New Frontier (Part Three)

Citroën didn’t have an engine worthy of their nascent 1970 flagship, but it wasn’t for the want of trying.

Image: The author

The highly unusual structure and operation of Citroën’s Bureau d’Études may have created a number of technical masterpieces, but it equally resulted in a number of serious operational drawbacks; perhaps the most serious being the lack of a cohesive singularity of purpose. Not only did the nominal Rue de Théàtre headquarters lack an effective figurehead (notably so in Lefèbvre’s wake), but the bureau itself was apparently scattered across a number of locations around Paris, each very much in effect its own personal fiefdom.

Of these, perhaps the least regarded represented the double chevron’s longest standing and most glaring weakness – engine development. This department, led by Italian former Fiat racing engine designer, Walter Becchia, seemed a largely forgotten outpost; the last meaningful programme to enter production after the cessation of hostilities in 1945 being the air-cooled flat-twin, fitted to the 2CV and its derivatives.

Two primary issues confronted Becchia: that of resource (there being little shortage of talent or ability), the other being philosophy. Under Pierre Bercot’s leadership, Citroën’s image was that of technological pioneer. But the technology the double chevron espoused (at the upper end at least) was that of advanced integrated hydraulic systems and sophisticated chassis dynamics married to progressive, aerodynamic bodywork. The engine tended to be viewed as something of an afterthought.

The roots of this lay in France’s post-war economy. As the country rebuilt, government imposed strictures upon engine capacity and power output penalised companies (and in turn, customers) who produced powerful, large-engined, upmarket models. This led to a situation where large-capacity, multi-cylinder engines were no longer built in France – the domestic industry by consequence losing the skillsets required.

While the pre-war Traction Avant had been offered in 2.8 litre six-cylinder form, the DS19, from its 1955 inception was only offered with a four-cylinder 1.9 litre engine of relatively modest output, itself a heavily revised version of the pre-war Traction unit, designed by Maurice Sainturat. A six cylinder engine had been considered for the DS, but since an in-line unit couldn’t be made to fit, a flat-six prototype engine was built [1], but it failed to meet expectations.

As the French market recovered and customers became more affluent, the demand for greater performance grew, as did Citroën’s growing belief that the DS was poorly motorisée. Capacity was progressively upped and a fuel injection system developed, but Bercot wanted more; one notable sidebar to the S-Vehicle programme being a powerplant suitable for a flagship DS model.

Not that Becchia and his engineers had been idle throughout this period, designing and building any number of prototype engines, including units of both V6 and V8 configuration; the Italian engineer conclusively favouring the latter, described by chronicler, Jan Norbye as being “light, compact and powerful“. Both of these units were fitted to Jacques Né’s nascent S-Vehicle prototypes, as was a more powerful version of the DS in-line four.

Dubbed 15N, the 1987 cc unit employed twin overhead camshafts and a 16-valve cylinder head. Fitted with four carburettors and mounted into a standard DS, it went like the proverbial scalded Goddess. The second official S-Vehicle prototype was also fitted with this engine, by 1966 developing around 130 bhp and said to be the first Citroën vehicle ever to exceed 200 km/h.

According to Norbye’s account, it appears that Bercot was impressed, decreeing that it be productionised immediately, at least until he was faced with the costs of doing so. However, in its wake, elements amid Citroën’s management would allegedly became gripped by the idea of a cheaper, four-cylinder version of the forthcoming flagship, a matter which would impact notably upon its eventual design [2].

Meanwhile, Bercot concluded that an external supplier would be required to provide Citroën with a suitable powerplant, approaching a number of carmakers, including (it’s been suggested), Alfa Romeo, BMW and Saab. Approaching Tridente owner, Adolfo Orsi in 1966, the request was filtered through to technical supremo Giulio Alfieri at Viale Ciro Menotti, who despite not having being furnished with the name of the potential French client, immediately discerned its identity from the highly specific nature of the brief.

Bercot had stipulated a lightweight alloy engine of around 2.5 litres capacity, demanding a V6 layout, for compactness. It is at this point that one of the most durable misconceptions around the SM’s Maserati V6 engine arises. The common orthodoxy suggests that the engine supplied to Citroën was derived from an existing Maserati V8 unit. Not so fast.

When the request from Paris came through, Alfieri took a pre-existing 4.2 litre 90° V8 unit from his workshops, and by effectively slicing two cylinders from the block, fashioned a 2.9 litre prototype engine which effectively met the brief. [3] Within weeks, it was in the Rue de Théàtre. Highly impressed by the speed in which the Italians could work, an order for a number of additional engines was fulfilled, quickly fitted to Né’s S-prototypes. Bercot wasted no time, negotiating an engine supply contract before Maserati’s competitors had even got off the blocks.

