Speed development of Piston-engined Military Aircraft

   Within 10 years of the Wright brothers first flight armies saw the value of aircraft for reconnaissance.  A pilot could range further, see more and report more quickly than a cavalry scout.  A particular flight soon after the outbreak of the Great War made a fundamental contribution to the outcome.  German armies were invading France and, in accordance with their grand strategy, the Western-most army was aiming to march West of Paris as part of a great encirclement of the French armies retreating before them.  On the 31^st August 1914 a reconnaissance pilot reported back from an observation of enemy troops just 47 miles NNE of Paris that they had changed direction to pass East of the capital (Ref. 1).  The French then saw that they could attack into the German flank by their troops assembling near Paris.  Given the warning they had time to organize that.  They did it (the battle of the Marne), it caused the Germans to recoil and thereby it foiled their grand strategy, which might otherwise have led to a defeat of the French which would have cost them the war.

Fighter development

   It was soon seen by both sides that aerial reconnaissance must be prevented and special aircraft were developed to do this.  The British called them “Scouts” originally (later they were named “Fighters”).  After various other ways of fighting were tried (eg forward-firing machine-guns (MG) in a pusher-propeller aircraft) the method of attack settled down to fuselage-mounted MG firing between the propeller blades of a tractor-propeller aircraft, using interrupter gear as pioneered by the Fokker E1 in 1915.  They were aimed by manoeuvring the whole aircraft.  Much later, after metal-skinned monoplanes replaced fabric biplane wings, reliable MGs could be fitted in the wings.

Fighter speed

   Fighte-to-fighter combat followed naturally.  The most valuable characteristic of a fighter was quickly seen to be speed.  The development of that quality is the subject of this Contribution 25.  The scope is confined to Piston-engined aircraft up to the end of WW2 in 1945, with just a note on the coming of jet power in the last two years in APPENDIX 2 on P.14.

The broad picture, 1914 to 1945

   Fig.1 below gives the development picture over 1914 to 1945.

Page 2 of 14

   The Fokker E1 was actually a monoplane but dog-fighting loads led to the biplane becoming the norm in the Great War.

   Biplane Fighters rose in speed from the Bristol Scout of 1915 at 93 MPH to the FIAT CR42 at 261 MPH in 1939 (+180% in 24 years but 21 of those at peacetime rates of spend).  Monoplane Fighters rose from the Seversky P35 in 1937 at 281 MPH to the North American P51H Mustang at 486 MPH in 1945 (+73% in 8 years).

   Bombers have been included in the WW2 period to show the margins available for Fighter interception.  All those plotted incurred the weight and drag of defensive armament, in accordance with orthodox policy*¹, bar one – the un-armed de Havilland Mosquito.  The great speed advantage as a bomber is clear*².  Before it flew it was called by some “Freeman’s folly” after the Air Chief Marshal who supported it.

*1.  The British and US bombers in particular carried heavy powered turrets.

*2.  The speed of the Mosquito, as well as enabling it to avoid interception, also meant that it could perform two sorties to Berlin in one long winter night of 1944, the second two-man crew also dropping a 4,000 lb bomb (Ref. 4).  Four men in a 2-engined Mosquito therefore equalled the 8,000 lb bombload of a 7-man 4-engined Lancaster heavy bomber.  The loss rate of the Mosquito over the winter 1943/44 was 0.5% where that of the Lancaster was 5% (Ref. 5, linked).

Correlation of aircraft speed with physical factors

   The physical factors readily available in Refs. 2 and 3 were:-

Wing Span (WS); Wing Area (WA); 

Average Weight (MA), which is ½(Weight Empty + Weight Loaded);

Total Power (PPT) @ Altitude (AA) of Maximum Speed (VM).  Certain assumptions about this are given in APPENDIX 1 on P.11.

 An omission which could not be made good without excessive work was Frontal Area.

Air Density @ Altitude (1000r) is taken from standard data.

These factors were used in a Multi-variable Regression Analysis (MRA), using Log factors, to provide correlations in the following form:-

VM  proportional to (WS^m1 x WA^m2 x MA^m3 x PPT^m4 x 1000p^m5 ).

