The Hispano-Suiza HS.404 20 mm Aircraft Gun in US Service
George M. Chinn, edited by Anthony G Williams
Amended 3 April 2009
[Editor's note: a recurring subject of debate is why the USAAF and USN relied almost exclusively on the .50 Browning MG as an aircraft gun during the Second World War, and did not make more use of the 20 mm Hispano cannon which was in production in the USA during this period. Some people assume from this that the .50 Browning was inherently superior to any and all aircraft cannon of 20+mm, despite the fact that all other major powers engaged in the war had adopted 20 mm cannon as the principal fighter weapon long before the end of the war, with some moving up to even larger calibres. The answer that the explanation lay in production quality problems with the American Hispano is often greeted with incredulity; after all, wasn't American industrial production the best in the world?
The history of the acquisition and development of the American Hispano is dealt with at great length in the classic work, "The Machine Gun", written by George M. Chinn, a retired USMC Colonel, in the early 1950s. What follows is two extracts from this account, which between them explain (almost) what happened, and why.]
Modifications and Attempts at Standardization
By the time production really began with all companies, the United States Navy was the only branch of the service actually placing Hispano Suiza cannon in planes in great numbers and demanding that aircraft designs of the future include 20-mm in lieu of lighter machine gun armament. Unfortunately, the gun, which was being delivered in prodigious numbers, was not proving itself totally reliable in America.
On the other hand, promising reports were coming in daily from the British, who had used practically the same procedure as our Army ordnance engineers in getting the gun into production status. The original Birkigt Type 404 gun was used as the model for the Mark I. This was followed shortly by the Mark II. Drawings and a sample of this improved cannon arrived in the United States for purposes of study and test in January 1942. The British strongly suggested that American ordnance officials confer with their representatives in order to accomplish an early standardization. It was further desired that all 20-mm aircraft cannon of this design procured for British use be of the Mark II type, the principal differences between the Mark II and the American M1 being pointed out as follows:
"1. The magazine carrier of the Mark II gun had a different latch; the ejector was provided with a buffer, and changes had been made in the magazine holding boss.
2. A heavier rear buffer was provided in the Mark II gun and the back plate was dovetailed into the receiver instead of being fitted with a simple groove as in the M1 gun. Inertia blocks were used in the breechblock slides of the Mark II gun.
3. The sear of the Mark II gun had been modified.
4. Triple wire driving spring and extractor springs were used in the Mark II gun instead of simple single-strand coil springs which were used in the M1 gun.
5. Minor changes had been made in the muzzle brake of the Mark II gun.
6. The receiver of the Mark II gun was substantially different from the one used in the M1 gun; much heavier guide rails were used and the receiver itself was larger, heavier, and designed in accordance with British manufacturing methods.
7. The chamber of the Mark II gun was 2 mm shallower than the chamber of the M1 gun."
Practically all the changes suggested were of a minor nature and slight modifications or alterations would permit complete standardization. The main difference in the two types of gun was in the chamber dimensions. Since both were designed to use the same cartridge, it was quite obvious that one size would best handle the round. The British were very insistent that their measurements were better, pointing out, in particular, that their chamber was slightly more than one-sixteenth inch shorter than the American one. In their opinion, such a length would solve the problem of faint strikes, since the weapon was inertia fired and depended upon the shoulder of the chamber to offer resistance and position the cartridge.
Tests were conducted at Aberdeen, Eglin Field, Wright Field, and Kenvil Proving Grounds to determine the relative merits of the British suggestions. The Army Ordnance engineers were not convinced by these tests that the British chamber was superior to the American design. However, it was agreed on 4 April 1942 that additional trials be initiated for the purpose of reaching a satisfactory compromise for both governments.
The only official action finally taken by the American representatives and approved by the Ordnance Committee was:
"(I) That the manufacture of the American 20-mm automatic gun Ml and AN-M2 be continued in the United States without modification to the chamber.
