"Fat Boy" 12/118 thrust engine

[racket]

Hi Patty

I have 2 X 12 V batteries coupled in series to give 24 volts , their cold crank amp ratings are >300 A.

The positive cable goes through a 24V 500A on/off isolation switch to a 24V 500A solenoid which is operated by 24V going through a 40A 12V relay activated by a toggle switch thats connected into a micro switch at the starter mount on the bellmouth , the micro switch deactivates the starter before the coupling socket disengages from the comp nut, it also makes it impossible for the starter to operate unless engaged onto the rotor.

Currently the starter can do maybe 3 spoolup sequences of ~20 seconds each in quick succession before it gets a bit hot to handle, the earth negative cable runs from the starter gearbox back to the battery/s.

At present with colder conditions the batteries aren't as "strong" and aren't able to spool the engine past 10,000 rpm with 70 psi of cold oil pressure , in the warmer months I'm able to get >12,000 rpm and the starter gets too hot to handle after 3 quick sequences .

The battery/s voltage drop from 26V to 8V after activating the toggle switch and slowly climbs up to 11V during a "dry" 20 second spoolup attempt, the starter is sucking the guts out of the batteries, its only a 1.4Kw starter on 12V .

For the next test run I'm currently experimenting with warming the batteries prior to driving out to the test area to see if it improves spoolup times at 10 deg C ambient conditions

Cheers
John

[racket]

Hi Guys

New test run done this morning , engine hotter, nozzle size needs to be increased .

youtu.be/s5XLZVBQL-8

A couple more vids to come

Cheers
John

[racket]

youtu.be/6A2sxQq4WJg

youtu.be/ImXN4HQxDVI

[racket]

Hi Guys

Some "numbers " from todays test run vs last test





Jetpipe temps were higher across all P2 power settings due to the jetnozzle being a tad too small in area

Fuel flow/pressures were similar at all P2 power settings .

T2 minus T1 were pretty similar

Where there was a big difference was in the jetpipe total pressure P4t its was considerably higher , at a 2.5 Bar P2 it increased from 0.63 Bar -9.2 psi to 0.86 Bar -12.6 psi , with a 23 pounds thrust increase from 155 lbs to 178 lbs .

Unfortunately I wasn't able to get data at 2.8 Bar P2 as the fuel tank was getting very low and air was being sucked into the fuel pump , .............I'd "wasted" too much time taking RPM numbers , but they needed taking, the quality industrial tach refused to work , maybe the RPM were too high , so I had to resort to a hand held one which wasn't the easiest to get a reading from , but the RPM are similar to what I was getting 3 years ago so alls OK there .

The jetnozzle needs to be opened up by a few more millimeters to keep temps down .

Cheers
John

[turboron]

John, sounds like steady progress. What is you overall goal at this point? 200 Shp?

Thanks, Ron

[racket]

Hi Ron

I think 200 SHP is realistic , I found with my TV84 engine that its 110 lbs of thrust as a pure jet converted to 115 HP on the bike dyno test , so my current 190 lbs of thrust should give 200 SHP.

The current design of the 12/118 was for a thrust engine using a turbine wheel that had been heavily clipped back to increase mass flow , and relying more on impulse energy from a choke NGV , but now I find myself with a NGV thats grossly oversized and feeding a choked wheel exducer thats powering the comp from reaction energy .

I'll run some more numbers today to better understand the "Corrected Flow" across the turb stage to see if anything can be "massaged" to improve the situation .

The current 61 Trim for the comp wheel isn't ideal for a shaft horsepower engine as its undoubtedly compromising things at higher P2 pressures , unlike Anders engine with its 48Trim which will probably be able to produce the same amount of thrust but from a higher P2 and P4t and is ideal for the 2 stages of expansion through the freepower.

LOL, I'm caught with a combination of parts that aren't the best match , but the positive is , its forced me to find ways around problems that shouldn't have been there in the first place .

Andy M's 110 mm inducered comp is probably a better match for our 129/112 mm turbine wheels , the lower mass flow should mean flowing in higher efficiency islands of the map , the NGV can be produced with a "lower" angle giving more impulse energy to the turb wheel, whilst the exducer throat area should still be capable of flowing his 3 lbs/sec design flow.

