Jesse Alexander driving Fangio W196
Jesse Alexander with Fangio Mercedes-Benz W196
Impressions of an Icon
The rare event of auctioning a post-war Mercedes-Benz racing car reminded one enthusiast that he had driven it when it was still quite new. He reflects on some details of the W196 design in the context of his trial on the track.
by Karl Ludvigsen
When news broke that Bonhams was auctioning an icon of Grand Prix racing, conceived sixty year ago, I was intrigued. The Mercedes-Benz W196 had long fascinated me. Of its complexity and ingenuity Laurence Pomeroy, Jr. famously wrote, ‘One may say figuratively that if the BRM be likened to a typewriter, the four cylinder Ferrari is the equivalent of an abacus and the Untertürkheim cars rate as an electronic calculating machine.’ Adding all-enveloping bodywork on their first appearance at Reims on 4 July 1954 only enhanced their all-conquering mystique.
I had written about these cars at the time and soon thereafter in my history of Mercedes-Benz racing, seeing them for the first time in the Museum in 1958. When M-B of America announced a press trip to Stuttgart for late summer of 1961 to flaunt a new automatic transmission, I urged them to include a drive in a W196 for this elite group of editors. I already knew that I was leaving
Car and Driver to take up a post at General Motors, so getting a W196 drive was a major box to tick. I was successful, which led to my being the first of our group to step into an open-wheeled W196 at the works test track on 14 August 1961.
Chassis number 00005/54, I knew, had been selected to be chromed and lacquered for display in the Daimler-Benz Museum. The chassis I drove was 00006/54, handled by Juan Fangio to win the Grands Prix of Germany and Switzerland in 1954. Hans Herrmann drove it to fourth at Monza and retired it in the Spanish
Epreueve at Barcelona with fuel-pump ailments. Updated for 1955 with the latest gubbins, including a brake-assist servo, Karl Kling used it to place second in the Italian G.P. Now, having escaped captivity, it is to be auctioned by Bonhams at Goodwood.
Before driving the W196 I reflected on some aspects of its unusual design. In aid of ample power for fast-track superiority, Mercedes-Benz built a fuel-injected straight-eight that they knew would be heavy and complex. At the time, though, they said that a vee-eight layout would have been even heavier. This view was greeted with scepticism by some designers. After all, they said, the straight engine’s extra crank and bottom-end weight—especially with the W196’s roller-bearing construction—should overshadow the vee engine’s extra camshaft-drive gears. Did it really make a difference? I decided to compare weights.
Though it was built around two intricate welded sheet-steel cylinder blocks, the 1954 engine was primarily of a high-silicon aluminium alloy (silumin) and weighed 451 pounds. To reduce weight for 1955 many low-stress parts like covers and cam housings were cast of magnesium, the result being a reduction to 429 pounds. A crankcase of magnesium was experimented with but durability testing showed that it could not hold its dimensions well enough.
The all-silumin Lancia D50 V-8 of the same era weighed 382 pounds, while the experimental Coventry-Climax ‘Godiva’ V-8 weighed 340 pounds—probably without some accessories. One of the most complex 2½-litre V-8s, the Brooke-Weston Speed Engines Limited design, weighed about 400 pounds. As a matter of interest the Mercedes-Benz 1½-litre W165
V-8 of 1939 weighed 442 pounds, including the added burden of a Roots blower.
An outlier that made a case for straight-eights was the weight of only 330 pounds for Gordini’s ultimate Formula 1 engine. This was such a paragon of simplicity, however, that it may not be fair to mention it. On balance it is likely that an unblown 2½-litre V-8 Mercedes-Benz engine might well have been lighter. In fact Rudolf Uhlenhaut told me that he was considering such an engine for the W196’s successor, along with four-wheel drive.
The W196’s powerplant was bedded down using a transverse mounting plate, as in a contemporary Indianapolis Offy-powered Kurtis, in a highly sophisticated chassis. Fully calculated for stress, its space frame consisted chiefly of tubes about an inch in diameter. Though it weighed only 79 pounds, it was considerably stiffer in torsion than the 181-pound 300SL space frame.
