Oddball or standard-setter? This was the mystery that has long surrounded Porscheʼs Type 645 Spyder, an unique and advanced design built to meet tough competition. We lift the veil to tell the story of the advanced prototype whose behaviour won its ʻMicke
Karl Ludvigsen looks back at the history of the wayward racing Porsche
One of the oddest episodes in Porsche history is the brief and literally meteoric career of the Type 645 Spyder. The 645 was in public view in Germany for less than four months from its first appearance to its last in the autumn of 1956. Its design, which dated to more than a year earlier, had been prepared as a successor to the Type 550 that would be lighter, more aerodynamic and superior in roadholding.
Porsche pitched its 550 Spyder into Europeʼs most competitive sports-car racing class. France was in the 1500cc category with its Gordini, Britain with Cooper and Lotus, East Germany with its six-cylinder EMWS, Italy with both Maserati and OSCA, and Germany herself with fuelinjected Borgwards. Although the 550 was off to a good start with the Fuhrmann engineʼs 1954 introduction, its chassis concept dated in some respects to Walter Glöckler ʼs racers as far back as 1950.
Egon Forstner decided to address this shortcoming. An Austrian who had joined the Porsche cadre in Gmünd in the 1940s, Forstner was a versatile engineer with patents in brake design, cooling systems, valve gear and tractor design among others. Moving to Stuttgart, he took over from long-serving Josef Mickl as head of the calculation department. It consisted of his assistant Ernst Henkel and, from 1956, newcomer Hans Mezger.
ʻThe calculation-department office was above the experimental department,ʼ said Mezger, ʻwhere everything was in one big area. We were on the third floor above the second-floor office known as the D-zug or D-train because it had side windows like those on a train, looking out on the experimental area below. In the D-zug office were about eight engineers in total with the chassis people on the left and the engine designers on the right.ʼ
In the latter part of 1954 Egon Forstner decided to start work on the design of a new body and chassis to carry the Type 547 four-cam engine. Counting on their enthusiasm for racing cars, he reached out to others on the Porsche staff for
help with the project, given the Type 645 designation. Two who signed up were engineer Ernst Fuhrmann and body designer Heinrich Klie.
Fuhrmannʼs involvement could well have taken place because he saw this project as an opportunity to burnish his credentials in the design of racing vehicles as well as the engine field in which his four-cam engine was already established as successful. Ambitious as he was to lead Porscheʼs engineering team, Fuhrmann needed to be seen as more than an engine expert.
So that their Type 645 could slip more smoothly through the air, the engineers reduced its frontal area by narrowing its track. Instead of the 550ʼs 49.0 inches this became 46.9 inches at the front and 45.3 inches at the rear. Producing a fifth-size clay model, Klie fitted its body closely around the wheels, partly shrouding those at the rear, and rounded its nose in plan view.
The usual drag-inducing opening for air for the oil cooler was eliminated by making the front lid itself a surface-type cooler, with a labyrinth of passages underneath its surface, left unpainted to improve heat radiation. Intriguing highly styled shapes were given to faired-in lamps at both ends of the car.
An aerodynamic feature that appeared on Klieʼs design model of the Type 645 was a headrest for the driver behind which was an oval-shaped grilled air passage into the engine room. As expressed in the patent granted Fuhrmann and Klie on the design, the headrest was shaped to create a turbulent zone of high pressure behind it, above the grilled aperture, to reduce the power lost in supplying cooling air to the engine.
Also patented by Klie and Fuhrmann was an alternative means of delivering cooling air to the engine bay. This was a rearward-facing slot almost the full width of the lid covering the engine, positioned and designed to preserve smooth airflow above the deck while admitting air under pressure into the rear compartment. A small central bulge covered the engine-fan housing. Rearward-facing air inlets
“REDUCED ITS FRONTAL AREA BY NARROWING THE TRACK…”
were above the carburettors.
The narrowness of Forstner ʼs Type 645 was to help make it lighter, as would a 6.2-inch reduction from the 550ʼs wheelbase to a mere 76.4 inches. It was to be bodied entirely in magnesium, lighter but less durable than the usual aluminium. There was no right-hand door and indeed there could not be one, for the fuel tank was placed along the right side of the body, counterbalancing the weight of the driver on the left.