However, while the subsequent production engine may have shared this unit’s included angle, it was in fact, new from the ground-up and designed specifically to Citroën’s specifications. The 2.7 litre capacity was arrived at to ensure an adequate power/economy balance coupled to the fact that it would lie within the French government restrictions on engine swept volume, above which punitive taxes would apply.

Bercot stipulated the engine must be no larger in physical size than the existing in-line four, that it should be as light as possible and must develop over 170 bhp. Alfieri chose to retain a 90° included angle, firstly because it aided packaging, both within the engine bay and around the engine itself and because it was a layout familiar to him. However, this was not an ideal included angle for a six cylinder engine, most V6’s employing a 60° vee, for sound technical reasons.

Alfieri’s production SM unit. (c)

The prototype Maserati engine was fitted to a DS, but was found to be too powerful, resulting in what former Citroën executive, Jacques Fleury described as “precarious, unsafe handling“, and since it was decided that it would not be possible to successfully market a DS at the vastly higher price required, the S-Vehicle programme gained further traction at board level, quickly becoming a production inevitability.

For Maserati, Citroën’s interest was fortuitous. Facing significant costs in meeting upcoming US impact and emissions regulations, and with an ageing product line, the necessity for collaboration became glaringly apparent. For Pierre Bercot, it also went deeper than a simple engine supply arrangement. Why this was to become so, will form the basis for the next chapter of our story.

More on the SM here.

Sources and references:

Robert Opron L’Automobile et L’Art: Peter I Pijlman
Citroën SM : Jan P. Norbye
Citroën SM : Brian Long
Sa Majesté – Citroën SM : Peter I Pijlman/ Brian Cass
Citroën SM – Accidental Death of an Icon : Stuart Ager
André Citroen – Engineer, Explorer, Entrepreneur – John Reynolds
Maserati- The Citroën Years 1968-1975 : Marc Sonnery

[1] Source: Marc Sonnery.
[2]  There were two flat-six proposals, employing both air and water cooling. Speaking with author, Marc Sonnery, Jacques Fleury intimated that the flat-six “was not a workable solution“.
[3] It remains unclear as to how far discussions got with other carmakers, but Alfieri’s initial cut down prototype V8-derived prototype still exists in an Italian Maserati collection. Source: Marc Sonnery.

Author: Eóin Doyle

Co-Founder. Editor. Content Provider.

28 thoughts on “New Frontier (Part Three)”

  1. I was appalled when the brilliant futuristic DS was unveiled that it had a prewar 1911cc engine.
    But the fiscal CV is still a thing in France, including for insurance, yet another way France traps itself into being a century behind the rest of the world. I expect they have an idiom for “shooting itself in the foot”.

  2. V6 engines with 60 degree bank angle are inherently better balanced than those with 90 degrees but they have a number of disadvantages, too.
    Such an engine is taller and above all there’s very little room in the V for inlet plumbing. That’s not so much of a problem if the cylinder heads are slim with parallel valves (Ford Essex or Cologne V6) or the valves are set at a V but the V is in line with the crankshaft, making it less of a crossflow but an S-flow design (Aurelia, Flaminia). As soon as the heads get wider because the valves are in a normal V or there are double cams there’s next to no room inside the V between the heads. Lamborghini’s 60 degree V12 (and some versions of Ferrari’s DOHC V12) didn’t run its inlet tract between the camshafts for nothing and the Dino V6 uses a 65 degree bank angle (with crankpins stepped by 115 degrees) and unequal length valve stems with highly asymmetrical heads to create room for the inlet tract.
    Therefore a V6 with 90 degrees between its banks isn’t so unusual. The PRV did it, all Audi V6s have 90 degrees (and a balancer shaft for the diesels), Benz’ V6 had 90 degrees (with stepped crankpins and the Maserati C114 needed room for its three huge twin choke Weber 42 DCNF2 carbs so the choice of a 90 degree angle wasn’t unexpected.

  3. Understood the V8 prototype engine carried over the architecture of the DS engine, similar to the infamous planned pre-war Traction Avant V8 (albeit the former at a smaller displacement compared to the latter). Yet interested to know if the same also applied for the V6 engine as well as whether it featured a 60-degree or 90-degree angle, also find it curious it put out 140 hp compared to the 130 hp in the V8.