A prime design goal was the lowest possible centre of gravity, hence the engine was canted over to the right at 30° to the horizontal. The use of a secondary drive shaft from the centre of the engine to the clutch allowed the designers to locate the heavy crankshaft 2.6 inches lower than the diameter of the clutch would otherwise have permitted.
Ground clearance was cut to the bare minimum at 4.1 inches. The result was a centre of gravity just 12.2 inches above the surface, which compares well with the 14 inches of the Grand Prix Scarab, which also used a lay-down engine.
‘The art of adjusting the springing and suspension of a racing car for cornering lies in selecting the point of neutral steer so the cornering power of all four tires is used to the very ultimate, with respect to the power available, and so the driver can easily reach and maintain that limit in a curve.’ So wrote Ludwig Kraus of Mercedes-Benz, who designed the chassis. ‘This,’ he said, ‘was a prime consideration during the design of our post-war 2½ -litre Formula 1 racing car, the W196.’
Kraus and Uhlenhaut felt that the best way to achieve this handling goal was to build in very slight understeer over most of the cornering range but to induce oversteer at the very limit. This was generally accomplished, though the cars did have more basic understeer than was really desired. Tweaking for the 1955 season was intended to reduce this.
Details of the design included roll centres 1.4 inches above ground in front and 6.9 inches above at the rear. Front torsion-bar spring rates varied from 61.5 to 72.5
pounds per inch depending on the circuit, while the effect of the anti-roll bar increased the rate to 111 pounds per inch for jounce of a single wheel only. At the rear the chassis tuners could select torsion bars ranging from 100 to 117 pounds per inch in stiffness. Weight distribution ranged from 47/53 front/rear with an empty tank to 40/60 with all fuel on board. The average, design distribution was 45/55. Dry weight for the variant I tested should be about 1480 pounds.
Front and rear tracks were 52.5 and 53.5 inches respectively, while wheelbases of 92.6, 87.1 and 84.8 inches were available. Changes in lengths as well as permutations of brake layouts and body shapes were effected to suit different circuits. All the 1954 cars had the long 92.6-inch wheelbase, with inboard front brakes and both open and enclosed bodies. For the beginning of the 1955 season, some of these cars were remodelled to 1955 style by fitting cleaner bodywork—though with a big bulge for the straight ram pipes of the lighter ‘55 engine—and by making many other detail changes like the addition of a brake servo. These first-year chassis are easy to detect because the top front frame crossmember still has the built-in pipe for the old curved-ram-tube inlet plenum.
Thoughts of these considerations as well as the criticisms of such as Fitch and Moss of the Porsche-synchronised five-speed transaxle were in my mind as I peered upward at the threateningly cloudy weather over the test track between the Daimler-Benz works and the Neckar River. I had rather hoped for a sunny day on the Nürburgring, but then you can’t have everything.
A cadre of blue-suited factory mechanics surrounded the W196, a dull silver object that breathes brute force rather than beauty. This is misleading, for the machine is sophisticated rather than stunning in action. After viewing the double-ended parallel- straight track I was led to a large grey-painted wooden box that I thought contained plugs or parts. The lid was flipped up to reveal six cycle-type crash hats, each in its own niche and each a different size, and six pairs of goggles. The legendary Mercedes preparation was front and centre.
Suitably lidded, I set about climbing aboard. The detachable wheel had already been removed; indeed the mechanics were quick to remove the wheel at every opportunity, apparently on the assumption that the driver wants to quit the car the instant it comes to rest. My feelings were just the opposite. They were going to have to drag me out of that car.