An important attribute of the new Spyder was a multitubular space frame. The concept was well known to Forstner and also to Fuhrmann, who had been with Porsche in Gmünd when it was working with Italyʼs Cisitalia, which specialised in such frames. Indeed Erwin Komenda had designed a space frame for the mid-engined VW Sports roadster of 1948, later hailed as the first ʻPorscheʼ car.
The frame designed for the 645 was breathtakingly sparse. One key element was a rectangle of large tubes located at the cowl, carrying the steering-column mounting. Three tubes at each side braced this to the assembly of crosstubes that carried the narrowed trailing-arm front suspension and its anti-roll bar. Steering was by equally divided track rods, operated by a small drag link from the steering box, unlike the standard car ʼs unequally divided track rods.
Another important element was a braced structure of small tubes above the familiar crosstube that housed the rear torsion bars. Rising rearward from it were small tubes that peaked at a high crosstube whose ends were mounts for the rear dampers. Hanging from it was a fabricated cradle that carried the engine-gearbox assembly, attached under its bell housing. Completing the frame was an X-brace at each side of the cockpit plus single diagonals of small tubes bracing each of the structureʼs open quadrilaterals.
For the Grand Prix Cisitalia of 1947–48, Porscheʼs Type 360 project, the engineering team in Gmünd, Austria designed a sophisticated rear suspension. Instead of the swing axles of the pre-war Auto Unions it used upper and
lower lateral links to guide each wheel hub, with brake and traction torque taken by a trailing arm. Giving precise wheel control with low unsprung weight and far less camber change than the usual swing axles, this was an immense step forward.
This suspension made its reappearance in the Type 645. Here the trailing arms were the usual Vw-porsche blades, drilled for lightness and set at a static angle of 10 degrees above horizontal. Upper and lower tubular links went inward and slightly forward to pivots attached to the back of the engine cradle. While the lower links were horizontal the upper ones sloped downward, toward the centre, at 13 degrees. This gave a rear roll centre that was higher than that of the Cisitalia yet not so elevated as that of a swing axle. The design provided for two degrees of negative camber at rest to enhance the grip of the eraʼs narrow 5.25 x 16 tyres.
Fabrications attached to the ends of the trailing arms carried each rear hub in a double-row ball bearing. Drive half-shafts had Hooke-type joints at their outer ends and pottype inner joints that could slide to adapt their length to suspension movement. Although simply arrived at, with its telescopic dampers this was a sophisticated linkage for 1955. Not until later in the 1950s would such suspensions begin to be adopted in Grand Prix racing.
This ambitious project was well on its way to realisation when Porsche chief engineer Karl Rabe, taking a break from his concentration on tractors, discovered what Forstner was up to. On 15 February 1955 he memoed Porscheʼs senior executives that he felt it ʻabsolutely essentialʼ that ʻa fundamental discussion take place with Herr Porsche about this vehicle.ʼ He made the following observations:
ʻI canʼt envision that one man carries this forward alone who at the same time remains the only remaining theoretician for the design office. I would not like to hide the fact that Herr Forstner has already asked several times for our help, which with the best will in the world I could not provide in view of the present workload in the design office.
ʻI consider it necessary to clarify the question of the cost of this vehicle. Hitherto Herr Forstner has only remarked that the vehicle will not be more costly because it will largely be built in-house. To this I would add that I have found no primary contract covering the creation of such a vehicle.ʼ
Rabe had rumbled Forstner ʼs end run around Zuffenhausenʼs procedures. The Type 645 hit the buffers, apart from some discussion about the addition of lightness by using magnesium instead of aluminium for its gearbox housing. It languished during 1955 when the existing 550 Spyders seemed able to hold the fort. For 1956, however, when Borgward was known to be readying its 16-valve fuel-injected four, competition looked to be intensifying. Type 645 was relaunched on 16 February by work order number 9159 calling for the production of two cars ʻas soon as possibleʼ.
