Centrifugal vs axial compressors in WW2 jets

[wuzak]

An excuse to post one of my favourite pictures the Meteor F1 with Metrovick F2 engines

View attachment 250621

And, from what I understand, performed better than the Whittle engined prototypes.

The F.2 wasn't put into production because of reliability issues. I wonder if they were any greater than those being experienced by the Germans.

 

[Shortround6]

There is little difference between a centrifugal compressor used on a piston engine and one used in a jet engine except size. In fact at least one company (and perhaps a collage research group) was able to make working turbojets for drone missile use in the late 40s by sticking a combustion chamber (one of the tricky bits) between the compressor and turbine of a standard GE "B" series turbo charger. By hanging the single chamber combustion chamber outside the diameter of the compressor and turbine I believe they were able to keep the same spacing between the two units (keep same shaft and housings).
Granted you need a working turbo charger to pull this trick

Centrifugal compressors were good for about 3:1 pressure ratio at the beginning of the war and went to 4:1 or just a bit beyond at the end (for a single stage). It seems that axial compressors needed 6-8 stages (or more) to get a 3:1 pressure ratio in the early part of the war in practice despite what theory said. As the war went on and more research/development was done things got better (but slowly). Pressure ratios of 5:1 or better from axial flow engines only really came into use several years post war. By the early 50s however Pressure ratios of over 10:1 were being run (even if not in service) and by the 60s pressure ratios of 15:1 were being exceeded by a a few engines.

Some companies with marine/industrial steam turbine experience were able to build successful aircraft gas turbine engines, some were not. It seems that steam turbine experience, while useful, was no guarantee of a successful aircraft turbine engine.

The highest pressure ratio I have found for a single stage centrifugal compressor jet engine is 4.6:1 and that is in the early 50s. A late model P&W J-48 (modified licensed RR Tay) that offered 7,250lbs thrust dry at sea level and 8500lbs wet for 2200lbs without after burner for 2200lbs dry weight. This was probably the last gasp of the centrifugal compressor in a big engine although it continues in use in small engines to this day.

It took the rest of the world until the early/mid 50s to get the axial engine to show a clear superiority over the centrifugal engine.

Lets not get too hung up on test results and service engine powers either. Many engines were tested to much higher powers than they were used in service several years earlier. RR ran a Merlin at 1800hp in 1938 and ran it for over 10 hours at 1600 hp if I remember right. A P&W engineer ran an R-2800 at 3500hp for a least a few seconds using a truly astounding amount of water/alcohol and it was a "B" series engine to boot which were never rated by P&W at over 2000hp dry or rated by the Army at much over 2500hp wet. The higher powered engines were "C" series engines. BTW the engine survived
AN engine hitting 4000lbs thrust on test means exactly that, 4000lbs on test, on the ground, for a time of possible a few seconds or possible several hours, it does NOT mean it was installed and flown in an aircraft at that rating at that time.

 

[nuuumannn]

The thing about early British centrifugal flow engines is that Rolls-Royce in particular had been working with centrifugal compressors for years in superchargers, longer than its German counterparts, so there was a lot of accumulated experience, knowledge and data to fall back on when that firm began to build them. It was only natural that Rolls begin engineering these compressors because of what it already knew. Its first production axial flow compressor was riddled with problems, particularly with compressor surge.

(Vampire and P80 got them as experimental fits in mid/late 1945, with the P80 beeing badly damaged due to a Nene engine failure)

The P-80 prototype was powered by a Halford H.1, not a Nene.

 

[Shortround6]

ALL German WW II combat piston engines use centrifugal compressors (superchargers) and the Germans had been making such superchargers since the early 30s. Most superchargers of the late 20s and early 30s were violable on commercial engines and were NOT military secrets. Granted with 70-80 octane fuel the need or actual use of a high pressure ratio compressor is rather limited so some rather bad designs could see a fair amount of use with no one being the wiser.

By the way the P-80 prototype badly damaged the Halford H.1 engine not the other way around. The intake duct was made of too thin a gauge sheet metal and not rigid enough. On a ground run of the engine the intake duct duct collapsed and part of it was sucked into the engine. Blaming the engine for the damage doesn't seem quite fair.

 

[delcyros]

I like Anthony Kay´s books, and am aware that he refers to the I40 engine run at 4,000lbs on the benchtest. However, the production version manufactured in 1944 and through 1945 were derated to 3,850lbs static thrust.
The germans appearently never proceed with a jet engine project of 4,000lbs to 5,000lbs range.
The HeS006, JUMO-004 and BMW-003 all were 1750 to 2450lbs range, the He-S011 and Db-007 was 2450 to 3000lbs range and the JUMO012 and BMW-018 was in the 6,000lbs to 8,000lbs thrust range.


snippet from
Lockheed P-80A Shooting Star

The first 345 aircraft of this contract (serials 44-84992 to 44-85336) were designated P-80A-1-LO. Some of them were powered by the 3850 lb.s.t. General Electric J33-GE-11 turbojet, the production version of the I-40 which had powered the XP-80A and the YP-80A. Others were powered by the Allison J33-A-9, a version of the same engine built by the Allison Division of the General Motors Corporation.