(2) That no chamber with the small cone moved one-sixteenth inch to the rear (as was done in the British chamber) be considered for manufacture in the United States.
(3) That the request made by British representatives that 20-mm automatic guns produced in the United States for British use be made with British chambers not be considered until after the 20-mm automatic guns Ml and M2 have been subjected to a thorough test in Great Britain."
After further comparative tests in late April 1942, it was again definitely decided by the Ordnance Department that all American-made 20-mm automatic guns continue to be made with the chambers longer by one-sixteenth inch than the British regardless of the employment of the same ammunition. This decision was final as far as American production was concerned, but in no way did it change the British representative's view on the longer chamber's performance.
[In 'Guns of the Royal Air Force 1939-1945' by G.F. Wallace - who was there - there is an account of British tests of the British and US Hispanos which took place early in 1942. The British were unhappy with initial supplies of the American-made guns: "there were frequent misfeeds and lightly struck cap stoppages, and the life of several small components was very short" so a comparative test between one British and three American guns was set up. The intention was to fire 5,000 rounds from each gun without replacing any components. "The British gun fired the full programme but the performance of the American guns was so bad that in each case the trial had to be abandoned before the 5,000 rounds had been fired." The British gun experienced 19 stoppages in firing 5,012 rounds. The American guns experienced 67 stoppages out of 4,092, 97 out of 3,705 and 94 out of 2,610 respectively. Incidentally, Wallace states that the US guns were "beautifully made and better finished than our own" and expressed surprise that although lightly struck caps were a major source of stoppages, even more frequent were mis-feeds.]
Oddly enough, the question was again raised, not by the English or our many proving grounds, but by manufacturers of 20-mm ammunition. In testing their cartridges for reliability of action, they encountered a series of malfunctions known as light-struck primers that were all out of proportion for such a weapon. These were not isolated cases, the reports coming in from practically every maker of 20-mm ammunition that was engaged in function firing his products.
Since the munitions companies pointed out that the faint strikes were due to lack of impact on the primer resulting from error in the gun, and not as a result of defective materials or workmanship, it was decided to conduct another test on an extensive scale at Aberdeen. Ninety of the 20-mm guns, Ml and AN-M2, selected from every facility producing them, were expended in this test with all types of ammunition, both from accepted and rejected lots.
A complete record was made of every malfunction during the entire test and the probable causes of the trouble. The engineers in charge of the project in the early stages of this test recommended that two modifications should be made to overcome the serious malfunctions:
"(1) Shorten the chamber one-sixteenth inch, thus modifying it to approximately the British chamber.
(2) Replace the extractor spring with a solid plug, thus positioning the rounds by means of the extractor. This change would include such modifications to the extractor, the bolt, and the ejector, as were deemed necessary."
The test began in June 1942 and continued until the last of January 1943. The final recommendations from Aberdeen Proving Ground were presented at a meeting attended by representatives of the Ammunition Branch, Industrial Division, Artillery Branch, Technical Division, and Field Service. All present accepted as official four much-needed modifications that were to be made on all 20-mm M1 and AN-M2 cannon: (1) The chambers were to be shortened one millimeter or approximately one thirty-second inch; (2) the extractor spring would be of the cantilever type; (3) the standard firing pin was to have one sixteenth inch removed from the back of the key slot to give it "float"; and (4) the breechblock slide springs would be strengthened.
This sanctioned change found the Army with 40,000,000 rounds of ammunition already stocked. While 56,410 guns had been manufactured to date, it would be easy to make external changes such as with the firing pin and extractor spring. Barrel chamber shortening, however, was a problem that generally cost as much in time and money as to make the whole barrel, to say nothing of the number of guns immobilized while the modification process was being performed.
Action was taken immediately by the Industrial Division to put the alterations into effect. There still remained certain differences between the British Mark II and the AN-M2. As the British Ministry of Aircraft Production had long advocated having both guns manufactured identically, the Army Ordnance Department ordered a comprehensive test in England as soon as the modified weapons came off the assembly line.
