What was so amiss with the engine congifuration on the General Electric CJ-805-23s used on the CV 990 (fan stage at the rear of the engine driven directly by the low-pressure turbine) that this configuration was never used again by another engine? Surely the weight saved by deleting the fan shaft must have been substantial.

Can this configuration be usefully exploited by a modern high-bypass engine?

Faro

Just a guess but I'm thinking the heat that the fan hub is subjected to in this layout combined with the huge centrifugal and torsional forces exerted on it by the fan blades would lead to poor service life. Maybe these problems could be engineered out but that would probably negate a good part of the weight savings. In addition you lose the fan as the first stage of compression for the core in this layout necessitating an additional compressor stage, further offsetting any weight savings.

Another issue would be supporting the turbine/fan combination. I don't know how they did it in the CJ-805 but without much of a shaft you need heavier bearings to effectively support that rotating mass, again marginalizing weight savings gained by eliminating the shaft.

DL757Md

Quoting Faro (Thread starter):

Perhaps the technology was practical for an engine the size of the CJ-805, but impractical for scaling to the size of a modern turbofan. If we look at picture of the CJ-805, we see that the aft fan blades are mounted directly to the tips of the turbine blades.



The problem I see in scaling the design, are the huge centrifugal forces generated by large fan blades, as well as thermal issues. IIRC, modern turbofans such as the RB211-524G on the 744 have a centrifugal force at each blade root of around 80-90 tonnes.

The LP turbine blades of a shaft-less version would need to be incredibly beefy to withstand this force, and such a shape may be very aerodynamically in-efficent. The additional centrifugal forces would also add immensely to the problem of turbine blade creep.

You would also then have to deal with the thermal problem of having the turbine blades subjected to 700-800 degrees gas temperature, whilst the fan blades are subjected to temperatures that may be as low as -60 degrees. If anything, you would be losing much heat and efficency due to heat conduction from the turbine blades to the fan blades. I'm not sure how a blade material could withstand this temperature differential in addition to the problem of creep and centrifugal force.

Regards, JetMech

[Edited 2007-12-27 08:33:46]

Good analysis JetMech. We haven't given up on this idea but the problems are just as you say. We now have much better alloys than we had "way back when", so we have more choices for the LPT blade and fan blade materials today. If you imagine a bimetallic system you could certainly make this work again. The question is whether it would win out over a more conventional architecture that also used today's materials. So far, it doesn't. Cheers

GHR

The reason the CJ-805-23 used an aft fan was because Convair was literally obsessed with the CJ-805 because of it's extremely responsive characteristics. They wanted a turbofan, but they didn't want to part with it's lightning quick spool up rate.

Re-designing the engine, would also be relatively costly and time-consuming too and routing a shaft through the main spool and adding an LP compressor a fan and stuff might affect spool-up rate too.

So they just added an aft turbine with fan blades on the outer-ring which gave the extra-thrust, it actually allowed a higher bypass-ratio (2.2 : 1) required minimum modifications, and the spool up rate was still quite good being only a fraction of a second behind the rest of the engine.

I did not consider the extra structural strengthening that would be required for a shaftless (aft) fan, but I know there is one thing that is definetly not desired about it. The fan can't double as a compressor stage and can't contribute to the engine's overall pressure-ratio.

Still if you want to add a fan to an engine, particularly a single-spool jet with a lightning fast spool-up rate without extensive modifications that's the way to go about it.


Andrea Kent

It's just too bad that the J-79 engine core didn't realize many effeciency gains when the fan section was added to make it the CJ-805...   The CV-880's and CV-990's had a reputation, even in the '60s when fuel was cheap, as being fuel hogs.

Was this due to "too much power", or just horribly inneffecient engines? How did the CJ-805's SFC stack up to other contemporary engines of the day, like, say JT3D's, JT4's, and the RR Conway?

KELPkid,

I would think the J-79/CJ-805 would be better than a J-57/JT3C because it had a higher pressure-ratio which is a good gauge of efficiency, could be wrong though. The CJ-805/J-79 did have inefficient combustion, though -- that's why it trailed such an awful black cloud of smoke.

In regards to your question of "too much power" or horribly efficient engines, I'd say it was mostly the latter. Turbofans really helped things out, but also improved combustor design (more efficient use of combustion space) better turbine metallurgy (allowed for higher pressure-ratios), and improved compressor-blade design (more pressure ratio with the same, or even less compressor blades)[/i].