Without any turbine map for the TV94 wheel its been a bit of a trial and error approach to figure out where its max flow is , at present the oversized NGV will be providing fairly modest gas velocities feeding into the turb wheel at less than ideal angles , the wheel's inducer tips are probably having to "accelerate" the gases "circumfrentially" as they go in , this is sucking power from the wheel leaving less for the comp and requiring me to substitute heat energy in the form of higher temps to compensate ................its a bit of a "dogs breakfast" , not very tidy , but its working :-)

Cheers
John

[turboron]

John, don't you just enjoy making the numbers and figuring out what the machine is doing? The gas turbine instantly solves the must complex flow equations while we scratch out heads and mumble.

Thanks, Ron

[racket]

Hi Ron

LOL.......yeh , the numbers seem to confuse things at times :-)

With my latest numbers it looks like I was getting very close to having a choked jet nozzle with a 1.86 PR in the jetpipe at the 2.5 Bar P2 setting , at a 2.8 Bar setting it would have been ~2.0 PR if we go by the previous tests P4t increase from 2.5 to 2.8.

My thoughts that the turb exducer is choking appear to be wrong , with a 3.5:1 PR ( 2.5 Bar P2 ) from the comp and with a 5% pressure drop across the flametube theres only a 3.325 PR going into the turb stage , and with a 1.86 PR in the jetpipe that only leaves a ~1.78 PR across the turb stage , so not enough pressure drop to choke anything .......bummer , no I've gotta find another culprit :-(

The 160 C deg temp rise across the comp at the 3.5 PR ( at 76.5% effic) will require an ~135 C drop across the turb stage which appears to be working at ~82% efficiency .

With a 1.86 PR across the jet nozzle from a temp of ~812 C average and 90% nozzle efficiency we should get a velocity of ~1880 ft/sec and a density of ~43 cu ft per pound , but 178 lbs of thrust from 1880 ft/sec only requires a tad over 3 lbs of mass flow and a theoretical nozzle of ~91 mm dia without boundary layer allowance , my current nozzle is nearly 98 mm , thats a fair bit of boundary layer .

It might be time to do a completely new set of calcs for the engine using the data I now have including at least a 10-15% decrease in mass flow from the earlier set of calcs, I think I might need to blank off a few NGV passageways

The saga continues :-)

Cheers
John

[racket]

Hi Ron

I need to blank off 3 NGV passageways , ~17% reduction to better match a 10% reduction in flow through the choked exducer and to provide a tad more gas velocity going towards the turbine wheel to better match blade speed at the 3.8 : 1 PR , basically fitting a smaller A/R scoll housing , hopefully these changes will reduce temps at higher PRs allowing them to be "explored" , might even get to 4:1 :-)

Currently the NGV throat area is a tad larger than the wheel exducer throats, not the right way round, it was OK with the highly clipped wheel and its more "open" throats .

Cheers
John

[turboron]

John, what I hear you saying is that you are going to improve the velocity triangle into the radial inflow turbine. Velocity triangles have always been a mystery to me. Can you say a few words to improve my understanding of the improvement to be gained?

Thanks, Ron

[racket]

Hi Ron

LOL.......now this will be "interesting", where to start :-)

I tend to work on temperature degrees , so if the comp temp rise is say 100 centigrade degrees then the temp drop through the turb will be 100 X 0.24 ( Cp air) divided by 0.28 (Cp gases) , so ~86 C degrees

100 C degrees during the compression equates to , .....100 X 0.24 X 1400 ( ft pounds/degree) divided by 550 ft lbs/hp/sec , ~61 HP

The 61 C degree drop through the turb stage can be converted into "gas deflection" required when the mean blade speed is taken into account , with our radial info turbines determining the mean blade speed is awkward , so lets use an axial wheel for simplicity with a mean blade speed of 700 ft/sec , we multiply our 61 X 32.2 ( gravity) X 1400 ( ft lbs/C degree) X 0.28 ( Cp hot gases) and divide by our 700 ft/sec blade for a gas deflection of 1100 ft per second .

From those numbers we can see that with a mean blade speed of 700 ft/sec that the velocity triangles can be juggled a bit to produce the 1100 ft/sec, we could have a "tight" NGV giving high gas velocities and providing that "extra" 400 ft/sec of deflection , with the gases exiting the turb wheel axially and providing no "extra" deflection , only keeping up with blade speed .

Or we can use a more open NGV with less gas velocity and having the gases entering the wheel axially but exiting with considerably higher velocity producing swirl in the exhaust as well as the 400 ft/sec of deflection.

Another method is to split the pressure drops in such a way that neither the inlet or exit of the wheel is axial , there being 200 ft/sec of deflection at inlet and exit , but that is predicated on the various components having the flow capacity to do so.