   The values of the exponents m1 to m5 are shown on the charts for the following groups:-

Fighter Biplanes;   Fighter Monoplanes;   and WW2 Bombers.

Fighter Biplanes with fixed undercarriages and open cockpits (except 1937 Gladiator had a canopy).

Fig.2 below presents this correlation.

Page 3 of 14

   The immediate question is “Why is the exponent for MA not negative?”.  It is contrary to common sense and physics that speed should increase with Weight!  Nevertheless, the computer says that it does!.  MA must be a surrogate for a parameter not yet identified.

   For 24 examples the average difference from the trend line is + /- 5.5%, reasonable for such a wide variety of machines over a period of 24 years.

Biplane Fighter Illustrations

Bristol Scout 1915

   The Scout and Camel had air-cooled rotary engines.  The DVII had a water-cooled 6 cylinder in-line engine.  The Gladiator and CR42 had air-cooled radial engines.

   The Gladiator was the first RAF fighter to have an enclosed cockpit.  In a novel feature the traditional pair of Vickers fuselage-mounted belt-fed MGs were supplemented by a Lewis drum-fed pair mounted under the thin wings, firing outside the propeller arc.

Page 4 of 14

Fighter Monoplanes with retractable undercarriages and enclosed cockpits.

Fig.3 below presents this correlation

   The anomaly of speed shown to increase with MA persists in the MRA of fighter monoplane data.

   The scatter of all 67 examples is +/- 4.3%.

Selected sub-groups of Fighter Monoplanes, compared to overall trendline

Fig. 4

   Apart from any doubt about the physical factors the FW TA152 with a bluff nose radiator may have suffered from local compressibility drag.

Page 5 of 14

Fig. 5

Fig.6

Page 6 of 14

Fighter Illustrations

*3.  After surprising a few unwary Germans in mid-1940, the Defiant was soon relegated to night operations.  The guns could be fired forward above the prop.arc (as used against Zeppelins in 1917 and a precursor of the German Schrage musik).  Without airborne radar in early 1941 the chance of a belly night attack success must have been very small.

Page 7 of 14

Both the Mustang (designed in 1940) and the Tempest (1941) had “Laminar flow” aerofoil wings.   “Laminar Flow” is placed in inverted commas because NACA, who pioneered this wing section in 1938, admitted that, while a hand-finished wind-tunnel test piece could achieve that type of flow over the first 3^rd of the chord with its reduced drag, an ordinary production wing could not be made sufficiently accurately to obtain much laminar flow in service.  The Mustang’s superior performance shown on Figs 1 and 4. was almost certainly due to having an under-body sufficiently-long and efficient radiator cooling-air inlet duct to obtain the true “Meredith effect” of zero drag from the system.  The huge radiator on the Tempest, needed for the Sabre engine, chin-mounted with a short inlet duct, is a stark contrast.

Page 8 of 14

A tale of two undercarriages!

The Martlet must have been tricky to land on a rolling carrier deck.  The wheel-track was greatly widened for the Hellcat and Bearcat.      The Bearcat did not reach WW2 service                    

.

Page 9 of 14

Bombers of WW2.

Fig. 7 below presents this correlation

   Discussion

   Considering the necessarily crudity of the assumptions with the available data, the correlation is quite good.  The average difference from the trendline for the 18 examples is +/-4%.  It covers –

                                           MA  Lb                PPT  BHP             VM  MPH              Difference from trend

from  Fairey Battle:             9,248                   1,030                  233                                  0.48%;

   to  Convair B36B           246,000              21,000                   372                                 0.26%.

