The Development of the DKW F9

On 25 January 1935, director of sales, Dr Carl Hahn, addressed a meeting of the Auto-Union board saying “In my opinion, with today’s models, we can no longer be competitive in the long term. Fundamentally we will have to create new models that, right from the start, are designed for lower cost.”

Since his earliest days in DKW, Dr Hahn had actively sought feedback from the company’s dealer network and was closely attuned to the ‘word on the street.’ DKW had made its name in the budget car market with its Front series of two-stroke light cars. However, in recent times DKW’s position had been steadily eroding against competition from more modern mass manufacturers, such as Opel and German Ford.

Additionally, a new threat had appeared on the horizon in the national ‘Volkswagen’ program. From the published specifications it was already apparent that the new, steel-bodied ‘People’s Car’ would make DKWs wooden bodied, two-stroke cars obsolete.

In 1938 the Volkswagen was unveiled to the German public, and tens of thousands of Germans immediately signed up to the purchase scheme. Almost half a million would have signed up by the start of the war.

The Volkswagen project was clearly an existential threat to Auto-Union as the group was entirely dependent on the cash flow its budget DKW brand generated. There may have been a lot of prestige in the luxury car market, but there was little profit and a lot of expense. The Group's luxury car brands Audi and Horch had been incorporated into Auto-Union in 1932 specifically because they were insolvent. Audi had virtually disappeared and Horch was only a small volume producer.

Dr Hahn outlined a completely new program for the group which would involve the development of a single, modern Auto-Union range that would include a budget entry, mid-class entry and a luxury entry. To save cost, the new range would use modern mass-production methods, share common components and have a distinctive ‘Auto-Union’ style. The new range would completely supersede the existing model range and production facilities would be rationalized. The Auto-Union board endorsed the program and on January 25, 1935, a meeting took place at the Horch plant in Zwickau in which Hahn, Technical Director William Werner, the factory directors Schuh (Audi, Zwickau) and Hoffmann (DKW, Zschopau) and the engine technicians Strobel, Kochen and Trost met to plan out the details of the new DKW program. At this meeting the car that would become the F9 was conceived.


The Engine
Auto-Union engineers were in agreement that DKW’s two-cylinder 700cc two stroke workhorse engine had reached its developmental limits. As the new car would be constructed in steel, a larger, more powerful engine would be necessary. Since 1930 DKW engineers at Spandau had been working on the V4 engine for the ‘Big DKW’ range. This complex engine, which used charging pumps rather than crankcase compression, had proven to be extremely problematic. Although it generated more horsepower than the two-cylinder engine, it had a prodigious consumption of fuel and spark plugs, as well as being prone to catastrophic failure. Despite years of development and improvements, the engine remained unreliable, so an entirely new engine was required.

In 1934, the Scott Motorcycle company of England had unveiled a new three cylinder two stroke motor. The engine drew the immediate attention of DKW engineers and an agent was sent to purchase three Scott triples from a Belgian dealership in 1935, shortly before Scott went bankrupt. The bikes and their engines were shipped to Germany for testing. It would not be correct to say that the DKW three cylinder engine was copied from the Scott engine, but DKW engineers certainly inspected the Scott engines during their studies.

DKW engineers used the standard CA 700cc twin cylinder as their starting point, adding an additional cylinder. In this arrangement the outer pistons were mirrored while the central piston was slightly different. Bench testing showed it generated 24 HP and ran very smooth. In December 1938, the engine plant at Zschopau was given approval to commence production of a preproduction run of 50 engines for road testing. All these engines received the suffix ‘V’ for versuch (research). In keeping with standard DKW practice, these engines were laid out transversely with power transmitted to the drive via a duplex chain, with a wet clutch gearbox on the right-hand side.