Assuming sensibly enough that Porscheʼs management wanted these cars to play some part in the 1956 season, on 28 February Egon Forstner advised Messrs Rabe, von Rücker, von Hanstein, Hild and Fuhrmann that he considered the timing ʻexceptionally tight and requiring the greatest haste.ʼ Raw materials and drawings were available, he said, for the cars to be built in the experimental department. The magnesium bodywork could be formed in parallel with the other work to speed things up, he considered.
The latest development with the Type 547 engine was to drive its distributors from the nose of the crankshaft instead of from the ends of the camshafts to eliminate variations in timing caused by the latter arrangement. Space for the new drive, said Forstner, was not obviously available in the tightly packed 645. Although he requested a drawing that would show space for the newer engine, this never materialised so his 645 would always be equipped by the older style of
“RABE HAD RUMBLED FORSTNER’S END RUN…”
engine, whose output disadvantage Forstner considered to be as much as 20 horsepower.
In mid-june of 1956 Egon Forstner recapped his brainchildrenʼs state of affairs. One of the two cars had been completed and tested on the Malmsheim skid pad. It was finished just in time to be taken to the ʼRing for trials on 15/16 May alongside a 550 and a 550A. Taking its wheel, Wolfgang von Trips just broke 11 minutes on his second lap but did not persevere further. In contrast, Herrmann in the 550A kept slashing his times, after various changes of tyre pressures and anti-roll bars, to a brilliant 10:35.2.
Having also tried the 645 Hans Herrmann said, ʻIt was certainly faster but totally undriveable. Von Trips and I declined emphatically.ʼ Hansʼs judgement that it was faster was a form of validation for what Forstner and his small team had wrought, but a racing car must be manageable as well as fast. Herrmannʼs verdict in particular was anything but positive because he was undeniably skilful.
Nonetheless the 645 was among the cars that Porsche fielded for the 1000-kilometre race on 27 May at the ʼRing, where it was driven in practice by Richard von Frankenberg. Still immature, it was rejected by Frankenberg in favour of the 550A Spyder in which he turned faster laps.
The Type 645 showed that its expected high-speed advantage was indeed present. On straights, Forstner reported, it had ʻvery steady roadholdingʼ. With suitable gearing its maximum speed was 162mph against 158mph for the 550A in its best 1956 factory trim, in spite of the horsepower deficit of its outdated engine. Forstner said its four was producing a meagre 98bhp.
Handling, however, was judged treacherous. Herbert Linge tested it at Malmsheim and concluded that ʻthe car was so terrific in the back that the front axle was overwhelmed.ʼ You must remember that it was always hard to get enough testing time. Drivers wanted the latest, fastest car immediately. In this case there wasnʼt enough time to set up the front suspension properly.
ʻRear adhesion was enormous,ʼ Linge told Jerry Sloniger. ʻYou started out with a great deal of understeer. But when the tail did break loose it came around like a cannon shot. Nobody could catch the spin.ʼ
The new rear suspension offered better grip which was not counterbalanced by the cornering power of the car ʼs front trailing arms, which leaned its wheels outward when the body rolled. ʻThe existing difficulties that have shown up with cornering,ʼ assured Forstner, ʻmust be able to be eliminated by rational judgement without undue difficulty.ʼ
Among the changes already implemented was a new
linkage arm that speeded up the steering by 28 per cent to help drivers catch the sudden breakaway. Interference between the new central steering arm and the abutment screw for the upper torsion bar was eliminated. The Silentbloc bearings supporting the front of the engine were reoriented to prevent tearing. The height of the fuel tank was reduced to cut its capacity from 130 to 80 litres, lowering the 645ʼs centre of gravity. Compensating would be a tank of 40 or 50 litres outside the frame to the driver ʼs left.
In spite of his car ʼs evident promise, Egon Forstner complained to Porsche, Rabe, von Rücker, von Hanstein and Hild on 19 June, ʻitʼs been standing in Experimental for weeks, completely abandoned.ʼ He enumerated the improvements made in the interest of better roadholding. Putting its front suspension and steering back together would take a mechanic only two days, he said. All that was needed in order to resume trials was an engine.