The wheel is central and the sides are very wide, so it’s a long stretch into the cockpit and essential to step on the seat first. I eased down into the very roomy cockpit and the equally roomy seat. Covered in the traditional plaid, the seat was well cushioned and canted deeply to the rear so that my knees were bent quite high and my torso leaned well back. It’s a relaxed position that has the added advantage of taking up a minimum of chassis space. My legs were spread almost a yard apart around the clutch housing and near-central drive shaft in an unique splayed stance that looks odd but feels fine.
The four-spoked wheel was locked onto its ten splines into a position that is physically near-vertical but actually feels more horizontal because the driver is leaning back so far. The grooved wooden rim has a hefty feel while at rest there’s a surprising inch of lost motion in the steering gear. Scanning the sculptured dashboard was easy because it was an axiom of Uhlenhaut to minimize the driver’s role in technical oversight. There are only three dials: oil pressure at left, 10,000 rpm tach with telltale in centre and water temperature at right. These are mainly ornamental since the engine will run long periods with no oil, the desmodromic valve gear allows wild over-revving and the coolant temperature is thermostatically controlled.
A switch on the dash sends battery current to the whirring fuel pump that prevents vapour lock in the fuel injection, while a big key under the panel at the right switches on first one magneto, then the other, then both. My only instructions were to get on with the clutch, not fooling around but engaging it fully, and during starting to hold out a small spring-loaded knob under the panel to the left. I reckon it retarded the ignition but it seemed to make little difference since the engine barked readily to life during a first-gear push-start even though I’d botched the knob routine.
My first problem in driving the W196 was to dominate this trial out of all proportion: getting the hang of the shift pattern. There are five forward speeds with first over to the left and forward. You can’t get into that side of the gate without first pressing a button on top of the knob, since first is intended for starting only, while an additional latch keeps you out of reverse. But it was the top four speeds that were baffling. In a normal four-speed box all your upshifts are made with a straight-back yank except for the 2-3 shift across the gate. On the top four Mercedes gears you shift up by shoving straight forward, and the normal 2-3 motion would jump you from second all the way to fifth!
There were lots of things to unlearn. It was consolation to know that Stirling Moss’s tendency to go from second to fifth led to the creation of an elaborate gate interlock—first used in the Mille Miglia on the 300SLR—and that notables like Fangio and Fitch were fooled by the gate at first. But you learn. A small assistance was spring-loading of the lever into the 2-3 slot. The first lesson was a spinal chill when I realised I’d ‘upshifted’ from third down into second with a straight-back motion and just barely engaged the clutch enough to send the exhaust note skyrocketing. I caught the error in time and didn’t make it again but it was never possible to forget the box entirely.
The ‘trouble’ with the Mercedes shift is that its Porsche-type synchromesh is so smooth and powerful that the lever feels the same going into any slot at any speed. It’s not like the Birdcage Maserati, for example, which gives your palm a slap and grates angrily if you’re off on your shifting speeds. The big lever moves in an arrogantly positive way in a machined, vintage-style gate with long fore-and-aft travel. In its use of synchromesh the W196 was unusual among modern G.P. cars while the gate itself definitely recalled the 1930s.
I took care to bring the clutch fully home before turning on the tap. It behaved impeccably in spite of its reputation as a weak spot in the design, especially when used in the heavier 300SLR. Its linkage has some over-centre action that eases the pressure required near the end of the travel, but it’s still a stiff clutch by any standards. To its left is a handy dead-pedal. Once I tromped on the clutch and thought, ‘That’s too stiff for the clutch. I must be on the dead-pedal.’ But it was the clutch after all.
The W196 power curve is extremely steep to a sharp peak at 290 bhp and 8,500 rpm. There’s not much torque until 5,000 rpm; the first torque peak is at 6,400 and there’s a ‘second wind’ at 8,100 rpm. This is all evident in action as on paper. The straight-eight pulls smoothly but with no sincerity to 5,000, where it suddenly comes alive with the hollow, booming roar that marks these cars—like two Porsche Spyders sounding off at once. It’s not deafening but it does induce a ringing in the right ear. I only took the engine to about 7,000, enough to get the full, sensual surge of that first leap of torque. It moves out—and right now.