The first test engine was mounted in pre-series F8 chassis number 7294. Some modification to the engine and gearbox mounts were required due to the engine’s greater width, but externally the vehicle was indistinguishable from a standard F8. The test vehicle’s first drive lasted only 50 metres before the gearbox gave out. The motorcycle style, multi-plate wet clutch and chain drive could not handle the greater torque of the new engine and disintegrated. A technical report indicates that six motors and gearboxes were shuffled through chassis 7294 in rapid succession - three engines; 429286-V, 429289-V and 429291-V, being cycled through in one six-week period! The placement of the transverse engine was regularly altered, sometimes with the axles and transmission to the front, as in the F8, and sometimes with the axles and transmission to the rear, as in the post-war F89P. A technical report from August 1939 reveals that the longest distance the gearbox was able to travel before failure was only 500 kilometres. Neither an improved cast iron gearbox casing or upgrading from a double to triple primary chain improved things and gearbox problems would continue to plague the test vehicles right through the development phase.

The surviving transverse engine, number 429296-V, is on display at the Chemnitz Fahrzeugmuseum https://fahrzeugmuseum-chemnitz.de/


Fresh thinking was required. Dr Ferdinand Porsche, technical director of the Auto-Union race team (1935-37), had faced similar problems with the gearboxes of Auto-Union’s ‘Silver Arrow’ racers. Not only did the gearbox need to handle the immense torque of the Silver Arrow’s engine, it also needed to be short enough to fit within the cars’ limited wheelbase. Porsche achieved this with a compact gearbox unit that doubled back on itself. The design team adapted Porsche’s gearbox for use with the triple engine. This required the engine to be turned 90 degrees and the gearbox mounted longitudinally, doubling back on itself to place the axles immediately behind the engine. With the gearbox problem seemingly resolved (although issues would persist through the early 1950s in the IFA F9), the new unit was returned for trials in the F8 test car.

By 1939 the development of a longitudinal three-cylinder engine and its gearbox were completed, which allowed the factory in Zschopau to begin preparations for series manufacture at the end of the same year. At least 49 three-cylinder pre-production engines were manufactured with numbers 429801-429900 being reserved as test engines. Test department documents mention three successive stages of production with engines up to number 429810 described as "old execution." Like their transverse predecessors, each three-cylinder longitudinal engine remained in their test chassis only a few thousand kilometres before replacement.

The car had two new features for the Front series. Earlier DKWs had the petrol tank mounted in the engine bay and used gravity feed, but new German road safety legislation required the tank to be moved to the trunk. This required a mechanical fuel pump to be installed. They also used a distributor ignition system that was mounted at the front of the engine. However, once road testing began it was discovered that vibration from the fuel pump progressively disrupted the ignition timing. Changing fuel pumps and modifying the ignition did not seem to help so, under time pressure, the engineers abandoned the distributor in favour of DKW’s tried and tested motorcycle technology of breakers and ignition coils. This simple and fool proof technology would remain as standard DKW practice until the DKW Munga was retired in 1970.

The mechanical fuel pump was also abandoned in favour of a vacuum pump driven by crankshaft pressure. It was a simple solution but would be an ongoing point of weakness.

One final problem would trouble the engine team – the exhaust. Two stroke engines depend on finely tuned exhaust back-pressure for efficient combustion and fuel economy, but the exhaust flow of triple engine proved a challenge. The early engines followed the pattern of the earlier twin cylinder engine, with an exhaust that turned 90 degrees from the engine and connected to a straight pipe that ran under the car. This arrangement was ineffective in terms of both tuning and noise dampening. Combined with the fact that the air intake also lacked a silencer, engine noise in the cabin of the first cars was considerable. Development of an effective solution was outsourced to the recognized exhaust tuning specialist, Eberspächer. A transverse pre-production engine was assigned to Eberspächer for testing, but they were never able to develop an exhaust that combined acceptable sound dampening without negatively influencing power, torque and fuel consumption. The design office decided to entrust the manufacture of the exhaust system to a new company called Bertram, who would ultimately succeed where Eberspächer failed. Although no one would realize it at the time, this side show would take on great significance in the future.

The Body
While engine development was following its own trajectory, Auto-Union's central body development department under Albert Locke, began styling the new “Auto-Union” form. Like other German companies, Auto-Union had dabbled in streamlined car projects from the early 1930s. Streamlining specialist Paul Jaray had bodied a Audi Front in 1933 and a rear engine DKW streamliner. Neither projects proceeded. The opening of Germany’s first Autobahn in 1935 gave a practical incentive to the movement and Auto-Union began to experiment more seriously. Improved Audi and DKW Jaray streamliners were tested in 1935 and these trials demonstrated conclusively that vehicle bodies with low air resistance were capable of faster speeds and lower fuel consumption than standard body designs. These findings were supported by independent studies performed by the Kamm Institute and Konig-Fachsenfeld.