The next 218 aircraft in the contract (44-85337 to 44-85941 and 45-8301 to 45-8262) were built as the P-80A-5-LO production block and differed by being equipped with the more powerful 4000 lb.s.t. Allison J33-A-17. The -5 also introduced a boundary layer control splitter plate inside the air intake. The landing light was relocated from the nose to the nosewheel landing gear strut. Later, the initial production P-80A-1-LOs were retrofitted with the uprated Allison engine during routine engine overhauls.

It might be noticed that the integration of the engine entailed thrust losses due to the long exhoust pipe and the intake ducts. Installed thrust was slightly lower than 3,850lbs rated for the first 345 P80A aircraft and has been calculated by NACA to the aequivalent of about 3,500lbs.

While the P80 prototype was indeed driven by an H1b, and the P80A prototypes were driven by I40/J33, one YP-80A was modified for an experimental fit with Nene engine as I mentioned:

serial no. 44-83027 was modified by Rolls-Royce to flight test the B-41, the prototype of the Nene turbojet. On November 14, 1945, it was destroyed in a crash landing after an engine failure.

So it was certainly possible to fit the Nene -at least theoretically- in an P80A airframe.

The HeS011A engine was initially buildt without reference to spare resources (the radial compressor milled out from a single block of aluminium took about 3,000 manhours to produce, roughly five times the total production time of a single series made JUMO-004B). It was quite complicated and offered only 1,115kp (=~2460lbs) at 9920rpm (HeS011 AV01, 1st prototype engine). That´s comparable to the performance offered by the experimental BMW-003D (which was lighter smaller)...
It was, of course hoped that thrust increase by an increase in rpm could be attained for 1,300kp in the - version and 1,500kp in the -C version but this is entirely speculation in absence of any engines made to these specifications.

 

[davparlr]

Yeah, there was something nagging at me on that 4000 lbs thrust for the YP-80. I just couldn't chase it down.

 

[nuuumannn]

On a ground run of the engine the intake duct duct collapsed and part of it was sucked into the engine. Blaming the engine for the damage doesn't seem quite fair.

Exactly. As for RR superchargers, despite the German engines having them in WW2 the knowledge Rolls accumulated did prove useful in the design of the compressors, but you knew that already, SR. The first RR engine with a supercharger was the Condor V of 1925 that had a exhaust driven turbo supercharger, although it did not go into production.

So it was certainly possible to fit the Nene -at least theoretically- in an P80A airframe.

I can't say I was aware of this. The problem with the Nene was that at this time it still needed work by the end of the war in Europe. Hawker did propose a Fury with a B.41 (Nene) jet engine with intakes in the wings and a split exhaust pipe in 1944 as the P.1035; this, via the P.1040 eventually evolved into the Sea Hawk.

 

[wuzak]

I have just read that there was another centrifugal jet project from Germany. It was from BMW.

BMW made turbochargers with centrifugal compressors and an air cooled axial flow turbine.

The BMW project took advantage of this knowledge, using an air-cooled turbine and a two stage centrifgal compressor. The axial flow turbine was rated for 900°C. This project, the P.3303, began in 1938 but was cancelled not long after the declaration of war (ie late 1939). It was decided to concentrate on the axial flow project, the P.3302 which would later become the 109-003.

The P.3302 project had its origins at Bramo before BMW took them over. It was originally designed for a combustion temperature of 600°C, but was soon changed to 900°C in co-operation with BMW, using their air-cooled turbine. Not long after, BMW took over Bramo.

Bramo had a second axial flow project, the P.3304. This differed in having counter rotating compressor stages. The theory there was there was no need for stator blades, and that the compressor would be more efficient (Griffith also experimented with counter rotating compressor stages when at the RAE, IIRC). The P.3304 would be officially known as the 109-002.

Bramo had ranked the types of compressors in order of desireability as 1. counter-rotating axial, 2. single rotating axial and 3. centrifugal. Bramo chose the single rotating axial compressor to start with, since it promised fewer development difficulties than the counter-rotating type. The latter began later (~1939), being cancelled early in 1942.

 

[delcyros]

BRAMO was very advanced before beeing bought out by BMW. There are believable reports suggesting that they had an experimental jet engine flying before ww2 in 1939 on a testbed.