                                                                        (illustrated below)

   There are examples (illustrated on P.10) where known features with an effect on speed are shown in the right direction of the difference from trend, although the size of the difference could not be relied upon, egs:-

• B17G:  difference  -4.2%:-  raised cockpit + radio & engineering stations, chin & top turrets (it is presumed that the ball ventral turret was retracted and fuselage gun hatches were closed for the max. speed test, but this may not have been done to get a combat measurement: 
• B29A:  difference  +7.4%:-  cockpit enclosed within glazed nose, low-line remotely-aimed turrets:  .  (The Heinkel He111 pioneered the cockpit-in-the-nose:  111H6 difference  +3.5%);
• Mosquito B IX:  difference  +6%:-  most-streamlined type in the examples, as noted above, without the typical British powered turrets, with in-line liquid-cooled engines and radiators embedded in the wings: 

Continued with illustrations on P.10.

Page 10 of 14

   The Whitley when designed originally without wing flaps was given the high value of 8.5⁰  incidence to give low take-off and landing speeds.  In level flight this produced a significant nose-down attitude.

   The special features of the B17G, B29A and the Mosquito have been described above.

Conclusion

   If they do nothing else these correlations do provide a useful way of illustrating the relative speed merits of numerous and varied aircraft.

Page 11 of 14

References

1. The World Crisis, 1914-1918  W. Churchill  Odhams ed  1949.
2. The Fighter Aircraft Pocket Book  R. Cross  Batsford  1962.
3. The Bomber Aircraft Pocket Book  R. Cross  Batsford  1964.
4. Development of the Rolls-Royce Merlin from 1939 to 1945  A.C. Lovesey  Aircraft Engineering magazine  1945.
5. https://military-history.fandom.com/wiki/Avro_Lancaster

APPENDIX 1

   The data from Refs. 2 and 3, although extremely comprehensive for such modest publications, needed three assumptions to make the correlations:-

1. That with supercharged piston engines the Total Power (PPT) corresponding to the Maximum Speed (VM) at the given Altitude (AA) was the same as reported at sea-level;
2. That the weight of the aircraft when tested for VM was the average (MA) of the reported Empty Weight (WE) and Loaded Weight (WL).

   No.1 is reasonable.  Most supercharged engines at high altitude used 2-speed compressors which reduced the 2^nd -speed power by the increased gear losses, but this was probably the same %age for all and so has been neglected.

   No.2 is rather crude.

  Frontal Area is also data that is needed ideally but is not available in the sources. 

   The standard variation of Density with Altitude was correlated as:-

1000p  =(0.0019*(AA)^2 -(0.2158*(AA)) +7.6337)     1000 lb/cu ft

3. To cover the whole range from 1914 to 1946 a calculation was needed for the power at Max. Speed altitude of Naturally-Aspirated (NA) engines in the years 1914 to 1926.  This used the standard adjustment for density variation:-

The method did give a seamless connection between NA engines and the later Pressure-Charged (PC) engines.

Comparison of the MRA exponents

   The MRA exponents for the 3 aircraft group correlations compare as follows:-

   

                                                                                                                                                                                                                

Page 12 of 14

Mixed group correlation

   As an experiment a mixed group of Bomber, Biplane and Monoplane aircraft was put into an MRA to provide suitable exponents for ASCFX as follows:-

                                                                                          

   The resultant chart, using competing Maritime WW2 aircraft, is shown below:-

                                                                                                                                                                                                                                       Fig.8

The flying boats were added separately.  Clearly, the Sunderland shows the need for a frontal area variable.  The Catalina had a much smaller hull, as seen in the illustration below.  It also retracted the floats to the wing-tips, which possibly reduced induced drag.

                                                          Partners in the Battle of the Atlantic.

Dive-bombers deadly against ships

See Contribution No. 10 for the damage inflicted by Luftwaffe dive-bombers on the RN and Allied navies before the Battle of Britain.  Over April and May 1941, during evacuations of the British and Commonwealth troops from Greece and then Crete, German dive-bombers had their greatest successes; with no air cover for the ships they sank 3 cruisers and 8 destroyers.

Page 14 of 14

  The Me262, Meteor and Airacomet were twin-engined.

The Me262 had the awesome armament of 4 x 30 mm cannon grouped in the nose.

The Mustang, in contrast, had 6 x 0.5’’ MG.  The Meteor had 4 x 20 mm cannon in the nose.