"Fat Boy" 12/118 thrust engine

[racket]

Hi Gavin

I guess it'll increase the lubes "pressure drop" across the bearings and effectively increasing flow, don't know , never ran a scav pump on any of my engines .

Cheers
John

[racket]

Hi Guys

Been busy making up the main supply lube pump and filter/relief valves, used the same pump as the scavenge and "repurposed" an old Subaru pump for the filter/reliefs , its sorta looking OK ,but might scrap the idea if it doesn't "look right" when finished.





Cheers
John

[racket]

Hi Guys

Getting some progress with the lube system , tank made , oilpump/s etc mounted , need some "specialised" tubing to finish the job , but as I'm self isolating to try and not catch the virus they'll have to wait .

Stay well everyone .

A few pics of the bits










Cheers
John

[racket]

Hi Guys

Getting bored being stuck at home without any material to continue with the freepower stage so decided to have a look inside Fat Boy to see how things are coping after a few test runs .

Once the outer can was off it was obvious that there weren't any problems , just a bit of extra soft soot in the lower half of the flametube from post shutdown burning of some lube bypassing the seals due to lack of a scavenge pump.



The outer wall of the flametube was looking good , combustion heating right up to the front wall , an even coverage all around the flametube





Removed the rear wall of the flametube for a look inside and found signs of combustion against the front wall, with an occassional "hotish" spot on the side walls, nothing too dramatic as the outside of the wall at those spots are still nice and smooth



the 18 air distribution caps feeding the Primary air into the flametube were still intact , though the alignment of vapouriser tube outlets and caps seems to have "adjusted" themselves a tad , probably because the rear wall that the tubes are fixed into has had a hard time of it and has inbuilt stresses and some buckling due to mistreatment over the years from all the "failures".

The "tubes" all look good , theres signs of being hot but not "burnt" as their surfaces are still nice and smooth with some "silver" towards the inlet end




I think I'll just reassemble her

Cheers
John

[turboron]

John, very impressive outcome considering all the issues you work through to get it to this point.

Thanks, Ron

[racket]

Hi Ron

Yep , I'm pretty happy with the outcome, there are still some issues but considering its most of the way there I'll leave well enough alone .

I'd like to experiment with using the dramatically clipped turb wheel again now that I've got the combustion sorted , I feel that combustion was the main reason for a lot of the earlier problems , I simply couldn't burn sufficient fuel fast enough before the combustion left the Primary Zone and started moving rearwards , the flametube dimensions are less than ideal , too short for the annular width , theres plenty of small RC turbines with short flametubes but their length/width ratios are >50% better , the little JetJoe flametube Smithy gave me years ago and which I've often used for inspiration has a length of ~48 mm but an annular width of only 22 mm , so better than a 2.1:1 L/W ratio , the 12/118 flametube is only ~1.2 :1 , its just too short for the cross sectional area required to keep air/gas speeds within reason.

The changes to the vapouriser tubes with the addition of their "square springs" to keep the fuel centrifuged against the wall and "distributed" through the "culverts" produced by the "square" spring shape, and the JP-3 type fuel appears to have produced plenty of fuel vapour that the Primary air caps can use effectively to get combustion going immediately so as not to waste any of the flametubes precious axial length.

The NGV probably needs to be redesigned with less throat area and a "lower" tangential gas angle now that I have the large G Trim turb wheel installed , the lower angle will produce a bit more gas deflection which in turn means less pressure drop required and denser gases with maybe a tad more flow ...............LOL, just the usual juggling act with turbine flows .

Its been an "educational" journey that has taught me a lot which will hopefully be of some use down the track, if not to me then someone else .

Cheers
John

[madpatty]

Apr 13, 2020 17:04:12 GMT -5 racket said:

Hi Ron

Yep , I'm pretty happy with the outcome, there are still some issues but considering its most of the way there I'll leave well enough alone .