    The 15N Twin-Cam 16-valve version of the DS engine sounds like one of those variants that should have been entered into production (costs notwithstanding), the likes of Alfa Romeo, Saab and others providing a suitable V6+ engine are a fascinating area to explore though would argue there was scope for Citroen and Alfa Romeo to collaborate at the lower end of the range.

    Engineers at Comotor partner NSU also apparently looked at a flat-6 engine for what became the Ro80, which makes one think whether that was another area both companies should have collaborated on (possibly in place of the Wankel engine).

    Know the first flat-6 prototype engine was not a success when tested, yet little detail exists about the later flat-4/6 engines planned for what became the CX via Citroen Project L both their exact origins as well as if they were from an external supplier or in-house.

    1. Walter Becchia had a BMW motorcycle – whether he bought it because he was a fan of flat engines or whether this bike made him a fan of such engines is unknown. At least he liked these engines and therefore you find lots of engines in boxer configuration, either air or water cooled and designed by him. The engine originally intended for the DS was an air cooled 1.8 litre flat six that didn’t deliver enough power.
      Daniel Puiboube quotes the DOHC engine for the DS as having three valves per cylinder and being transversely mounted behind the front axle.

    2. Apparently the story was the Citroën engineers and designers had used motorcycle engines for their prototypes, with Walter Becchia being inspired by the twin cylinder boxer BMW R12 owned by Citroën stylist Flaminio Bertoni.

      The 2CV engine did very well yet would it have benefited from some of the features from other Flat-Twins such as the BMW R67-derived BMW 600/700, Erich Ledwinka’s Puch Flat-Twin, Loius Delagarde’s Panhard Flat-Twin or Juan Ramirez Montepeo’s BMW style Orix 610 Flat-Twin?

      Know BMW Motorcycle Flat-Twin engines conversions in 2CVs and related cars by the likes of Sparrow Automotive is a thing, with one or possibly two water-cooled Flat-Twin engines being used in the odd Citroen ECO 2000 concept though of the view it could have a received a shade more development then it did (even Project F envisioned a 2CV-derived 750cc Flat-Twin).

      The Rosengart Ariette-based Rosengart Sagaie was another that was to be powered by a BMW motorcycle derived Flat-Twin from CEMEC in place of the Austin Seven-based engine.

      In some respects it is little surprise the originally envisaged air-cooled 1.8 Flat-Six planned for the DS turned out the way it did, even Porsche was said to have had a difficult time developing a suitable Flat-Six for what eventually became the Porsche 911.

      It is the later Flat-Six engine projects that are of particular interest, at the moment only know Citroen developed a water-cooled 95 hp 1654cc Flat-Four in Project L as well as a water-cooled 1450cc Flat-Four for the GS post-Peugeot takeover, unfortunately the information for all three is scarce and it is unknown whether any relationship exists between them.

      Curious the DOHC engine was stated to feature either 12-valves or 16-valves, compared with the other exotic alternatives available to Citroen seems to be the relatively feasible for a more accessible sporting DS below or in place of the SM.

    3. Can’t see how the decidedly pre war 2CV engine design could have been inspired by a decidedly post war BMW R67.

    4. Never said the 2CV was inspired by the BMW R67, that does not though preclude the idea of the 2CV engine could have experienced a similar level of re-working / revision as 1965 DS engine (let alone the modifications it took for the pre-war 11 Traction Avant engine to become the DS engine). There were a number of post-war Flat-Twin designs a re-worked 2CV engine could have potentially drawn inspiration from.

  4. Why would Bercot have approached Saab when their only engine was barely bigger than Citroën’s own flat-twin?

    1. Because Saab was also on the look-out for bigger engine alternatives, in the end going for a co-operation with Triumph for their slant four. Initially they were considering a V8, what would be the Stag-engine, but went for the smaller engine + turbo-charged path in the end.

    2. Thought Saab were working on their own Slant-Four project prior to being introduced to Triumph’s similar Slant-Four project via Ricardo?

      Even so the possibility of Citroen being interested in what Saab were developing at one point does open up a fascinating what-if had the joint-development between Saab and Triumph also included Citroen. The same goes with Citroen and Alfa Romeo (less so with Citroen and BMW).

  5. BMW’s R75 sidecar 750cc with telescopic front forks engine was reverse engineered for the 2CV downsized engine. The R75 has 4 forward and 1 reverse plus overdrive with a drive shaft driven sidecar wheel. A total of 8 F & 2 R not bad for 1930’s technology.
    Both boxers were Air and Oil cooled. Some 2CV prototypes were even 2 strokes.