The 1937 rally and endurance racing season presented the company an opportunity to present a number of body designs for road testing. For the Rome to Liege endurance race, Auto-Union fielded a series of Wanderer W24s with streamlined aluminum cabriolet bodywork. Unfortunately, the Wanderers did not perform well in the race, but this was entirely due to their un-supercharged engines, which proved extremely unreliable. The team returned the following year with improved engines and performed better. https://dkwautounionproject.blogspot.com/2017/07/auto-union-streamliners.html

A more mature and practical design was applied to the 1938 Berlin Rome endurance race. A new aerodynamic body was designed by Albert Locke and manufactured in Spandau and fitted to pre-series F8 chassis numbers 7301 to 7307. The cars were fitted with a variety of engines, including a tuned 700cc two-cylinder engine and a 700cc charging pump twin cylinder racing engine develop for sidecar racing. It's likely that the pre-production transverse 3-cylinder engine was also trialed.

In the body development department, stylist Gunther Mikwausch and modeler Wilhelm Bohm adapted these competition concepts for practical use. Sculpted models were tested in wind tunnels and the results assessed. They commissioned an expert opinion from the leading scientist in the field, Professor Wunibald Kamm. Kamm was able to share comparative figures of the ‘Volkswagen’ and the Adler Autobahn, and suggested some minor changes to the body to reduce sensitivity to side winds.

By 1939, Dr Hahn’s vision of a modern ‘Auto-Union’ program had begun to crystallize. The first example was unveiled at the Berlin International Motor Show in March 1939. The Horch 930S was a far cry from the proposed DKW ‘Hohnklasse’ (high-class). The beautifully sculpted, streamlined body was cutting edge for the period. Fittings were luxurious. The most notable feature was a fold out basin for hot and cold running water that popped out behind the front wheel. The Horch 930S was not a mass production car however as the body was entirely hand crafted.

The Berlin exhibition car stunned motoring journalists and the motoring public alike.

The DKW Hohnklasse would not receive such an elaborate body, but its similarity to the Horch 930S was obvious. After mock-ups, a hand-built body was mounted on F8 chassis number 7350 with a transverse mounted engine numbered 429290-0 in 1939. Road trials demonstrated an impressive Cd factor of 0.42. The regular failure of the early gearboxes resulted in constant switches of engine and gearbox units. Longitudinal engine numbers 429804-V and 429817-V successively replaced the transverse engine.

Chassis 7350, with sunroof and two-tone paint scheme was used in a promotional photo shoot for the new model, now designated the F9 in accordance with DKW nomenclature. A brochure was prepared for the 1940 production year but plans for series production were placed on hold by the start of hostilities in September 1939.

With all elements of the new car’s design settled, Auto-Union had begun preparation for series production. The Zwickau engine plant was given the go ahead to manufacture another hundred engines. At least five vehicles were built and registered for the road by 30 September 1939. Apart from the original chassis 7350, the others were recorded in the archives as chassis 7294, 7344, 7347 and 7358, with chassis 7344 built as an open-topped tourer.

Dr Carl Hahn received one of these early cars and would clock up many thousands of kilometres, providing extensive reports on all aspects of the cars’ handling to the design team. Dr Hahn’s car constantly cycled back through the factory for improvements and repairs. One of the problem areas quickly identified by road testing was the inadequacy of the car’s brakes. The F9 was a much heavier car than the wooden bodied F8 and the F8’s cable brakes were not up to the task. Hydraulic brakes were therefore installed as standard and, as these were sourced from the supplier who was gearing up for Volkswagen mass production, costs per brake unit actually decreased.

After Germany’s lightning victory over Poland in 1939, the expectation in Germany was that an accommodation would soon be reached with Britain and France and hostilities would come to end. Therefore, while Auto-Union and other auto companies were officially banned from developing new models, work progressed discretely behind the scenes. Several more pre-production prototypes were hand built in Chemnitz for expanded road testing.

