The Norden Bombsight and the U.S. Naval Proving Ground
Robert V. Gates
The Norden Bombsight and the Army Air Force are forever linked in public memory. Less well remembered, however, is the fact that the bombsight was developed by Carl Norden for the U.S. Navy.
Prior to World War I, military theorists thought of aircraft, when they thought of them at all, in scout or reconnaissance roles. More thought was given to their utility as bombers during the war. Bombing missions were attempted from low altitudes by aircraft carrying only small bomb loads. Bombs were dropped by the pilot or observer without the benefit of an aiming device, or bombsight. Needless to say, bombing was hit or miss—mostly miss.
Later in the war, more attention was given to developing aircraft capable of carrying larger bomb loads. The British began experimenting with bombsights in 1916. The most promising, developed by Lt. Cdr. Harry E. Wimperis of the Royal Naval Service’s Imperial College of Science, was described as “little more than a board fitted with a bubble level and two adjustable rifle sights.” Predetermined bombing tables and levers adjusting for altitude and speed were used to achieve an accuracy of “hundreds of feet.” A primary source of inaccuracy was the aircraft’s random pitch and roll during bombing runs.
The U.S. Navy considered ships the primary targets of its bombardment missions. In its search for an effective means of accomplishing this mission, it considered level bombing, dive bombing, glide bombing, and aerial torpedo attack. In June and July of 1921, Army airmen and Navy aviators dropped bombs on a variety of targets including the anchored German battleship Ostfriesland. Navy pilots dropped bombs from an altitude of less than 2500 feet and achieved hits with only 19 percent of them. Brig. Gen. Billy Mitchell’s Army pilots did somewhat better: 30 percent of their bombs hit the Ostfriesland.
| Bombing Ostfriesland, c. 1921.
Photo courtesy of the Naval History
A 1924 report for the Secretary of the Navy concluded “it is absurd to think that either the aerial bomb or the submarine torpedo have furnished the effectual answer to the capital ship.” The Army came to the same conclusion.
The Navy’s Bureau of Ordnance (BuOrd) had the responsibility for developing Navy bombsights. In January 1920, BuOrd contracted with Carl L. Norden to improve the Navy Mark III bombsight, a modified Wimperis device.
Norden, born in Semarang, Java, to Dutch parents, studied mechanical engineering at the Federal Polytechnic Institute in Zurich and immigrated to the United States in 1904. He worked for Elmer Sperry for two years developing ship gyrostabilizers. Their relationship was a rocky one; Sperry disliked Norden’s appetite for “vile black cigars” and Norden resented Sperry’s proposal that Norden sign over his future gyrostabilizer patents to the Sperry Gyroscope Company. They parted ways in 1913 although they worked together on various projects during World War I.
Norden’s first efforts included adding gyro-stabilization to the bombsight along with a telescope to better sight the target and a means for providing flight directions to pilots. When the results were unsatisfactory, Norden used Navy funding for three pilot direction indicators (PDIs) for the Mark III bombsight and family funds to continue work on a better bombsight. In June 1922, impressed with his progress, the Navy contracted with him for three experimental bombsights designated the Mark XI.
A year later, the Navy was concerned the project was too big for one man, especially the man they knew as “Old Man Dynamite” because his generally unsociable and reclusive nature made him so difficult to work with. They sent him a collaborator, Theodore Barth, who was known as a practical engineer and a man who could get things done. This successful relationship lasted until both men retired after World War II.
Norden worked out of his home and Barth’s apartment and, using the equipment and skilled labor of the Witteman-Lewis Aircraft Company, delivered the three PDIs and three experimental Mark XI bombsights—all handmade—to the Navy in the winter of 1923 and spring of 1924. Mark XI bench and flight testing was conducted at the U.S. Naval Proving Ground in Dahlgren, Virginia, in 1924. Neither Norden nor the Navy was pleased with the performance of the bombsights as test bombs fell with “alarming irregularity.” The Navy also believed the sight was too complicated.
