Because the car is going to look something like a Barbot, the frame rails are 10″ (250mm) shorter than the originals. The overall length of each rail is  130″.

The front and rear of the frame have tuck-ups for ground clearance. These I created by:

• cutting a 1/2″ strip off each end of the frame: one 31″ for the front of the car, and one 37″ for the rear of the car. To get a straight cut, I used a piece of 1/2″ box tubing clamped to the rails as a guide.  This was a laborious and dirty process.   For anyone trying this, use ear protection and a dust mask. Some sites recommend the use of a jig saw.  Don’t bother.  They wander too much even with a guide in place.
• cutting the 1/2″ strips loose from the rails and bevelling the cut edges to get good weld penetration.
• removing a wedge out of each end of the rails
• re-welding the 1/2″ strips to the rails with my MIG welder, and grinding the welds.

There was some minor horizontal warping of the rails, but nothing that cannot be pulled back into place when the cross-members are welded in.

.

Given the cost, and a small pocket book, that leaves building a new chassis and there are two ways to do that: by manually folding a new one out of sheet steel or by building one out of rectangular tubing.

So, how thick should the material be?  A bit of research shows that the original 2CV chassis metal was 0.8 mm medium carbon steel, the Ami-6 and Mehari used 1.2 mm medium carbon steel, and many of the after-market frames use 3mm mild steel, some of it galvanized.  Given that medium carbon is not readily available here and rusts like crazy because of that, mild steel is the order of the day and I would rather err on the side of strength, for safety sake. So 3mm should do it (0.120″ or about 11 guage USS).

I do not have access to a sheetl metal brake that will bend 11 guage and manually bending 11 guage is over MDF forms is out of the question. The MDF will not stand up. 18 guage is about the limit for that. Besides, building all the form is a pile of work. So rectangular tubing is the way to go.  Metal suppliers here can provide 2″ by 4″ x ,125 cold rolled but nothing thinner.  The penalty of using this ubing is weight.  The original chassis weighs about 60 lbs., but the replacement rails along weigh about 1oo lb.  Yikes!
Still there are ways to lighten this considerably without sacrificing strength and I will cover all that off in a later blog entry.

As part of the restoration, I removed the louvered engine protection panel under the frame and saw something I should not see.  The inner flanges of the frame had rusted away and the web had pinholes in it from rust.   Then I noticed that the two frame rails had a third unexpected piece of rusty metal running along them where there should only be two layers of metal.. So I called Bob and we examined the chassis closely.  The extra piece of metal turned out to be a previous repair which had also started to rust through.

So we also decided to use a mirror and a flashlight to look under the new piece of bottom skin that had been welded to the front platform.  We noticed that the new skin was 16 AWG, which is a rather heavy replacement  but we could also see that the flange on the lower part of the internal frame was rusted completely away and had not been repaired before the new thick skin had been welded on.

To repair all this means cutting off the front legs of the frame and forming new ones.  To get them into place would also mean cutting off the thick metal skin which had been added to the underside of the platform.  Because this has been welded all around the periphery, removing it would mean hours of grinding and would further damage flanges of the frame rails So, hours of grinding, rebuilding the flanges of the frame within the chassis, building new front legs – and the frame would still be full of repairs and therefore possibly weak.

So, after several days to mull things over, I have decided that a new frame is in order.  It is the only safe thing to do.

Bruce Grant took this photos in Rue du Chevreuil in Brussels.  However, he did not know the whole back-story.  For that, I contacted our favourite resident of Brussels, Étienne Musslin, who also happens to be the Editor-in-Chief of the French magazine Planète 2CV.  Étienne replied:

“This poster is located in Place du Jeu de Balle in Brussels (where there is a Flea Market every day).  I have no idea how it got there, but it originally was from the 1998 ICCCR which was held at Chevetogne.  It was located at the entrance to a display of Citroëns that were specific to the Belgian market.”

If you look carefully at the poster, first you notice that it is drawn very much in the style of the covers of the Tintin books by Hergé (who was Belgian).  Also, in the background you can see two famous tourist sites from Brussels: the Atomium and the statue of the Manneken Pis.

Sadly, it looks like the poster is slowly degrading, since I found this image on the web where it looks like it was a little fresher.