At this stage early plans for monocoque, pressed steel construction were finally abandoned in favor of traditional body and chassis construction methods. Earlier, in 1938 Konrad Schulz and Dr. Rudolf Slaby from Auto-Union’s Spandau bodywork began discussions with independent vehicle designer, Friedrich Maier, in Berlin. Maier had been an aircraft engineer with Junkers before moving into vehicle design in 1934. He held a series of wide-ranging patents for features such as a height adjustable driver’s seat and, importantly, a manufacturing method for self-supporting steel car body. As a demonstration of his method, Maier built a single test vehicle, called the Maier Lightweight. This interesting vehicle was powered by a rear mounted 600cc DKW two-stroke motor. It had a spacious interior, four doors, adjustable height seats and a single, centrally mounted headlight that turned with the steering wheel.

http://heinkelscooter.blogspot.com.au/2014/02/1935-maier-leichtbau.html

Nevertheless, Auto-Union management could not convince themselves that Maier’s small-scale operation would be able to meet their requirements for mass production and they were fearful that any infringement of Ambi-Budd’s patents would embroil them in costly litigation. They decided to continue with traditional body and chassis construction methods. In fact, it wouldn’t be until 1957 with the DKW Junior that Auto-Union finally adopted monocoque construction. Panel bucks were constructed at the body development facility and passed to the industrial tool company Allgaier to build the panel presses necessary to commence series production.

The war situation however, did force the company to consider alternatives to steel body panels. Auto-Union had been experimenting with thermoset plastics for several years in the mid-1930s in partnership with several different chemical firms. The main technical challenge was mass production of panels in the required thickness. This problem was finally solved in conjunction with Dynamit AG using resin pressed sawdust, but these plans too were derailed by the war.

Against domestic expectations, the military situation escalated rapidly with the German invasion of Denmark and Norway in April 1940, followed immediately by the invasion of France and Low Countries in May. Surprisingly it took the Germans only six weeks to defeat the French army and push the British Expeditionary Force across the English Channel. In the short pause that followed peace feelers were put out to Britain and there was hope in Germany that the war would end. The British however, doggedly refused to negotiate and the road was opened for a long war. Government bans on new civilian vehicle production were extended and would become progressively more restrictive as the war went on.

In 1941, despite the war having expanded on all fronts, the German government began to wind down military production and another round of discrete production began at Auto-Union. Although penalties for breaching the ban were severe, an internal report from 1943 suggest that at least another fifteen F9s were built during the second half of the war, including six experimental chassis. All the F9 pre-production cars were improved through multiple iterations of modifications identified by extensive road testing, including the detailed reports and recommendations provided by Dr Hahn himself. Dr Hahn was to observe that in the DKW F9, Auto-Union finally had a car that could hold its own against all other auto makers.

As a sidenote, its important to understand how all this fitted within Dr Hahn’s overall program. Having addressed DKW’s budget offering, attention focused on Auto-Union’s mid-range car brand, Wanderer. Auto-Union had begun cross branding Wanderer and DKW products in the later 1930s. In 1936 Wanderer W24 sedan bodies were fitted with the unreliable DKW 4=8 engine and sold as the DKW ‘New’ Sonderklasse. The Sonderklasse received a different radiator grill and fittings were downgraded for the lower-mid range market, and despite the DKW’s inferior and troublesome engine, the new Sonderklasse still managed to outsell its Wanderer counterpart. Nevertheless, the Wanderer-Sonderklasse was always recognized as a stop gap until the new Auto-Union range hit the market in 1940.

Auto-Union’s new mid-range offering would be the Wanderer W4 (also known as W31), which was basically the new Hohnklasse body fitted with a four-cylinder Wanderer engine. A mock-up was prepared, followed by at least three prototypes for road testing, including a two door sedan, two door cabriolet and a four door sedan.



Thus, by the end of 1941, the new Auto-Union program for the post-war period was clear. At the top of the range would be the luxurious Horch 930S. The mid-range would be serviced by the Wanderer W4 (with the six cylinderW6 model as an upgrade option), and the lower mid-range would be serviced by the DKW F9 Hohnklasse. All other classes and models would be retired except the small engine 600cc F8 Reichklasse, which would remain as the sole budget offering until the Volkswagen entered full production when it too would be retired.