Andrea Kent

Quoting Jetlife2 (Reply 3): We haven't given up on this idea but the problems are just as you say.

I've been thinking some more about the various issues. Perhaps an aft fan could be designed such that the ring of material between the turbine and fan blades is made radially thicker and axially wider. It could then act as an annular fan disc, and react the centrifugal forces of the fan blades as pure tensile loads in the annular ring. The extra axial width of the annular disc will allow a normal, fir tree fan blade attachment to be made.

The creep issue could be mitigated by having the turbine blades designed such that they transmit torque forces only. Centrifugal loads can be prevented from reached the turbine blades by having a connection to the annular fan disc that is free in the radial direction. The thermal difference issue could be minimalised by using ceramic insulators between the tips of the turbine blades and annular fan disc.

If the turbine blades had a radial pin at their tips to transmit torque, a cylindrical shaped ceramic insulator could be made to fit over it. The outside of the cylindrical shaped insulator then fits into a corresponding radial hole on the inside surface of the annular fan disc. This gives the radial freedom required to prevent transmission of centrifugal loads to the turbine blades, and ensures that the ceramic insulators are loaded in compression only when transmitting torque. If the subsequent conductive heat transmission is low enough, the fan blades could then be made of a material with a very high strength to weight ratio such as advanced composites.

This would leave the problem of accurately placing the axis of rotation of the turbine assembly concentric with that of the annular fan disc and fan blades. For this, I propose two sets of concentric bearings. One set would support the turbine spool, and would be similar to the arrangement already used in jet engines.

The bearings for the annular fan disk would be large, ring like assemblies of the same diameter as the annular fan disc, similar to the bearings used on the rear hot stream nozzles of the RR Pegasus engine (Harrier). This second set of bearings would mount to the front and rear edges of the annular disc, and the outer parts of the LP turbine casing.

The annular fan disc bearings could also be designed to seal the gas path. The casing structure around the LP turbine and annular fan disc would then be designed to support both sets of bearings, thus ensuring that the spin axes of the various rotating turbo-machinery components are concentric.

The remaining problem would then be issues with annular fan disk growth due to tensile straining, and somehow allowing for this in the bearing design without compromising concentricity.

Regards, JetMech

[Edited 2007-12-27 21:25:35]

Quoting JetMech (Reply 7): Quoting Jetlife2 (Reply 3): We haven't given up on this idea but the problems are just as you say. I've been thinking some more about the various issues. Perhaps an aft fan could be designed such that the ring of material between the turbine and fan blades is made radially thicker and axially wider. It could then act as an annular fan disc, and react the centrifugal forces of the fan blades as pure tensile loads in the annular ring. The extra axial width of the annular disc will allow a normal, fir tree fan blade attachment to be made. The creep issue could be mitigated by having the turbine blades designed such that they transmit torque forces only. Centrifugal loads can be prevented from reached the turbine blades by having a connection to the annular fan disc that is free in the radial direction. The thermal difference issue could be minimalised by using ceramic insulators between the tips of the turbine blades and annular fan disc. If the turbine blades had a radial pin at their tips to transmit torque, a cylindrical shaped ceramic insulator could be made to fit over it. The outside of the cylindrical shaped insulator then fits into a corresponding radial hole on the inside surface of the annular fan disc. This gives the radial freedom required to prevent transmission of centrifugal loads to the turbine blades, and ensures that the ceramic insulators are loaded in compression only when transmitting torque. If the subsequent conductive heat transmission is low enough, the fan blades could then be made of a material with a very high strength to weight ratio such as advanced composites. This would leave the problem of accurately placing the axis of rotation of the turbine assembly concentric with that of the annular fan disc and fan blades. For this, I propose two sets of concentric bearings. One set would support the turbine spool, and would be similar to the arrangement already used in jet engines. The bearings for the annular fan disk would be large, ring like assemblies of the same diameter as the annular fan disc, similar to the bearings used on the rear hot stream nozzles of the RR Pegasus engine (Harrier). This second set of bearings would mount to the front and rear edges of the annular disc, and the outer parts of the LP turbine casing. The annular fan disc bearings could also be designed to seal the gas path. The casing structure around the LP turbine and annular fan disc would then be designed to support both sets of bearings, thus ensuring that the spin axes of the various rotating turbo-machinery components are concentric. The remaining problem would then be issues with annular fan disk growth due to tensile straining, and somehow allowing for this in the bearing design without compromising concentricity. Regards, JetMech

Many thanx JetMech for your thoughtful and detailed input. I guess that perhaps one day -materials and technologies permitting- this configuration may well reappear as outlined by Jetlife2.

