3 engine. I bought the plans for that engine years ago so I have detailed engineering drawings of it. I'll scale it out for my engine and give that one a shot - should probably work well given it is already a proven design.

The vaporizers look nice and silver - they do not have any of the coloration on them to indicate they were getting hot at all.

As for the bearings, what appeared to have happened is the retainer section that holds the balls simply failed - the little "fingers" that hang onto the balls, that is. That is the most delicate part of the retainer since there's not a lot of material there. The bearings were placed back to back so the rearmost bearing's retainer got blown out of the engine entirely. The front bearing's retainer was sitting in front of the bearing when I remove the turbine and it was missing all of the little ball-holding fingers. The retainer wasn't a solid ring with holes for the balls - these retainers had a solid ring on one side and small clips that held the balls in place. On the other side of the bearing there was no "ring" per se - you could see straight down to the balls themselves and only see the two little clips that held onto the side of the balls.

I'm having a really hard time locating bearings made of M50 steel races. I've put an inquiry in to a couple of places that make bearings to see if they have them or can make a set of them. Interestingly, production model gas turbine engines are using 440C SS bearing races with a full complement of SiN3 balls. I know the 440CSS material is good for higher temperatures so what I may do is get my hands on some 440C deep groove bearings, disassemble them, carve out the outer race so I can stuff a full complement of balls into those and give those a shot in the engine. I would also put them in a face to face arrangement with a preload spring washer between the outer races and appropriate shimming between the inner races so the pair would have some extra axial support and load to keep the balls from skidding, and then still use a preload spring like I have now at the front of the front bearing's outer race to stiffen the shafting and take a little bit of the axial load off the front bearings. That spring also helps to keep the outer races from free-wheeling in the bearing sleeve as well.

In the other parts of the engine, it is clear that the NGV has been exposed to the highest temperatures, not surprisingly given this thing always had flames coming out of the jetpipe. The inner and outer liners of the combustion chamber also look pretty good - only slightly brown in 90% of its surface and all of it still having that polished sheen to it. It just looks like all of the combustion is occurring right at the inlet throat of the NGV and through the turbine itself. This would also explain why trying to throttle her was so difficult - this engine was not a lag monster at all in her previous runs years back - she would happily throttle up and down, idling around 650C and even throttling her up quickly did not result in any visible flames out of the tailpipe. Kinda wish I didn't trash the old combustor, LOL - I just thought I could do it even better this time around.

[racket]

Hi Ash

Have you tried Barden Bearings www.bardenbearings.com/content.barden.us/us/index.jsp .

Yep, sounds like the combustion is seriously delayed , poor vapourisation , need more heat on the evaporators .

Cheers
John

[ashpowers]

Hi John,

I ordered more tubing (0.300" i.d.) and the plan is to open the combustor and make new evaporator tubes, longer, dropping further down into the chamber length and not coming up so close to the front face of the combustor. That process will go pretty easily and not have to waste so much work I've already done if the vaporizer lengths are the only problem.

I've got an email in to Barden asking about having them make up some M50 steel raceways and stuff them full of SiN3 balls, or make a recommendation on a bearing to use.

Funny thing, I've spoke with a few different bearing companies today and all of them are like, "its impossible to make a bearing run at those speeds". And then I fill them in on the fact that there is an entire industry out there for model gas turbines that are running at these speeds and greater. Then they all just seem to get stupid on me and say things like we can guarantee anything (like they would in any other case, LOL) or we just dont have a bearing for your application. You would think that talking to a representative of a bearing company, that they would know the breadth of the technology in which they are trying to sell. Frustrating always talking to a sales rep when you know more about their product than they do! Grrr.

I have found a company that sells 6901 full ceramic full complement deep groove bearings in SiN3 (not the zirconia stuff). I've never seen the balls in any of my rear bearings come apart - its always the raceways. So I'm thinking I might go with a set of those in the rear. Thoughts?

[racket]

Hi Ash

LOL.............yeh, bearing guys shudder when we talk about the Dn numbers we want to run .

A full SiN3 bearing would need very little lube , that stuff is super tough even at elevated temps .