At the suggestion of Capt. E. R. S. Adams of the British Air Mission, two guns each from International Harvester, Oldsmobile, and Bendix were shipped to England for the purpose of competitive aerial tests with the Mark II. Representatives of the Army Ordnance Department were present to observe the 2,000-round tests which were held during July and August 1943.
Two British Mark II's were mounted in the left wing of a Hurricane fighter with two AN-M2's made by Oldsmobile and International Harvester in the right wing. Combat flying, dives, G-loading, straight-away, etc., were simulated. One stoppage was attributed to the Oldsmobile gun. The International Harvester weapon had no stoppages but a cracked breech-block was noticed at the completion of the trial. Each Mark II had one sear failure and one of them had a cracked breech lock after 1,400 rounds. The Bendix guns were fired on the ground in competition with the British-made guns and made a creditable showing.
In reporting the findings of the test, Mr. Hansen, of the British Ministry of Aircraft Production, declared: "American guns are as good as British guns and are acceptable for service use".
[There is an inconsistency here. As will be seen below, the performance of the American Hispanos remained unsatisfactory throughout the war, yet the British expressed no concerns and were obviously happy with their guns. There would seem to be only two possible explanations; either the British tolerance of unreliability was considerably greater than the American, or the American guns had been assembled from selected components and thereby performed better than average.
Wallace states that although thousands of American-made Hispanos were supplied under Lease-Lend to the UK, none was ever installed in RAF aircraft. Some were modified for use as AA guns by shortening the chamber and fitting triple-wire recoil springs, but these non-standard guns were never used in operations.
However, the news from America wasn't all bad. Wallace states that: "we had continual trouble with the recoil units of the British Hispano 20mm gun - the Mark V gun in particular never worked satisfactorily with the British designed unit. The American Edgewater Mounting unit...was tried on the Mark V gun and was found to be a great improvement...In consequence all Mark V Hispanos were fitted with the Edgewater design of Front Mounting Unit."]
Performance of Hispano-Suiza Cannon During World War II
It soon became apparent that the Navy would be the largest user of the 20-mm cannon; in fact, records show that this branch of the service mounted over 90 percent of the cannon actually placed in American aircraft. The first 20-mm Hispano-Suiza automatic gun in a mock up by the Navy was installed at the Bureau of Aeronautics test facility, then known as the Aircraft Armament Unit, Norfolk, Va., on 11 March 1942.
There was nothing slow about the Bureau of Aeronautics on armament decisions, for the gun was officially accepted for aircraft use the next day. The installation was in the wing of an SB-2C which had been shipped separately by the Curtiss-Wright Aeronautical Corporation, Columbus, Ohio, in order to proof fire the new cannon.
The Aircraft Armament Unit continued to test the weapon in its mock-up mounting and it was as late as August 1943 before the guns were actually placed in the planes at the factory, SB-1C, serial number 200, being the first to be so armed. It was quickly followed by plane number 50, SB-2C.
This aircraft, the scout bomber type, proved to be one of the most widely used during the war and the resulting damage inflicted upon the enemy by Naval pilots was tremendous. However, the records reveal such destruction was done by dive bombing rather than by the use of the 20-mm cannon. The SB-2C first reached the combat area on 11 November 1943 and its first action took place in March 1944.
As an innovation in Navy ordnance, factory representatives accompanied the new cannon to the front. These expert technicians sent back voluminous reports that explained that the malfunctions that did occur were due to one of three things: failure of the feeder, bad ammunition, and improper maintenance. Their zeal in clearing the gun itself in every instance casts doubt on the validity of the reports.