With my engine the turb wheel exducer is the limiting factor flow wise and I need to juggle the velocity triangles to maximise the flow through it whilst still producing the required gas deflection and horsepower to power the comp .

A problem I run into is if I can't get sufficient deflection at higher P2s to power the comp with the "angles" I have , then I need to increase the gases Cp to compensate , but hotter gases are less dense even if they travel faster so mass flow suffers, and theres a limit to how hot I can make them whilst keeping the turb wheel in one piece.

My engine was originally designed for 3.6 lbs/sec , but with the G Trim turb wheel installed I think that needs to be reduced to ~3.3 lbs/sec , or maybe even a tad less, but my NGV throats were sized for the extra flow so currently their gas velocity production will be low and along with that the gas deflection component of the comps horsepower requirement , I'm relying more/entirely on the gas deflection produced by the wheels exducer angle , which is mildly clipped as standard.

At present , probably once past the RPM of a ~3:1 PR there is "negative" deflection at the inducer tip requiring even more deflection from the exducer , but if its already choked it can't produce more , so my temps need to rise sharply to compensate , which they are , some 200 C degrees between 3 and 3.8 PR .

I'm hopeful that by reducing the NGV area the gas velocity increase will remove that "negative" deflection with the need for only a mild increase in PR across the NGV , just a couple of psi , the increased tangential velocity also comes with increased radial velocity to help get those reduced density gases into the wheel , the increase in radial component of the velocity triangle also means less pressure drop required in the wheel to get a "sonic" velocity at the exducer throats , so what pressure we lose in the NGV we hopefully gain in the wheel, but its the removal of that "negative" deflection thats most important.

Clear as mud :-)

When I was playing around with the TV84 engine it was only once I was running the tightest scroll that a 4:1 PR was achievable , with larger scrolls the temperatures were getting out of control at lower and lower P2s as the scroll A/R increased , the smaller A/Rs produced more inflow deflection relieving the choked exducer from some of its power producing duties.

LOL........its a bit more complicated than this because the varying scroll sizes also change the mass flow and along with that the comp efficiencies which in turn changes the power requirements and the required deflection , thankfully our radial turb stages are rather forgiving of our mishandling and tend to "self sooth"

The Garrett turb maps make life a lot simpler as its possible to easily match comp and turb flows.

Hope this goes some way to help explain things.

Cheers
John

[turboron]

John, this nuts and bolts explanation is very helpful in understanding your corrective action. It sounds like you are trying to achieve 50% reaction (equal pressure drop in the stationary NGV and the rotor) by juggling the NGV throat area. I need to look up the velocity triangles for 50% reaction.

Thanks, Ron

[turboron]

Johm, 50% turbine reaction velocity triangles.




Thanks, Ron

[racket]

Hi Ron

Yep , thats getting like it .

Its more efficient to have two medium gas velocity increases than a large and small one , its that square root part of the equations.

Cheers
John

[racket]

Hi Guys

Bit of an update ...............I've been getting the jetnozzle opened up by a few more millimeters to lower temps and doing some battery checks and decided to remove one of the old batteries and reposition its supply to the main teststand battery as I only need 12V now since removing the 24V PWM control on the old fuel pump.

I was thinking of blanking off 3 of the 18 NGV passageways to decrease the NGV flow area and increase gas velocity into the turb wheel to better match turb inducer tip velocity at higher P2 speeds ,..............a nice simple solution I thought ..............BUT , then I realise the turb exducer would have less effective flow area as those blanked off NGV passageways wouldn't be feeding any gases into the turb wheel passageways as they passed them , so back to the drawing board :-(

Instead I've been modifying an older jetpipe/exducer shroud to provide ~2 mm of extra radial clearance at the throats , this should provide at least another 10% of flow area at that restriction .

The extra flow through and around the exducer would mean more flow through the NGV increasing the gas velocity without any penalty hopefully as the comp is more than capable of flowing extra .................LOL, I really have no idea what will happen , but it'll be worth trying , 90% of the flow will be getting deflected in the exducer as per normal operation and helping to power the comp and the bit of extra speed from the NGV might just make up any shortfall from the 10% bypassing the exducer ...........lotsa unkowns , but the "numbers" should soon indicate whats happening .

I'll fit the 107mm diameter jet nozzle I used in earlier tests when I had the test stand "dancing" , and I might open it up to ~112mm to cut back a bit on the backpressure to help with temperatures if theres more mass flow .

I love experimenting with something new :-)

Cheers
John