Moreover, Forstner added, the frame for 645 number two was halffinished. Some expenditure was needed to source such parts for it as the rear uprights and half shafts, torsion bars and dampers. ʻThe outlays already made oblige that the final step should be taken, namely to commence trials of the vehicle to win all the knowledge that is there to be won.ʼ
Egon Forstner concluded his report to Porscheʼs senior racing cadre with a heartfelt query about the future of the Type 645, its half-finished sister and its assortment of components. Always insightful, the experienced Karl Rabe had raised reservations a year and a half earlier about the potential of what was essentially a one-man project in the demanding Porsche environment where every man and every expense had to count. Now he was asking — no, begging — the powers that be for their support of what would be his only attempt to create a complete automobile.
Their response was to leave the second car unfinished and to authorise completion and testing of the first Type 645. Even after modifications its handling was demanding. Causing heavy initial understeer, in turns its rear tyres would grip much better than those in front. This would be followed by a sudden and hard-to-catch transition to oversteer at the limit. Contributing to this were its short wheelbase and low moment of inertia about its vertical axis, a function in part of its centrally-mounted fuel tank.
Having tasted the Type 645ʼs speed in practice at the ʼRing in May, Richard von Frankenberg was willing to cast his lot with it. Then 34 years of age and wearing glasses with heavy lenses, Frankenberg was not only skilled as a sportscar driver but also renowned for his bravery. Although he took this new kind of Porsche under his wing, he had no illusions about its attributes. He dubbed it ʻMickey Mouseʼ, not because of its smaller size but in recognition of its tricky behaviour.
Mickey Mouse first raced close to home on the Solitude circuit on 22 July 1956. Von Frankenberg qualified the new model in the front row, second only to Herrmann in a 550A, but lost first gear even before the 99-mile race began. On his first lap braking problems surfaced, a regular feature of the car which made approaching corners as exciting as driving through them. Power faded, too, as oil temperature soared — a limitation of the front-deck radiator? — but von Frankenberg soldiered on to a fourth-place finish behind winner Herrmann, von Trips in a Porsche and Edgar Barthʼs East German AWE.
In August Mickey Mouse practiced for the sports-car race that accompanied the German Grand Prix at the Nürburgring but did not compete. All Porscheʼs drivers concluded that its handling idiosyncrasies were too daunting a challenge for a 312-mile race over that treacherous track.
The Type 645ʼs next appearance was at West Berlinʼs Avus for the 152-mile Grand Prix of Berlin on 16 September, round six of the German Sports-car Championship. Because high speed was decisive at the Avus, with its long straights and steeply banked turn, von Frankenberg chose the Type 645 for this race in spite of the car ʼs known quirks. Like
“CHOSE THE TYPE 645 IN SPITE OF THE CAR’S KNOWN QUIRKS…”
Porscheʼs other Avus entries it was given a tighter tonneau covering the cockpit with a low wraparound windscreen sheltering the driver.
Practice at the Avus promised no dominance for the Mickey Mouse, which clocked the third fastest time behind Barthʼs AWE and Roy Salvadoriʼs mid-engined sports Cooper-climax. Soon after the start, however, von Frankenberg took the lead from Salvadori on braking for the flat South Turn and moved ahead of the field, trailed closely by the 550A of Wolfgang von Trips.
On their third of 30 laps, the two silver Porsches droned down the back straight and banked onto the vast brick bowl of the North Curve. Built in 1937, it provided a running surface 24 yards wide, 13 yards of this banked at 43 degrees. It was a constantslope banking deliberately intended to allow cars to run both high and low. Rudy Uhlenhaut of Mercedes-benz said it was ʻvery dangerous and you had to treat it exactly as if you were driving on a normal road. If you went too fast you just slid over the top.ʼ
Among the 50,000 Berliners all eyes were on the leading von Frankenberg when his car veered abruptly to the right and powered at an angle across and over the vertical lip at the top of the banking, which flipped it. Cartwheeling as it vaulted two yards above the rim, the car bounded off the outer earth wall, over a wire fence and landed, a crumpled flaming mass, in the street between a Mercedes 300SL and an open-topped Opel Rekord.