Sounds like you only need some minor mods to the flametube and you'll be back in action , excellent news :-)

Cheers
John

[ashpowers]

OK, so I ordered a pair of 6901 deep groove bearings with 440C Stainless raceways and SiN3 balls. Those will be here tomorrow. I ordered additional balls so I can convert these to a full complement arrangement. I also ordered 12mm i.d. shims (0.1mm thick, 10 of them) and spring washers which I will use to shim between the bearings' inner races and put a spring washer between the outer races so I can preload the bearings against each other. I'll cut the outer races of these bearings to allow me to pack them full of balls and arrange them face to face on the shaft so they will not be able to fall apart.

Another thing I discovered is that the more forward bearing in the rear appears to have gone to more elevated temperatures than the rearmost bearing, which is not what I expected to see. At least, the inner race's color indicates this to be the case. I'm wondering if with all of the combustion occuring at the NGV and out of the tailpipe, that with lower case pressures some of that combustion was actually pushing into the bearing tunnel from the turbine end and possibly igniting some of the fuel/oil mix in the bearing tunnel. The outer races of both bearings look identical heat-wise - just the inner races show a good bit of difference in how much heat they took. The preload spring that is between the bearing sets has also lost all of its temper so it did get exceptionally hot within the bearing tunnel - dont know why I didn't see this at the thermocouple reading though - I'll have to check the calibration of that again.

I'm going to disregard what I'm seeing in the bearings until I rebuild the combustion chamber and get the engine running properly. There's just too much going wrong with the burn and that is going to throw everything else off.

[ashpowers]

OK, I figured out why I was not getting proper readings on the bearing tunnel. The probes I was sold are GROUNDED to the sheath. GRRRRR! The manufacturer did not tell me this when I purchased them and honestly, I didn't know they made them with the junction connected to the sheath - seems idiotic to do that when used in more complex electrical systems. They do have a quicker response because the junction is physically touching the sheath but it creates a ground loop and this was pulling the signal down - as well as the EGT probe too, so she did get hotter than what the readings were telling me. Fortunately I never ran her for any length of time to soak the turbine to failure but the parts got hotter than what my equipment was telling me.

So, I've ordered new thermocouples, the ungrounded type which will be here tomorrow and I'll re-test the setup.

I've separated the outer liner from the rest of the combustion chamber so I have access to the vaporizers. They definitely were not seeing much heat at all. Problem is that the dilution ports on the inner liner are kindof high up and making the vaporizers much longer will expose them to this cooler air as well as disrupt the flow path of the dilution air.

I'm going to sit on it for a day or two to make a decision on what direction to go from here. Really thinking I need to start from scratch with this one. Pondering to go with 8 vaporizers that run along the inner wall of the outer liner with side-feed ports like what is shown in one of my previous drawings here:


This would be more of a conventional approach to the design - simiar to 9-94, Fat Mumma, and other production turbines. Or going with Phil's Phoenix layout... Either way, a day's work. Leaning towards Phil's layout though..

[racket]

Hi Ash

With our TV94 engines the flametube holes are positioned so that they lie between the evaporators , the airjets entering the flametube will "entrain" some hot flametube gases resulting in hot gases circulating up and around the evaporators.

With the "side entry" design you'll need to be careful that the airflow to the back of the flametube, and any dilution holes positioned there, isn't restricted by the fuel manifold, and is probably the reason why the rear entry has become more or less the standard design .

The side entry does make for more turbulence within the evaporator which should force the fuel to the walls of the evap, but the rear entry evaps seem to work OK as long as the fuel is squirted onto the inner wall and not straight down the evap axis, the "D'ing" of my evap tubes causes considerable turbulence and fuel "movement" on the evap wall.

Cheers
John

[ashpowers]

OK, got all the goodies in this morning. =)

After more consideration I think I'm going to move forward with a layout like the Phoenix 30.3. Its a proven design and not terribly difficult to build. As much as I would like to re-invent the wheel here, I do have a much larger engine I want to get around to building. ;-)

The main purpose of resurrecting this engine was to take another stab at the ball bearing configuration; employ some additional engineering into the setup to better preserve the rotorbearing group and take from that to the TV94 engine's design. I would just like to see the T04 engine run reliably, throw an ST50 freepower on it and chain her up to the scooter to rile the neighborhood kids up now and then. =)

I'll post pics up soon as things come together. BTW, these new OMEGA thermocouples and their modular connector cables are NICE pieces!!

-Ash