Many changes were identified during the initial testing, and BuOrd contracted with Norden for modifications to two of the Mark XI bombsights. The modified bombsights were delivered to Dahlgren for flight testing in 1925. Tests during the summer and fall of 1925 showed that the changes were worthwhile. The 18 bombs that were dropped from an altitude of 3000 feet in the final test achieved a mean impact point that was 9 feet short in range but 187 feet to the right of the flight path.
The Navy test bombardier was impressed but reported the sight was too complex and required “both hands, both feet, and the teeth” to operate. In an open cockpit, the wind and cold made fine adjustments to the sight nearly impossible. Norden viewed the basic design as good and the problems correctable. He left Barth in New York to consult with the Navy and went to Zurich for a year to work on design changes.
The Navy completed Mark XI testing in October 1927 and, despite continuing problems with leveling, vibration, and the PDI, began negotiations with Norden and Barth to purchase eight Mark XI bombsights and PDIs. Norden and Barth balked at the proposal because they considered themselves consulting engineers not production contractors. In 1928, after additional encouragement from the Navy and some unwritten agreements, they agreed to form Carl L. Norden, Incorporated. They agreed to produce and deliver 80 Mark XI sights with spare parts and toolkits for $384,000. They also agreed to transfer all patents, models, and designs to the government two years later. Norden said he was paid $1 for these rights, although Navy records show he was paid $250. A very low price in either case.
Bureau of Ordnance testing related to the Mark XI bombsight development at the Naval Proving Ground in Dahlgren, Virginia, began in 1922. In the five and a half-year period leading up to the production contract, Norden and Barth visited Dahlgren 51 times. The bench and flight testing at Dahlgren is credited with uncovering numerous design and performance issues. Dahlgren was also the site of the first school to teach mechanics how to maintain the Mark XI bombsight.
Mark XI production began slowly, and Norden and the Navy tested and improved each sight as it was produced. Norden shipped the first three Mark XI bombsights to Dahlgren for testing in early 1929. The bombsights were essentially handmade, and production continued at three units per month. With all of its shortcomings and complexity, the Mark XI represented a significant improvement over other bombsights. However, it did not resolve the limitations of high-altitude horizontal bombing.
After signing the contract with the Navy, Norden went to his mother’s Zurich home to work on his next design, the Mark XV. This was the bombsight (known as the M-series by the Army Air Force) used by both the Army and the Navy during World War II. Two prototypes of the Mark XV, a timing sight and a synchronous sight, were delivered to Dahlgren in February 1931 for evaluation.
Testing at Dahlgren was intended to identify deficiencies in a new concept, not as acceptance tests. For this reason, the Naval Proving Ground conducted extensive bench tests of the components of the sight as well as intensive flight testing. Dahlgren provided a final report to BuOrd containing 33 pages of deficiencies and suggested corrective actions. Flight tests showed the Mark XV was twice as accurate as the Mark XI (i.e., the percentage of hits was twice as high). Testing ended in August 1931 when BuOrd issued a production contract for the Mark XV bombsight.
The Mark XV was given more tests than any other sight ever developed by BuOrd. Life tests of various components and analytical studies continued into 1932. On April 18, 1932, the first order for the new sight was placed—32 for the Navy and 23 for the Army. The Navy received its first production unit in September 1932, and the Army received its in April 1933. The sights continued to be nearly handmade and every unit went to Dahlgren for calibration and acceptance testing.
The Naval Proving Ground received Norden’s next improvement—the Stabilized Bombing Approach Equipment (SBAE)—in February 1935. The SBAE, an automatic flight control system, transferred adjustments of the bombsight’s controls through mechanical linkages to the azimuth gyro and allowed bombardiers to fly aircraft in roll and yaw. Testing revealed both the strengths and weaknesses of the prototype. Flight tests showed a 30 percent improvement in Mark XV accuracy in smooth air and 39 percent improvement in rough air. The first production models were available in late 1936, and production began in June 1937 at the rate of seven to ten per month.