I'd like to experiment with using the dramatically clipped turb wheel again now that I've got the combustion sorted , I feel that combustion was the main reason for a lot of the earlier problems , I simply couldn't burn sufficient fuel fast enough before the combustion left the Primary Zone and started moving rearwards , the flametube dimensions are less than ideal , too short for the annular width , theres plenty of small RC turbines with short flametubes but their length/width ratios are >50% better , the little JetJoe flametube Smithy gave me years ago and which I've often used for inspiration has a length of ~48 mm but an annular width of only 22 mm , so better than a 2.1:1 L/W ratio , the 12/118 flametube is only ~1.2 :1 , its just too short for the cross sectional area required to keep air/gas speeds within reason.

The changes to the vapouriser tubes with the addition of their "square springs" to keep the fuel centrifuged against the wall and "distributed" through the "culverts" produced by the "square" spring shape, and the JP-3 type fuel appears to have produced plenty of fuel vapour that the Primary air caps can use effectively to get combustion going immediately so as not to waste any of the flametubes precious axial length.

The NGV probably needs to be redesigned with less throat area and a "lower" tangential gas angle now that I have the large G Trim turb wheel installed , the lower angle will produce a bit more gas deflection which in turn means less pressure drop required and denser gases with maybe a tad more flow ...............LOL, just the usual juggling act with turbine flows .

Its been an "educational" journey that has taught me a lot which will hopefully be of some use down the track, if not to me then someone else .

Cheers
John

Hi Racket.

Though I agree with the L/D ratios of greater than 2.

Isn’t it counter intuitive that if you reduce the D(annular width) of the combustor the L(length) requires also reduces and vice versa.

Shouldn’t the volume be a better measure?

Thanks

[racket]

Hi Patty

The 12/118 flametube has a volume of ~200 cubic inches - 0.115 cubic feet and I burn ~4.5 lbs/min of fuel at 3.5 atmospheres , the combustion intensity is kinda high ........LOL, do the numbers :-)

If I reduce the cross section to improve the L/W ratio , the combustion intensity goes even further into "difficult" territory, but more importantly the Primary Zone air speeds will go up , further creating combustion problems

The distance between comp and turb are fixed by the shaft length, and with a need for airflow passageway to the inner wall/s the flametube dimensions are pretty well fixed .

The volume could be increased to improve dwell time , but that only makes the L/W ratio even worse :-(

Suggestions gratefully accepted .

Cheers
John

[turboron]

John, it might be instructive to compare your liner's L/W ratio with recently designed aircraft gas turbine which use annular combustors. The Rolls-Royce/Allison AE combustor would be a good example. It was designed in the late 1980's. My guess is the L/W is near 1/1 from online pictures. The main difference is that the AE uses atomizing fuel nozzles versus your vaporizer tubes.

Thanks, Ron

[racket]

Hi Ron

Would you have a Link for that combustor.

Andy M didn't appear to have problems with his similar length combustor jetandturbineowners.proboards.com/thread/1025/hx-102-255-money-pit maybe the direct injection/atomisation provides air/fuel mixing as the fuel exits the injector virtually parallel with the front wall of the flametube which is riddled with air entry holes , the combustion might be getting anchored right up against the front wall unlike the 12/118 combustor in its original configuration which had a "cold" front wall with combustion not beginning until some distance downstream of it , like most micro RC turbines www.youtube.com/watch?v=yWfWlEUp4MY .

Now that I've installed the air caps to supply Primary air parallel with the front wall , the fuel/air vapour exiting the vapouriser tubes impacts the cap/frontwall and spreads out across mixing with the primary air and combusting virtually right against the wall by the looks of the colour witness marks.

I've been looking back through my testing journal and theres been some big changes in the T2 vs P2 efficiency numbers , from low 70s comp efficiency in the early days until low 80s in the most recent testing , best of 83% at a 2.5 PR with the jet nozzle test , and even an 84% with the freepower test that may have reduced mass flow a tad more and into a better "island"

Maybe I should try and rig up a thrust testing arrangement to get some definitive numbers as I'm currently only speculating on the thrust level from calcs using P4t , T4 and jetnozzle sizing which are all roughly ballpark from my original calcs of 4 years ago .