 

[stona]

BRAMO was very advanced before beeing bought out by BMW. There are believable reports suggesting that they had an experimental jet engine flying before ww2 in 1939 on a testbed.

Reports from where?

I don't know much about jets but I'm familiar with the various machinations of the RLM in relation to some of these programs

Bruno Bruckmann and Hermann Oestrich at Bramo were visited by Hans Mauch of the RLM in 1938 (he also went to BMW, Daimler Benz and Junkers) to discuss gas turbines and jet propulsion. The Bramo men made it clear that they did not think that they could produce an efficient gas turbine and that they would continue trying to increase aircraft performance using a piston engine to drive a multi blade ducted fan. An experimental system was flown on an Fw 44 in October 1938. The Fw 44 was a two seat, biplane trainer for those unfamiliar with the type.
A rather more ambitious power unit consisting of a Bramo Twin Fanfir driving a large variable pitch ducted fan was also tested at DVL Gottingen in 1938.
Neither is an experimental jet flying before the war.

It was as a result of these projects giving disappointing results that the Bramo team started research and development of a turbojet. They collaborated fully with BMW from October 1938 until the acquisition of Bramo by BMW shortly thereafter, July 1939 I think, but haven't checked.

The BMW team at Munchen-Albach had already started a two stage centrifugal turbo jet and the Bramo team at Berlin-Spandau an 'axial counter rotating unit' [quote because I'm not sure exactly what that is!]by the time of the merger. This Bramo developed engine was placed under RLM contract 109-002.

The BMW 109-003 was an axial turbojet which used Loehner's turbine from the centrifugal engine which was dropped. This was the engine which finally ran (at Spandau) in August 1940, was eventually coaxed into giving 440 Kg of thrust and was cleared for flight in November 1941. Two of these engines flew in Me 262 V1 on 25th March 1942. They both flamed out. The developed 003A-0 didn't fly until 17th October 1943 under a Ju 88.

Cheers

Steve

 

[delcyros]

They experimented with not yet authochthonious (self sustained) jet propulsion aggregats, which lead to P3302 series types.
Ok, experimentation with jet engines was not so fundamentally new and it was still a long way to a working examples but nevertheless these approaches were very early.

 

[wuzak]

the Bramo team at Berlin-Spandau an 'axial counter rotating unit' [quote because I'm not sure exactly what that is!]by the time of the merger. This Bramo developed engine was placed under RLM contract 109-002.

In a conventional axial compressor each compressor stage is followed by a fixed stator.

In the axial counter-rotating unit the stators are replaced by another set of rotors which rotate in the opposite direction. There are, theoretically, benefits, but there is a big downside - complexity.

 

[fastmongrel]

Theoretically as Steam is a Gas everything from Parsons onward is a gas turbine. I naively thought that people who worked in Steam Turbine development would have been the best people to work on Gas Turbines but that doesnt seem to have been the case are the two disiplines not compatible.

 

[davparlr]

Theoretically as Steam is a Gas everything from Parsons onward is a gas turbine. I naively thought that people who worked in Steam Turbine development would have been the best people to work on Gas Turbines but that doesnt seem to have been the case are the two disiplines not compatible.

Not knowing anything about steam turbines, I would guess the aerodynamics of a steam turbine vs a jet turbine would be similar, however, vane cooling would be more of an issue with a jet turbine, and, I suspect gas velocity maybe a factor. I think there may be some similarity between a steam turbine and a compressor but not as much.

 

[delcyros]

Steam turbines didn´t used airfoil shapes for the rotors. So they were more simplier to be constructed but nevertheless there is some relationship between both technologies and some shared layouts. Needless to say that the Steam turbine wasn´t self-sustainable but relied on a feed steam manifold pressure provided by seperated boilers, which canceled the need to buildt a hot turbine.

 

[b0ned0me]

Steam turbines didn´t used airfoil shapes for the rotors.

I did originally post that this was what Griffith pointed out, but on further looking around, I think I'm wrong. His paper seems to be specifically on gas turbines, and focused on compressors.
Which points to the other major challenge for steam people/companies - they had no experience with compressors. I think the temperature and the compressor design are the two trickiest bits, which explains why areoengine firms with prodigious experience of high temperatures and compressor dynamics basically ate their lunch.

 

[Aozora]

Steam turbines didn´t used airfoil shapes for the rotors. So they were more simplier to be constructed but nevertheless there is some relationship between both technologies and some shared layouts. Needless to say that the Steam turbine wasn´t self-sustainable but relied on a feed steam manifold pressure provided by seperated boilers, which canceled the need to buildt a hot turbine.