Faro

Quoting JetMech (Reply 2): The LP turbine blades of a shaft-less version would need to be incredibly beefy to withstand this force, and such a shape may be very aerodynamically in-efficent.

I recall reading that when the CV-990 was being designed that Boeing engineers predicted that it would not realize its objectives because of the aerodynamics of the engine/wing interface. They had done extensive research on engine mountings for the B-47 which carried over into the B-52 and 707 programs; they determined that there must be a substantial gap between the engine and the wing or excessive drag resulted. Later they discovered that mounting the engine almost completely in front of the wing worked also, hence the 737 Classic (and later the 777.) But a large engine directly under the wing caused huge problems, and the CV-990 suffered because of it. This makes the concept of an aft fan much harder to realize, because it is dictating that the fan must be substantially forward of the wing. If the engine is forward of the fan this causes huge structural problems. The other solution (hanging the engine far enough below the wing to give the necessary clearance) would require ridiculously long landing gear. Of course you could go back to tail-mounted or overwing engines, but those arrangements have their own problems.

There actually is another engine with an aft fan, the CF-700 series engine by GE. It is the original powerplant for Falcon 20's and most Saberliners. It has a 2 stage aft mounted fan on it. The front part of the engine is basically a CJ-610, which powers 20 series Learjets. The CJ-610-6 has about 2950 lbs of thrust and by adding the rear mounted aft fan, the CF-700-2D2 gets upwards of 4500 pounds of thrust. The aft fan also makes the engine a lot quieter than the 610 although it is still pretty loud. It is a very reliable engine though and still in use today on the planes I fly but they burn kind of a lot of gas. The aft fan is an interesting concept!

Quoting Tb727 (Reply 10): There actually is another engine with an aft fan, the CF-700 series engine by GE. It is the original powerplant for Falcon 20's and most Saberliners

Just want to clarify, only the Sabreliner model 75A used the GE CF700 engine, this model Sabreliner was not very popular and not many were sold.

Almost all Sabreliners and the military version T-39’s used the P&W JT12 engine, either the –6 or –8 version. Later Sabreliners used the Garrett TFE 731 engine.

According to (cough) wikipedia (cough), one of the issues with aft fans is hot gas leakage from the LP turbine to the fan. Is this in fact not much of problem? Can it be solved easily like previous posts seem to assume?

Quoting SEPilot (Reply 9): Of course you could go back to tail-mounted or overwing engines, but those arrangements have their own problems.

Even so, don't rear mounted engines have certain advantages as well? Can't remember them OTOH, but I think you've answered your own question.

The Unducted Fan of the 1980s had an LP spool that was "shaftless". The current Open Rotor design is very similar.

See: http://www.flightglobal.com/articles/2007/11/30/219990/civil-engines-will-cfm-leap-to-open-rotor.html

And if you can find it, Bill Sweetman's article gives an excellent description and diagrams of how the whole thing worked.

Title The short, happy life of the prop-fan
Authors SWEETMAN, BILL
Journal Title Air & Space Smithsonian Vol. 20 No. 3
Period Covered Aug./Sept. 2005
Publication Date 20050800
Description p. 42-49 : ill.; Illustrations

Quoting Jetstar (Reply 11): Later Sabreliners used the Garrett TFE 731 engine.

Now that is an awesome engine. We have 1 FA20-5 as they are designated after the conversion. What a huge difference in fuel savings.

Quoting Tb727 (Reply 14): Now that is an awesome engine. We have 1 FA20-5 as they are designated after the conversion. What a huge difference in fuel savings

Besides the Lockheed JetStar re-engine program, Hawker 125’s and as you mentioned the Falcon 20 were also re-engined with the 731, and it still is a very popular engine for new corporate jets and will be for many years to come as they keep making improvements and upgrades.

They had a lot of reliability problems back in the early years of the engine, but those have been fixed many years ago and now it is a very good engine