Cheers
John

[finiteparts]

John and Ron,

Comparing the L/h ratio of modern engines would be very unfair because most of them would have the advantage of higher cycle pressures and temperatures. Reaction rates scale to the 1.8 power of pressure and exponentially with inlet temperature, so our low pressure/temperature machines would be highly misrepresented by such comparisons. If you look at the evolution of combustors in turbofans you will see their dramatic reduction in size which corresponds to the dramatic increases of core pressures and thus combustor inlet temperatures.

- Chris

[turboron]

Chris, can you explain your post in more detail? I looked through my Lefebvre for insight with no joy. What is the effect of the reaction rate scale on the volume of a combustion liner that is subject to an increased pressure ratio, for example. Your post implies that the higher pressure liner will be smaller or that the mass flow could be increased.

Thanks, Ron

[finiteparts]

Apr 15, 2020 11:59:02 GMT -5 turboron said:

Chris, can you explain your post in more detail? I looked through my Lefebvre for insight with no joy. What is the effect of the reaction rate scale on the volume of a combustion liner that is subject to an increased pressure ratio, for example. Your post implies that the higher pressure liner will be smaller or that the mass flow could be increased.

Thanks, Ron

Ron,

Yes, the higher pressure, higher combustor inlet temperature works to reduce the required volume.   Now of course there are tons of other things at play too, but the pressure/temp are very large influences on required combustor volume.

If you look in the Combustion Efficiency chapter in Lefebvres, the theta parameter (Loading parameter) is discuss multiple times.    In my first edition, there is a section on "Reaction-Rate-Controlled Systems" and he walks through his Burning Velocity model.  The burning velocity model is based on a sort of quasi-Arrhenius rate equation that has been correlated to data.  If you look at the loading parameter, it is apparent the 1.8 power on pressure and the exponential dependence on temperature. 

If you remember back to basic chemistry, this should all make sense.  The temperature dependence is easy to understand since higher temperatures mean that the gas molecules have higher bulk averaged kinetic energy and thus the probability of collisions between moleculers is greatly enehanced (remember that reactions occur at the molecular level...so they need to come into contact with each other to exchange electrons and atoms...i.e. collisions).  The pressure dependence can be thought of in a similar manner.  Higher pressure means a higher density of molecules in a given space, thus there is a higher probability of molecular collisions.

A second factor is simply the basic gas dynamics of increased pressure and temperature.  Typically, the increases in compressor pressure ratios has been accompanied by increases in compressor efficiencies, but even if you hold the same level of efficiency, the temperature rise in the compressors does not rise as fast as the pressure and thus you get an increase in density.  This means that you need a smaller through flow area to achieve the same Mach number in the combustor.  So if you are trying to scale a combustor and hold the internal local velocities the same, you make the combustor smaller.

Now of course, the other "big" factor is the need to reduce emissions.  One of the primary mechanisms to form NOx (oxides of nitrogen) is the thermal pathway ("thermal NOx") and this mechanism drives combustion designers to minimize the residence time of the combustor to as low as possible in order to minimize the reacting species time at temperatures where the Zeldovich NOx mechanisms can occur.  This has substantially reduced combustor sizes, but of course during this same time the combustor designers have developed much more efficient swirlers and primary airflows so that the reduced residence times do not negatively impact the combustion efficiencies and cause the CO emissions to spike.

I hope that helps,

Chris

[turboron]

Chris, thanks very helpful. I have the 2nd edition of Lefebvres book. I will it check tomorrow and study the relevant sections. I may want to ask you some additional questions after my study.

Thanks again, Ron

[turboron]

Chris, the 2nd edition also uses the name theta. There is no reference to "loading parameter" that I saw. In any case, when I went through the numbers for my combustor using a 2.457/1 pressure ratio, mass flow in lbs/minute and the dimensions in inches I calculated a Theta of .7447 times 10^-4. Figure 5-3 indicates a combustion efficiency of 90%. If I increase the pressure ratio to 4/1 while holding mass flow constant I get a Theta of 2.8772 times 10^-4 which is still a 90% combustion efficiency per Figure 5-3. Do these number seem correct?

Thanks, Ron