"Fat Boy" 12/118 thrust engine

[racket]

Hi Ron

Even a G Trim is too small , they use them on the PT118 turbo www.precisionturbo.net/products/details/Street-and-Race-Turbocharger---PT118-CEA--174-/329 , my current cutback exducer is already at the ~111mm PT118 exducer dia , they can get away with the "match" between comp and turb because they're probably running higher exhaust manifold pressures than inlet manifold pressures , something we can't do , our turb inlet pressures are always lower, so less dense which means more turb flow area required :-(

LOL...........its a balancing act that I can't seem to master .

Cheers
John

[ripp]

Hi John,

Maybe this design will help








Cheers
Ralph

translate.google.com

[turboron]

John, if you cutback the G trim would that give you enough increased flow area?

Thanks, Ron

[racket]

Hi Ron

Nope :-( the wheel I use is basically the same as a G Trim except for the "Trim" change at the exducer outlet, it comprises an ~10 degree radial taper on my wheel's exducer which is removed when the clipping is done.

The turbine wheel is simply too small for the compressor its driving , I'm attempting the impossible ;-)

Cheers
John

[racket]

Hi Ralph

You might need to explain that design a bit more

Cheers
John

[turboron]

John, if you can not get more horsepower out of the turbine then perhaps the horsepower absorbed by the compressor can be reduced by injecting a water mist at the inlet for example.

Thanks, Ron

[racket]

Hi Ron

Currently the comp is working at ~85% efficiency between 1.5-2.0 PR according to the T2 temps vs PR vs T1 , even allowing for measuring "variables??" its still at better than 80% , which for a centrif comp of this size is pretty good , so its not putting a big load on the turbine wheel and is one of the frustrations with things at present.

When I do the turb stage velocity triangles and power production theres plenty of power being produced to drive the comp , thats why I'm looking at possible mass flow restrictions that may be "lowering" gas velocities through the "oversized" passageways .

Because of the extreme exducer clipping and the ~55 degree exducer angle theres minimal gas deflection from that part of the power production , I'm relying more on "impulse" from the NGV which are at ~33 degrees , the fact that the diffuser exhaust was creating low density gases , even though at high velocity , the mass flow was being restricted by the exducer , this in turn meant less "volume" through the NGV passageways even if they were "choked??", this then reduced gas deflection and horsepower which further exacerbated the situation as higher temps were required to "fill up" those passageways ..............its a very fine line I'm trying to walk with this engine , the theory says it should work , just.

The fuel flow rate vs temp rise in the combustor was indicating a lower than design mass flow for the PR being produced , consistent with there being some sort of restriction

Cheers
John

[turboron]

John, A leak between the compressor discharge and the combustor would also explain the mass flow shortfall.

Thanks, Ron

[racket]

Hi Ron

Its a bit hard to leak air , theres a silicone O'ring sealing the outer can to the comp cover , and a slip joint at the hot end that only has a few thou clearance when cold to allow assembly but which seals when the inner wall heats up , I'm looking at at least a 10-15% shortfall in mass flow :-(

Cheers
John

[turboron]

John, if there is no leak then the compressor curve must not have the raise to surge that you are expecting. If the impeller has radial blades sometimes curve the left (lower mass flow) of the design point is almost flat (constant pressure ratio). Therefore the compressor flow is really 10 to 15 % less that you think at a given pressure ratio.

Thanks, Ron

[racket]

Hi Ron

LOL.............I should have written this up properly rather than being lazy , so here goes.

My fuel pressure at 1 bar P2 was 20 psi , so if we take off the 1 bar we're left with ~5 psi actual pressure drop across the fuel injectors , from flow testing of the injectors that equates to ~1,000cc/min 0.8 kgs/m- 1.75 lbs/min , and with a ~650 C rise in the combustor it requires an~0.017 F/A ratio , so a max air flow of ~1.7 lbs/sec

With a 2.0 PR out of the comp , I'll hopefully have a 1.95 PR into the NGV , assuming its running choked with a calculated gas velocity of 1800 ft/sec and gas density ~41 cu ft/lb at~920K , through an area of 0.048 sq ft of NGV flow area taking into account some boundary blockage from its 0.053 sq ft measured area , 0.048 X 1800 = 86.7 CFS being passed , divide by 41 cu ft/lb = 2.11 lbs/sec ., or nearly 25% more than my fuel flow numbers projected .

Now the velocity triangles produce a radial inflow gas speed of ~1,000 ft/sec into the wheel( 1130 ft/sec with a tad of "angle") which is worth ~40 C degrees , theres a 1.19 PR drop between NGV and diffuser exhaust entrance which is at a measured 2 psi below ambient according to my static pressure pickup, the 1.19 PR drop is good for ~37 C degrees which when combined with the 40 inlet degrees gives me 77 degrees of drop which should produce a velocity of ~1380 ft/sec .

Density at the 12.7 psia at ~900K is ~47.25 cu ft/lb , exducer flow area is ~0.05967 sq ft , 1380 X 0.05967 = 82.3 CFS flow , divide by our density of 47.25 = 1.74 lbs/s , pretty close to my fuel flow "number" .

Now my static pressure probe is some way downstream of the exducer face and there is considerable flow area increase over that distance so one would expect some diffusion and static pressure increase which would make the "numbers " even worse due to even lower static pressures and gas densities.

I've had to make a number of assumptions about things I really don't know a lot about , but my gut feeling is the diffusing exhaust isn't helping matters .

The next test without it should produce some changes in running parameters.

Cheers
John

[turboron]

John, I love it when you talk dirty.

Thanks, Ron

[racket]

Hi Ron

LOL...............glad to oblige :-)

I'm flying a bit blind on this engine , slowly feeling my way along a path that I don't know where we're going , its testing me , but keeping these old brain cells working , if in the end it doesn't work , then so be it , I've learnt a lot along the way about what not to do,and more importantly , why not , something that wouldn't have given me the opportunity if I'd simply made a "normal" engine using conventional well matched bits..............I'm trying to make two wrongs make a right .

Cheers
John

[turboron]

John, I looked on a Garrett turbine map to find the pressure ratio at which the curve flattens. The curves showed the curve flattens at a pressure ratio of 2.25. Using your pressures of 28.7 psia at the NGV(1.95 times 14.7) divided by your exducer exit pressure of 12.7 psia we get a pressure ratio of 28.7/12.7 = 2.25 so the nozzle is choked. I calculate a sonic velocity of 1937 fps using a k of 1.315, a molecular weight of 28.85 and a temperature of 1656.33 degrees R. Instead of calculating a mass flow I use your mass flow of 1.7 lbs/sec and your specific volume of 41 ft^3/lb to get a volumetric flow of 41 times 1.7 = 69.7 ft^3/sec. From this we get an effective throat area of 69.7/1937 = 0.036 square feet. Using your actual area of 0.053 sq. ft we get a blockage of 0.036/0.053 = .68 or approximately 30 percent. This implies a flow separation trigger of some kind such as a forward facing step or flow path ingress of a axial flow air stream.

Thanks, Ron

[racket]

Hi Ron

Different maps show different points at which the "Corrected Flow" rate flattens ,the the TV91 map I have flatlines at 2.0 PR whereas the GT6041 doesn't flatline until 2.75 , a normal nozzle like our NGV should choke ~1.9 PR , but that only happens if downstream "backpressures" allow it to reach that PR , it gets complicated unless we know the actual pressure at every point along the flow path , it'd need several pickup points which is beyond me to provide.

Cheers
John