Initial sale price for the DKW F9 was forecast to be RM1750, but this was expected to be reduced to RM1180 when full production of 250 cars per day. However, in order to achieve this volume would require a comprehensive rationalization of Auto-Union's dysfunctional distributed manufacturing process. DKW bodywork was manufactured at Spandau, outside Berlin, while engines were manufactured at Zschopau in Saxony. Miscellaneous fittings were manufactured at the Framowerkes in Frankenburg or sourced from specialist companies across the country. All the parts were then shipped to Audi's Zwickau plant, where chassis were constructed, and final assembly took place. Even if there were exemplary logistical cooperation between all these plants, this arrangement could not achieve modern mass production. In 1936 Auto-Union had purchased the former Prestowerks plant in Chemnitz, taking over Presto's modern, new office buildings as their company headquarters. In 1938 Auto-Union began fitting out the vacant factory floor as an integrated mass production facility for the new Auto-Union range. All elements - engine, chassis, body and assembly - would be manufactured here on a modern assembly line. Once completed, the bodyshop at Spandau and the Wanderer plant at Seiglitz would be sold or converted to other purposes. By this consolidation and simplification of production, Auto-Union would remain a viable automobile manufacturer in a post-Volkswagen world.

But it was not to be. By 1944 Germany’s position was dire and all industrial resources were diverted to war production in a vain effort to stave off inevitable defeat. Despite this, at least one more F9 was built in late 1944 as the personal car for technical director William Werner. This car, the very last built, incorporated all of the improvements and modifications made to the pre-production series, but was fitted with a very early pre-production engine no. 429801-V, suggesting that it was retrieved from the testing bench.

Werner did not get to enjoy use of this car for very long as it was badly damaged during a bombing raid over Berlin in late February 1945. The car was transported to Chemnitz for repair and, as the car was not running, it was not evacuated with the other prototypes to Saupersdorf in April 1945, thereby escaping their unfortunate fate (see further below). As part of an attempt to evacuate Auto-Union technology, plans and tools out of the Soviet zone, the car was transferred to the Auto-Union branch in Hanover in May 1945. In 1946 the British Ministry of supply seized the car as part of the Allied program to confiscate and evaluate German technology and shipped it to the School of Tank Technology in Cobham, Surrey for evaluation. An article from July 1946 indicated that no detailed technical assessment was made although it was briefly mentioned in BIOS report 21. (‘The Motor Car Industry in Germany during the period 1939-1945’, pg. 21), which noted:

“the F9 car was developed solely to meet competition offered by the Volkswagen. The Volkswagen (subsidised) was cheaper than their F8 model and had more room and was faster. They reckoned that in producing the F9, which had a much better appearance than the Volkswagen and was about 15 kph faster, they would hold the market somewhere between the Volkswagen and higher priced cars.” (quoted from Karl Ludvigsen’s “Battle for the Beetle” page 373.

The seizure of Werner's F9 was reported in both British and German automotive magazines. This German copy of a British report (above) contains many inaccuracies caused by the confusion of the postwar period. 

"The only copy of a 3-cylinder DKW, a model developed in Germany during the war in the hope of better days, has recently been presented at an English exhibition. This model, which "Auto-Review" briefly reported in issue 23/1945, was recovered unfinished by British forces in the Soviet occupation zone and shipped to England. It was repaired by the English and then analyzed, its particularities requiring a thorough analysis. Its characteristics are as follows: 3-cylinder in-line, 2-stroke engine, front engine and transmission, 4-speed gearbox, radiator behind the engine, very large interior."

How thoroughly the car was repaired is questionable as it was soon handed to the Australians for assessment. The war had convinced the Australian government that the country needed to develop an indigenous automobile industry and in 1944 proposed that Australia begin manufacturing “a car similar to the German low priced two-stroke DKW.” However, by the time the vehicle arrived in Australia a deal with General Motors had been agreed for domestic vehicle manufacture by GM Holden. Additionally, the complex legal web of patent ownership that was soon to play out between East and West German successor companies was sufficient to dissuade the Australian government of any plans to build an Australian DKW. 