    1. That’s a story I don’t buy. The BMW R75 engine like all BMW boxers up to the /7 series had a one piece tunnel type crankcase with bearing carriers bolted in at the front. The 2CV engine had a two piece vertically split crankcase carrying all crankshaft bearings. Inspiration maybe. Reverse engineering, no.
      2CV engines were habitually driven at rev levels no BMW motorcycle engine would have survived for any length of time.

    2. Interestingly Citroen engines did find their way into the odd motorcycle, maybe there were other motorcycle companies that made use of Citroen engines.

      Though likely doubtful, is it known whether Citroen themselves either pre or post Michelin contemplated diversifying into another industry like motorcycles?

    3. Bob – stop! You will make me want to get a motorbike and I know they are trouble. That machine looks really good and I like the way it uses a proper engine. This is not something I was aware of until now (and maybe a good thing too). It´s a pity the project did not get so far. I wonder if any remain in existence?

    4. Amongst all motorcycles with car engines like Münch TTS (guess what engine it used)

      or the Van Veen OCR with the comotor wankel engine

      the BFG was the only one that made some sense because it wasn’t excessively overweight and with its cooling fan it was ideally suited for the low speed work many police bikes suffer.
      There even is a prototype of a supposed NSU motorcycle with the 1,300cc Polo engine. Motorcycle lover Fugen-Ferdl encouraged its build but stopped the project after driving the excessively heavy and unwieldy prototype.

  6. When working for Citroen in the mid-seventies I saw two flat-six engines under a work bench in their service training centre in Nanterre in the north-west of Paris. I was told they were prototype engines for the DS. One was air-cooled with its pushrod tubes located above the cylinders rather than below as on the 2CV engines and the other was water-cooled. I later saw the same engines at the 50th anniversary celebration for the Traction model in 1984. Sadly, they had been hand-painted and were, as far as I remember, part of a low-key display at the Place de la Concorde. I’m sure they must still exist.

    On the topic of the original production DS it was not simply the engine of the Traction carried over to the DS. It had inclined valves in hemi-spherical combustion chambers cast in light-alloy against the Traction’s in-line set up in cast-iron. To maximise valve area the exhaust valves in the DS were set at a compound angle, a technical feature that another volume manufacturer adopted decades later and decided to market as their “CVH” engine standing for “compound valve hemi.” So the first DS engine wasn’t quite so outdated, it was just a shame that it ended up virtually in the passenger compartment, was never intended to be there, so got a poor reputation for refinement, a reputation it lost when fitted to the CX.

    1. Thanks for that, Paul.
      As a boy I just saw “1911cc” and assumed it was the Traction Avant undeveloped.

    2. Oh dear; please tell me they weren’t seriously considering non-crossflow cylinder heads for their flat six. (I mean, I get it, from a packaging standpoint, but no one should have been surprised if it didn’t meet performance expectations.)

  7. I don’t think anybody’s yet mentioned Citroën’s barking mad proposal for a two-stroke V4, with a supercharger driven by its own four stroke engine.

    Described thus in Citroënët:

    “In the early nineteen sixties, the Bureau d’Etudes worked on developing a V4, two-stroke, supercharged engine.

    Displacing 1.800 cm3, it was predicted to develop 120 bhp and lots of low end torque, more like that of a V8.

    The supercharger was driven by its own 200 cm3 four-stroke engine.”

    I say “barking mad”, but somebody among the commentariat will probably explain earnestly why there are several strands of sound logic in the idea.

    While checking that this engine wasn’t something I’d imagined, I also discovered a 1959 proposal for a 5CV ID, although on reading further, I discovered it was a well worked out suggestion from the magazine L’Automobile, to use the Panhard Dyna engine to bring the DS experience within the reach of the masses:

    1. Citroen should have either placed the 750-850 Panhard Flat-Twin into the Ami and Dyane (if not the later versions of the 2CV above the 602cc unit) or further develop the 2CV engine above 602-652cc to make the Panhard engine redundant.

      The 1.8 V4 two-stroke prototype engine would have been too leftfield an idea to have progressed any further.

      The idea of using the Panhard engine in the larger Citroen DS also brings to mind of the engine finding possible use in an alternate production version of the smaller Citroen C 60 prototype, where there would be less of a gap between the 850 Panhard engine and the C 60’s planned 1100-1400 Flat-Fours compared to the 850 Panhard engine and the 1911cc DS engine.

    2. As ridiculous as the 2-stroke concept seems, I can’t help but notice that 1800 cc + 200 cc = 2000 cc. Were they hoping to effectively offer the performance of a ~3.5 L V8 while avoiding Italian (and, I assume, a French) taxes on engines >2000 cc. That would have been quite a scam! (I can’t help but wonder what racing series’ rules might have been similarly circumvented. Does anyone know if the Citroën engineers ever published any SAE papers or anything about the concept??)