Available records show that the AN-M2 was installed in 5,800 Naval planes requiring the total mounting of 11,600 guns. The SB-2C and SB-W aircraft were the principal planes carrying this weapon into combat, along with a very limited number of F4U-1Cs. Statistics on enemy aircraft shot down in World War II credit the AN-M2 in SB-2C aircraft with destroying few enemy aircraft. The F4U-1C planes brought down an even smaller number. However, it must be remembered that the primary mission of the SB-2C was not to shoot down aircraft.
[The utilisation figures given in this section cannot be correct, at least as far as WW2 production is concerned. Some 7,200 SB-2C/W were built, using a total of 14,400 cannon. In addition, over 1,400 F6F-5N night fighters, which each carried two cannon, were made. Finally some 200 F4U-1C were equipped with four cannon. This gives a total of around 18,000 guns used by the USN.
The figure of 90+% of production to the USN also cannot be right. The USAAF principally used the 20mm Hispano in two aircraft: the P-38 (over 9,000 built, carrying one cannon each) and the P-61 (650 built, with four cannon), giving a total of nearly 12,000 guns. Nearly 30,000 guns were therefore fitted to aircraft, of which the USN made use of just over 60%.
This use breaks down (in round figures) as follows: USN attack planes 48%, USN night fighters 10%, USN day fighters 3%, USAAF night fighters 9%, USAAF day fighters 30%. ]
The United States Navy has always permitted the introduction of evidence even when contrary to what it would like it to be. There existed two distinct schools of thought on the reliability of the gun. One was that the 20-mm Hispano-Suiza automatic cannon could not be considered satisfactory as an aircraft weapon as long as it was necessary for the ordnanceman to coat the cartridge case with a heavy lubricant or wax. The other was that this was unimportant as long as it bettered the performance of the weapon. But everyone even remotely connected with weapon development agreed on one thing, namely, that 20 millimeters was the minimum caliber for aerial warfare.
[There is another puzzle here. The RAF were totally opposed to having to wax or oil cartridges, so according to a primary British source the cartridge was modified so that such oiling or waxing was unnecessary (unfortunately, the source doesn't say what changes were made). This is corroborated by the official RAF Hispano manual, which explicitly forbits oiling the ammunition. Yet all US Hispanos relied on oiled or waxed ammunition. So how could the British and American guns and ammunition have been compatible, as was supposed to be the case? The most likely explanation is that the British did indeed settle for a hard wax coating, applied at the factory.]
During war all that can be done is to install and make function as reliably as possible that which is issued. With the mounting of the 20-mm cannon in Navy planes a series of malfunctions began that could not be properly corrected at the time because manufacture was at the peak of production. The slightest change would practically mean retooling. The most serious problem was the oversize chamber. There still remained considerable variance in dimensions between the chambers of the British and American cannon, even after the latter chamber was made one thirty-second inch shorter.
Due to an outmoded agreement of long standing, everything above caliber .60 in the Army is considered artillery and the manufacture of the Hispano-Suiza cannon therefore came under this classification. In other words the production of this high-speed machine gun was done under artillery manufacturing tolerances. The resulting poor mating of parts, coupled with the inherent fault of all gas-operated weapons whereby the breech locking key in the receiver is immovable and the position of the gas port in the barrel is permanently fixed, made it impossible to adjust the relationship between barrel and breech lock to establish head space. Thus the most vital measurement in any automatic weapon was governed by chance in this instance.
An unfortunate discovery was that chamber errors in the gun could be corrected for the moment by covering the ammunition case with a heavy lubricant. If the chamber was oversize, it served as a fluid fit to make up the deficiency and, if unsafe head space existed that would result in case rupture if ammunition was fired dry, then the lubricant allowed the cartridge case to slip back at the start of pressure build up, to take up the slack between the breech lock and the breech lock key. Had this method of 'quick fix" not been possible, the Navy would have long ago recognized the seriousness of the situation. In fact, this inexcusable method of correction was in use so long that it was becoming accepted as a satisfactory solution of a necessary nuisance.