Engine throttle response was so swift as to be startling at first. A jab at the long-travel pedal for a downshift brings instant, eager response that overdoes it until you learn to reduce the dosage to suit the needs. This is a key advantage of fuel injection on a racing engine, since throttle response can be a problem with multiple big-venturi carburettor throats. I used full-bore to 7,000 in third, fourth and fifth around the test track, which felt like a maximum of perhaps 120 mph with the low gearing and small back tires fitted to this much-flogged demonstrator. It used up the short straights very quickly indeed—which brings us to the brakes.
When I started out, the chassis—if not the engine—was cold. When I applied the brakes the first time I was sure I was headed for the bulrushes. They began to warm up, however, and soon were answering the pedal with smooth, strong deceleration. Braking is aided by a hydraulic servo developed by ATE for truck use, fitted for the first time to the SLR in 1955. It runs from a little vane pump driven from the back of the gearbox that wastes half a horsepower at peak speed.
It was only possible to probe the handling on two turns, one long, tight right and one fast left-hander, but it was obvious at once that this was an amazingly easy and safe car to drive. It sometimes seemed that it wasn’t just a case of being ‘forgiving,’ of allowing you to rescue yourself with ease; the W196 felt quite capable of handling the required rescue all by itself. I felt this biddability when I entered the tighter, closing turn a hair fast and the tail swung smoothly, automatically out, saying, ‘Okay, correct like I told you and you’re in business.’ (in German, of course)
At constant cornering speed the W196 gently understeers, but as you apply power and increase speed you feel the tail shifting easily outward toward balanced neutrality. More power carries the attitude over to oversteer in an unbroken progression. The whole handling spectrum of this car is available at all times to the driver. Feel of the steering is light but still highly informative; it has a way of instilling confidence. At 2½ turns lock to lock it is fast enough to cope with the car’s very progressive motions—a product of a high polar moment of inertia.
Maurice Smith, then editor of
The Autocar, also tried the W196. Several years earlier Maurice had driven Stirling Moss’s 250F Maserati, so he was able to make interesting comparisons of these mid-1950s rivals. He reported that the Mercedes felt far more stable than the 250F, which tended to leap and jump about on its hard springing and needed constant correction of its very quick steering. Maurice also said the ride of the Mercedes was far superior. It certainly was comfortable, for a racing car, and the low centre of gravity and relatively high roll centres assured that lean in corners was minimal.
You look almost through the shallow windshield, over a bonnet that’s high in relation to your low, reclining position. There was relatively little sense of driving an open-wheeled car, since the left front wheel is only partially visible and the right one is almost completely hidden by the high ram-plenum housing. The W196 had a curiously light, hollow feel as of a stiff shell to which the drive elements were attached. This one was obviously much-used by ranks of eager journalists, for it was a little loose both in appearance and action.
Finally one mechanic started waving flags at me and threatening to throw himself in front of the car, so I had to switch off and come to a halt back by the box of helmets. The many onlookers around the track began to disperse and the mechanics packed up their gear. As the echoes of its exhaust subsided, this W196 Mercedes again became one of the world’s most fascinating mobile museum pieces. If its new owner would like some tips on driving it, I’m available.
1 Pedants will point out that I am not being strictly correct in my use of ‘W165’ in discussing an engine. Mercedes-Benz refers to car types with a ‘W’ prefix, for Wagen, and to engines with an ‘M’ prefix for Motor. Thus W196 refers to the whole car and M196 to the engine only. I have used ‘W’ throughout for the sale of simplicity.
2 In the 1960s Ferrari did essentially the same thing with his outboard-clutch transmission design.
3 Here readers may remember the parody of this unusual seating position by Peter Ustinov in the ‘Schnorcedes’ of his Grand Prix of Gibraltar.
Juan Manuel Fangio, W196 Mercedes-Benz, Spa, Belgium, 1955
Many more Fangio images here.