Photos of Parsons ship's turbines; the blades still needed to be given a modified airfoil shape to direct the steam through successive sets of rotors, otherwise the turbines wouldn't work at all well:

Steam Turbines

 

[Shortround6]

The steam turbine guys have an advantage in that they know how to make disks with many replaceable blades. They Know how to and have the machinery to make the airfoil shaped blades.

A minor problem is that the materials they are used to working with don't work at the temperatures a gas turbine needs.

A bigger problem is that the compressor works the exact opposite of a Steam turbine, the steam turbine starts with a high pressure gas/fluid and allows it to expand in carefully controlled stages to get the most work out of it before the pressure drops too low to be useful.
The axial compressor of a jet engine (or gas turbine) starts with a low pressure gas/fluid and compresses it in carefully controlled stages to get the highest possible pressure while using the least amount of power.

Many steam turbine firms were also used to thinking in terms of HP per ton instead of HP per pound (or KG). Of course Steam turbine firms were also used to thinking of overhaul life in thousands of hours not tens of hours (or a few hundred)

 

[wuzak]

Photos of Parsons ship's turbines; the blades still needed to be given a modified airfoil shape to direct the steam through successive sets of rotors, otherwise the turbines wouldn't work at all well:

Steam Turbines

For turbines there are two kinds of blades - reaction and impulse. I think that the reaction blades are aerodynamically shaped. Less so in the case of the impulse blade.

From the looks of the first picture, the five rotors to the right at impulse blades and the rest are reaction.

That image shows how importantthe stators and nozzles are as well. These aren't shown in the rotor set pictures.

 

[delcyros]

I think the latter picture quite elobately shows this. However, even then, the early gas turbines were not reaction turbines but impulse types. Müllers HeS030 was the first axial Reaction type gas turbine to my knowledge. This yielded significant improvements in efficiency of the engine.

Regarding the late ww2 german jet engine developments, I´d like to give some further thought on BMW projects.
BMW developed the P3302/3 into the BMW-003. By late 1940 they started another, initially preliminary project which would eventually end up in the large BMW-018 engine, of which at least two prototypes have been buildt, though evidence suggest that only the compressor has been trialed before the facilities were captured and the BMW-018V2 was destroyed (the -V1 was destroyed in oct. 1944 by an air raid).
However, Schelp went into the class II buissness with the HeS011 triggering the project in 1942. The RLM was not convinced that the diagonal compressor is necessarely the most efficient form of a class II jet engine and urged Schelp to order BMW to exploits it´s experiences in the more developed BMW-003C/D projects to scale the axial engine up into a suitable class II engine as a backup.
This was the late 1943 or early 1944 starting point of the BMW-P3306 projected jet engine. At wars end, the design of the BMW-P3306 V1 had been completed, though no parts have been assembled. The outlines of the projected engine read like:

diameter: 850mm
length: 3200mm
compression ratio: 4.0:1 (calculated)
thrust: 1700kp at 7500 rpm (calculated)
sfc: 1.15 (claculated)
7 stages axial compressor (reaction type)
1 stage axial turbine
annual compressor with 20 burners

Altough not run in ww2, one engine was made to this specification and trialed in march 1948. The head of the BMW jet engine project was hired by the french to set up a jet engine design bureau in sept. 1945. Just 3 weeks later, he presented to complete design of an engine "scaled up from the known BMW-003 and able to produce 1700kp". As You can imagine, he didn´t created this engine in three weeks in 1945, but what he was able to do is to recollect the complete set of ww2 BMW-P3306 drawings. He never referred to P3306 but I have a couple of drawings and it´s nothing else, it´s the realisation of the BMW ww2 jet engine.
The design was submitted and after the german design team signed a second contract in 1946, S.N.E.C.M.A. started preperation for part assembly of two jet engines, redesignated ATAR-101. The -V1 prototype was still a perfect BMW-P3306 with Riedel starter unit, hollow air cooled turbine blades from rather mild steel and adjustable exhoust area controlled by a jet needle. In the second prototype -V2 the mild steel was replaced by high Nickel and chromium content steels, resulting in simplier construction and increased rpm and allowed exhoust temperature.
Delays in S.N.E.C.M.A production (particularely in producing hollow turbine blades) made compressor parts aviable in late 1946 and in march 1948 the -V1 prototype was benchtested for the first time. It worked according to specification and developed 1700 kp thrust at 7500rpm as intended.
Just two months later, the-V2 became avaiable and produced 2,200kp thrust at 8,050 rpm. The following development of the ATAR-101A and -B (they removed the Riedel starter engine in the hub and changed the tail to exhoust eye-type lid shutters) is entirely post ww2.



In my mind, a much more reasonable engeneering approach than the HeS011. Note how much the inavaiability of proper heat resistent materials affected the performance of this engine.