This state of affairs continued until the war's end, at which time all complaints and suggested improvements were carefully evaluated by the Chief of the Bureau of Ordnance, and a letter outlining past faults and suggesting improvements was dispatched on 26 December 1945 from that authority to the Army's Chief of Ordnance under the subject, "Reactivation of Certain 20-mm Automatic Gun Development Projects and 20-mm Ammunition Development Projects-Request for." The following paragraphs are quoted from the letter:
"There is a firm requirement on the part of the Navy Department for use of 20-mm automatic guns in practically all Navy combat aircraft currently in design and currently designated combat operational aircraft -
The 20-mm automatic guns M2 and M3 in their present stage of development have certain objections and defects which make continued development of this type weapon highly desirable. The following features are considered objectionable and are believed capable of improvement:
(a) The profile of the gun is too bulky for proper installation in VF type wings.
(b) The cyclic rate of the gun is too low.
(c) The belt pull is too low.
(d) It is believed that the over-all weight of the gun and its associated equipments can be materially reduced.
(e) The accumulated tolerances in the manufacture of the weapon are too great to give uniformly efficient operation in these guns.
Other objectionable features which are believed capable of rectification are listed below:
(a) The need to oil the ammunition prior to loading for use in this weapon is undesirable. Self-lubricated ammunition, or the elimination of the need for lubrication, is strongly desired.
(b) The ballistics of the projectile can stand much improvement. It is believed that ballistics similar to that of the Caliber .60 projectile can be closely approximated.
(c) It is believed that an electrically primed round can be developed for the gun which will give more efficient performance.
(d) The ammunition should be manufactured to fit the chamber of the gun in which it will be fired and not to fit two of these weapons - namely, the American and British 20-mm automatic guns.
From the above, it can be seen that the Navy's need for improvement in the gun and ammunition is immediate and will be continuing until the Army's long range development program of an optimum gun for aircraft materializes. It is understood that the optimum gun will require from 15 to 25 years for development to be completed. Continued improvement in the present cannon will certainly contribute materially in experience gained to the development work leading toward the optimum gun....
Inasmuch as the Chief of Ordnance is definitely interested in this development program, this Bureau wishes to indicate its active interest in and requests that the following program be undertaken:
(a) Improve the present 20-mm Automatic Gun M3 for immediate needs.
(b) Continue development projects of such guns as the 20-mm T32 and T33 to arrive at a reliable lightweight, high performance gun within the next four to six years.
(c) Improve the ammunition for these guns in order to achieve a family of matched projectiles of relatively high performance.
(d) Through the experience gained in this development program obtain information, data and experience which, combined with current gun research for an optimum gun, might materially aid in the development of an optimum gun for aircraft within the next ten to fifteen years.
To support such a program, this Bureau will initiate projects complementary to those undertaken by the Ordnance Department (ASF) to provide competent and experienced personnel and afford Navy Ordnance facilities to assist in the program. In addition, this Bureau will furnish funds to support a proportional part of this development program as established by the estimates of the Chief of Ordnance."
This letter resulted in the cooperation of the Army, with Navy engineering personnel, familiar with the conditions that needed remedying, in solving the various problems. Today, barely five years after the war, every point brought out by the Navy's Chief of the Bureau of Ordnance has been answered. Nothing was basically wrong with the weapon. Its wartime performance, good or bad, was the result of having been bought in desperation, put into mass production without first having been adequately proved, and then modified regularly to meet a future commitment before the previous model had been made to function reliably.
[Chinn goes on to describe the postwar development of the American Hispano, implying that all of the problems were solved within a few years in the much-modified M3 and M24 variants. However, a separate source mentions that when the French received a batch of F8F-1B Bearcat fighters for use in SE Asia in the early 1950s, they found that a high proportion (25 out of 64) of the installed 20 mm M3 cannon were defective, due to parts being outside the specified margins, and had to be replaced].