Not until some five minutes after this spectacular crash did a Porsche technician, Harry Lörcher, discover the unconscious von Frankenberg lying in the greenery on the earth wall. Having plummeted from the Mickey Mouse as it brushed through an acacia tree he was not, after all, a victim of the white-hot magnesium flames that were consuming the inverted Type 645.
Richard von Frankenberg was never able to remember what happened during this ʻMiracle of the Avus.ʼ The three minutes before his crash were erased by the impact of his fall. ʻThat he survived with only slight injuries or none at all is mythical,ʼ said his son Donald. ʻHe had to lie five weeks in a Berlin hospital, several weeks of those on his belly.ʼ This was the result of severe injuries to the skin of his back caused by the forces of his ejection from the cockpit.
The organising auto club, the AVD, asked the fastidious engineer from the Glöckler racing days, Hermann Ramelow, to examine the wreckage on its behalf. He found no evidence of sudden failure or malfunction of the steering or suspension.
Some speculation centred on the right-side suspension, which had been virtually immobilised on all the Spyders to cope with the g-forces on the banking. Be that as it may, the Type 645 had worn out its welcome at Zuffenhausen. ʻThe Mickey Mouse type,ʼ wrote von Frankenberg dryly, ʻwas not subsequently recalled to life.ʼ
It did not expire without teaching the Porsche racers some useful lessons. The Type 645 had shown that frontal area could be reduced and that performance benefits were be derived as a result. Further improvement of its suspension seemed possible, even essential, especially at the front end. The oil cooler built into its front deck lid had proved its potential and would be used in the RSK.
Best of all, its frame design pointed out the new direction for the successful 550A, which Trips drove to victory at the Avus. It showed the clear advantage of the tubular space frame that later Porsche racers would use. ʻThe car led its race,ʼ Herbert Linge reflected of the 645. ʻIt couldnʼt have been a complete failure and it taught us a lot.ʼ
As for Richard von Frankenberg, he was able to attend the Porsche companyʼs Christmas festivities but only with the aid of a cane. He managed to maintain his work as the editor of Porscheʼs Christophorous magazine but, wrote his son Donald, ʻNumber 23ʼs appearance was delayed.ʼ Richard von Frankenburg would race again, and win, in 1957. CP
“LED ITS RACE… IT COULDN’T HAVE BEEN A COMPLETE FAILURE…”
Below: Co-creator Ernst Fuhrmann, in jacket and tie behind the car, showed fatherly concern for his Type 645 at Avus in September 1956. This would be the Spyder ʼs final race
Below: Engineered by Egon Forstner with the support of Ernst Fuhrmann, the Type 645 took shape in 1955 as an advanced racing Spyder built on a tubular-steel space frame
Above: The Type 645ʼs first public appearance was in practice at the Nürburgring for the May 1956 1000kilometre race, showing the right-hand filler for its cockpit-side fuel tank
Below far left: The Type 645ʼs rear suspension used double lateral links to guide its rear wheels. This was a radical advance, indeed too much so for the car ʼs trailingarm front suspension
Left centre: Instead of the head-rest inlet the engineers patented a slot across the rear deck which, tunnel tests showed, delivered a supply of air under pressure to the cooling blower
Left: The wind-tunnel model showed a planned headrest which was designed to capture air for the engine in a grille behind it. Fuhrmann and Klie were credited with the patent
Above: At the head of the field after the start, next to the 550A Spyder of Wolfgang von Trips, the smaller size of Richard von Frankenbergʼs ʻMickey Mouseʼ was evident
Below left: The Type 645ʼs front view at Solitude displayed its faired-in headlamps and the deck lid that served as a surface cooler for the engine oil, the filler for the tank behind it
Below right: When the Type 645 raced at the Avus on 16 September 1956 additional vents were cut in its tail to assist engine cooling. Dents showed the fragility of its magnesium body
Below: Amidst a stellar field of top 1.5-litre sports cars at Solitude on 22 July 1956, von Frankenberg in the Type 645 made a lively start in second gear from the front row
Above: Fortunately Richard von Frankenberg was thrown out of the Type 645 when it left the steep Avus banking. The car had a nearly-full fuel tank when it crashed. Little of value survived its terminal inferno – it is amazing nobody was killed…