The Army Air Corps had long worked with the Sperry Corporation to equip its aircraft with autopilots. Carl Norden, who continued to compete with Sperry, preferred to work with the Navy rather than the Army. (He once told an Army colonel “No man can serve the Lord and the Devil at the same time—and I work for the Navy.”) He argued that it was a duplication of effort for the Army to equip its aircraft with both Sperry autopilots and Norden SBAEs.
The tests at Dahlgren compared the Norden SBAE with the Sperry autopilot and concluded that the SBAE “… is at least the equal of if not superior to the Sperry gyro-pilot.” The Army continued to try to connect the Norden bombsight and the Sperry A-2 autopilot—without success.
When the Air Corps continued to pursue an SBAE replacement, the Navy responded by developing an adapter that allowed the Norden bombsight to connect to the Sperry A-3 autopilot. Dahlgren completed tests of the adapter in August 1941, and 40 units were produced between September and December.
The Army Air Force asked Minneapolis-Honeywell Regulator to develop new automatic flight control equipment (AFCE) with electronic parts to link the A-3 autopilot and the Norden bombsight without the Navy’s adapter. This system (designated the C-1), ordered into production in October 1941, was the standard autopilot/AFCE/SBAE for the remainder of World War II.
Procurement became a major headache because the Navy refused to share production with the Army. Between 1932 and 1938, the Norden Company produced only 121 bombsights per year. Even after Norden added additional production sources to meet Army Air Force needs, shortages of materials, specialized machine tools, and skilled labor kept production below required levels. There was a major shortage of bombsights that extended to late 1943.
All bombsights continued to go to Dahlgren for bench and flight testing. It was estimated this process delayed delivery for four to five weeks. Although BuOrd refused to eliminate the Dahlgren testing, they did make some concessions: one bombsight of every ten produced was sent to Dahlgren for testing. They also agreed that bench testing would be completed on the day the sight was received. Further, Dahlgren would only flight test the number of sights that could be completed within 15 days of bench testing.
As the war went on, it became clear that Army Air Force performance requirements exceeded those of the Navy, and the Navy had little interest in modifying the sight since it had chosen dive bombing as its preferred means of attacking moving targets. Thus, improvements to the bombsight were motivated by the Army and, by late in the war, were being developed by someone other than the Norden Company.
Between 1932 and the end of World War II, nearly 90,000 Mark XV (or M-9) bombsights—81,537 for the Army Air Force and 8,353 for the Navy—were produced at a total cost of $1.1 billion. Production began to catch up with demand by late 1943, but mass production techniques also led to declining quality. The Norden Company was not interested in helping to solve the problem and in late 1944, 75–80 percent of all sights produced failed to meet specifications.
The accuracy achieved at Dahlgren was never duplicated in combat. The Navy specification was for 2.5 mils (or 2.5 feet mean miss for every 1000 feet of altitude). The inherent accuracy of the 1944 Norden sights was 14 mils. By some reports, the accuracy achieved in combat was more than 50 mils.
While some used the discrepancy between design and operational accuracy to question the effectiveness of high-altitude bombing, the performance of the Eighth Air Force in Europe refutes this. In the end, 7.5 million bombs were dropped from an average altitude of 21,000 feet with 31.8 percent of them falling within 1,000 feet of the aiming point. While this did not meet prewar expectations for precision, it did stop German oil production and destroyed 20 percent of German war production in the last 16 months of the conflict.
This article began by noting the Navy’s role in the development of the Norden bombsight is less well known than is its use by the Army Air Force in World War II. Even less known is the role the Naval Proving Ground in Dahlgren, Virginia, played in the development, testing, and acceptance of the Norden bombsights beginning soon after World War I. It’s clear that both deserve credit for their significant contributions to the breakthrough capability represented by the Norden bombsight.