  8. On Citroën engines in motorcycles, there is a very tenuous, one step removed. connection with BMW insofar as the first prototype of the K589 K100 (in-line water-cooled four cylinder) used a Peugeot XA engine from a 104.

    It seems to have been n0 more than proof of concept – the A30 series engines which BMW subsequently developed had no connection whatsoever with the XA engine.

  9. I’m curious what else the author knows about the conventional wisdom that the SM engine was based on a stillborn V8 design that Maserati ultimately couldn’t afford to develop. Are we saying that’s not true, or that the stillborn V8 was already a running prototype at the time Citroën’s request for proposal was received? And I’m specifically curious about footnote #3; where does the prototype currently reside??

    1. Joe: I will endeavour to answer your questions, but it will require a return to the source material to check attributions, who said what and so forth. Please bear with me…

    2. Joe: To return to your queries regarding the genesis of the Maserati V6 as fitted to the SM. Like most areas of historical record, views on the initial source engine differ. According to journalist and writer, Jan Norbye, the initial prototype was derived from a 2965 cc V8 that had been designed in 1965, but not developed. This engine he suggested (minus two cylinders), and a “slight adjustment of the bore and stroke could easily bring that up to the stipulated 2.5 litre”.

      Ermanno Cozza on the other hand, a sixty year veteran at Maserati, who was interviewed at length by author, Marc Sonnery, told him that the day after the initial request came to technical director, Alfieri, they had taken “a new machined V8 block, cut off two cylinders, made a crank for a six, welded shut the ports that needed to be shut, and to cut a long story short, tried out the engine.”

      He went on to say, “It was a 2.9, almost a 3-litre, it had been made from a 4.2 litre… The V8 minus two was done quickly just to try in-house for experimentation and see what output it would deliver.” It would seem that this experimental engine never left Modena, and a prototype 90 degree V6, created by Ennio Ascari was made in about 40 days, which was then sent to Paris for evaluation. Cozza: “The old one [engine] had the chain at the end, the new one, the definitive SM engine had the chain in the centre”.

      This cut down V8 which was used as proof of concept by Alfieri for Bercot and his directors languished for years at Maserati in storage, but around 1998 was acquired (along with a number of other prototype engines) by the Panini family, who apparently own a museum in Modena. This particular engine is, according to Marc Sonnery, not on display there, but is, along with a number of other experimental Maserati units, in storage.

      Personally, I’d be inclined to accept Cozza’s account, which sounds a good deal more plausible – and he was in a position to know. Certainly, from the accounts of the former Maserati staff interviewed for Sonnery’s book, there is unanimity about the uniqueness of the SM power unit. It was designed specifically for the car and not only seems to have shared no commonality with any production Maserati engine, but was built in a separate area of the factory. 

      Hope this clarifies matters. 

    3. Thanks for following up! Part of my curiosity had, in fact, stemmed from the mention of the prototype engine being in “an Italian Maserati collection,” which I figured almost had to mean Panini–which I’ve visited twice (2004 and 2010) and seen no such engine on display. Of course, that begs the question–if they actually have this mythical engine (and other prototypes), why wouldn’t they display it? (If I recall correctly, the only Maserati engines on display in their museum are a ’60s inline-six, a very stock-looking fuel-injected Tipo C114 SM/Merak V6, and a thoughtfully-sectioned Biturbo V6.) Has Mr. Sonnery actually seen this engine, or is it merely an assumption that Panini acquired it from the factory (since they acquired most or all of the former corporate ‘museum’ collection in the ’90s)?

      There’s no question in my mind that the Tipo C114 has no commonality with any other production Maserati engine; my curiosity has always been about whether it may have been related to an all-new V8 that never saw the light of day on account of the fuel crisis (or other economic factors). If I understand correctly from your summary of Cozza’s account, that must be a myth, and the V8 from which the prototype was derived was the long-serving Tipo 107, with the confusion/mythology stemming from the fact that the production V6 has essentially nothing in common with the Tipo 107.

    4. ADDENDUM: I think I found it! With the clues in your article and your follow-up comment, and much creative Googling, I found this photo which would appear to confirm that the mythical ‘prototype’ V6 was little more than a cut-down Tipo 107 V8 and thus really not a prototype at all, strictly speaking:

      I don’t speak French, but I’m pretty sure that caption is saying what I’ve already observed (albeit maybe focusing more on the off-center Maserati logotype than the technical details of the engine, to confirm that it’s a cut-down V8).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.