In September 1950 the car was sold by public auction of surplus military vehicles along with a twin cylinder DKW F8 and a V4 DKW Sonderklasse. DKW enthusiast, Steve O'Meager purchased the car but found that its engine was unserviceable, having broken pistons, a broken conrod and one third of the lower crankcase missing [probably the engine was run without oil until destruction - ed]. Being a creative mechanic, O'Meager fabricated replacement parts and got the engine running again. He also converted the car to left hand drive to meet Australian licensing requirements. The car provided him with years of reliable service, including several long trips interstate. Ill health forced him to sell the car in 1983 to Mr Leo Redfern, a fellow DKW enthusiast. By this stage the engine had failed and Leo assessed it as being beyond repair. To keep the car on the road, he acquired a 1953 DKW F89P and performed a body swap, transplanting the body onto the F89P running gear. The original bonnet and radiator screen was badly rusted out so was replaced with that of the F89P. Fortunately, Leo retained the engine and chassis.

The Werner F9 in the late 1980s when owned by Peter Thorogood.

In the late 1980s, the car was purchased by DKW enthusiast, Peter Thorogood. Peter also obtained the original chassis and engine (which had been dismantled and stored in a box) and planned to do a full restoration. However, when the economy turned bad in the early 1990s, he was forced to put the car up for sale. By this stage Audi AG had become aware of the car and were keen to obtain this important vehicle for their collection. Forty odd years after it had left Germany the prototype finally returned home. Audi Tradition sent the car to a restorer in Estonia who undertook a largely cosmetic restoration. This however involved a substantial amount of body repair. An early IFA F9 bonnet and grill replaced the missing original and the car was reverted to right hand drive. The engine however was not fully repairable. The car is now on display at Audi's Ingolstadt museum where it can be enjoyed by everyone.

The Werner car has been exhibited at many auto shows.

The car getting a thorough inspection at an exhibition at the Chemnitz Fahrzeugmuseum in 2015.
https://dkwautounionproject.blogspot.com/2021/08/the-werner-dkw-f9.html

The fate of the other cars:
Surviving photos of the cars allow some cross referencing of their registration numbers with their chassis numbers.

Chassis 7347 was originally registered as IA-230889 in Spandau before going to Chemnitz where it received the license number IV-73330.

As for 7350, despite the registration number IV-34564 appearing in the press photos, it was actually registered at the Audi factory in Zwickau as V-12126. The registration number IV-34564 appears on a number of different vehicles in a number of publicity shoots.

Chassis 7352 was owned by the Spandau testing department under registration number IA-228349. Chassis 7353 was registered in the district of Chemnitz as IV-4267.

Chassis 7358 was also registered in the same district as IV-4410.

The above car, chassis 7358, was presented to the Soviet Administration in Leipzig before being shipped to the USSR for evaluation in 1946 where it received the license number Proba 13-10. This very early car would be discovered in Estonia in 2016 and has since been shipped back to Germany for restoration. https://dkwautounionproject.blogspot.com/2017/06/a-second-dkw-f9-prototype-discovered.html

We do not know which chassis were hidden under the registrations IA-34561, IV-5188, IV-0630 and IA-225868.

The 4-door prototype along with all surviving F9s and pre-war endurance racers in Auto-Union possession were evacuated to a vacant mill in Saupersdorf for protection in March 1945. Unfortunately the mill and all the vehicles were destroyed during fighting there in April 1945, immediately prior to the end of the war. The remains of the 4-door car were transported to the DKW Spandauwerkes for assessment but was too badly damaged for restoration.

By 1946 neither the new Auto-Union in the west nor its East German counterpart had an F9 in their possession and it seemed that all the pre-war development was lost. It would take a miracle for this particular phoenix to rise from the ashes. Surprisingly, the F9 would be resurrected twice.

The resurrection of Auto-Union: https://dkwautounionproject.blogspot.com/2020/05/collapse-and-reconstruction-history-of.html
The development of the DKW F89P: https://dkwautounionproject.blogspot.com/2017/06/dkw-f89p-new-meisterklasse.html
The development of the IFA F9: https://dkwautounionproject.blogspot.com/2020/07/the-development-of-ifa-f9.html

This post incorporates detailed information from the historians Ralf Friese and Frieder Bach published in German. Frieder Bach's books are available here: https://fahrzeugmuseum-chemnitz.de/das-museum/publikationen/