But suppose I had a billion dollars and lots of time to build exactly what I want, what I really, really want.  Would it be an exotic super car fabricated out of unobtanium with 700 hp capable of 300 kmh? Nope, it would be a simple light, efficient, practical car – very much a modern day 2 CV. It would have many of the features designed into the original 2CV but with improvements to reduce its shortcomings such as rust and noise.

I thought about updating the original 2CV spec to “able to carry a two-four up Carling avenue in early spring without breaking any” but that really doesn’t quite evoke the pleasant pastoral image of carrying a basket of eggs across a ploughed field.

I’d start with a platform chassis like that of the deuche because of its high strength to weight. However, I would make it out of corrosion resistant materials to avoid 2CV vanishing frame syndrome.  Unlike the 2CV, I would use some of the space between the upper and lower skins for things like the battery and the fuel tank.  The platform would be much wider to form the entire floor of the passenger compartment to help with side impacts.   

Just like the 2CV, the car would have trailing rear suspension arms and leading link front arms attached close together to reduce the amount of material required in the platform and therefore keep the weight down.

I would also use the horizontal spring and damper arrangement in the 2 CV but with a few changes.  First, the front leading link suspension would have two parallel arms in a configuration that prevents a change in caster as the suspension compresses over a bump. That should take care of the heavy steering that can sometimes occur when braking in a sharp turn.  It would also provide an opportunity to move the steering rack into the space between the upper and lower arms where it would be much easier to service. Second, I would select progressive gas dampers. Third, I’d add a front sway bar like the one in the AMI 6 to control lean in the corners and keep the tire patches a little flatter on the road.

Speaking of tires and tire patches, there would be no wide low-profile tires on the car. They add a harsh ride, are hard to keep balanced, expensive to replace, and tend to skitter on snow and ice instead of cutting through.  Give me tall skinnies any day.  I’d rather keep the center of gravity low and the weight of the car down to maintain handling than adding wider rubber and rims.

I would also keep the flat twin because of its low weight and low vibration, but again with changes. To extend valve and seat life with the lean mixtures that go along with low HCO emissions, and to keep engine noise down. I would reluctantly opt for water cooling. To avoid a drive belt for the water pump, I’d screw it directly to the front of the crankshaft. 

To further quell engine noise, I’d dump the pushrods and opt for overhead cams and finger-operated valves with hydraulic fulcrums.  I’d use a lightweight belt to drive the cams provided the engine could be built as a non-interference engine.  Failing the ability to do the non-interference thing, I’d drive the cams with a chain. Direct injection would allow a higher compression ratio. Then by increasing engine size to about 800 cc it should be possible to up the power but maintain fuel economy. Although I worry about adding the complexity of direct injection, I only have to look at a carb to realize that adding direct injection may actually not add any more complexity at all. Naturally, a slightly offset crank and lighter pistons would be used to keep engine friction at bay. As for 4 valves, they should not be necessary to allow the engine to breath.  The deuche’s 602 engine can happily wind to 7k with only two of them, and all that with very little valve overlap.

Instead of an alternator and the lossy drive belt, I would install rare-earth magnets around the periphery of the flywheel which would pass by coils installed in the bell-housing to generate electricity.  With a bit of savvy it might be possible to use the coils to start the engine, like the dyna-starters of the past and therefore get rid of the starter too.  A distributor-less lost-spark electronic ignition and plasma plugs would complete the picture. Turbocharging would be unwelcome because of the complexity and cost.

Now for utter heresy! Because the engine would be equipped with direct injection, the long intake manifold with the carb perched on top of it would be superfluous making the engine much thinner. By rotating the intake ports to the front of the cylinder heads and the exhaust ports to the rear the engine would be thinner still. This would allow a relocation of the engine to the rear of the car inside the platform and make the car rear-wheel drive.  Moving the engine to the rear means a very much shorter exhaust with less opportunity to radiate noise and less weight.  Unlike the VW and Porsche, I’d keep as much weight ahead of the rear axle as possible to reduce the bugbear of trailing throttle oversteer.

The heater core would be at the rear of the car along with the heater blower to avoid long heavy and possibly leaky heater hoses full of water. I have a long memory of how long it took to get heat to the windshield of the VW. If the ducts were smooth and made of a low-specific heat plastic instead of steel, airflow would be good and it would not be necessary to heat up all that metal before the heat arrived. The Renault Dauphine used a similar strategy with remarkable success. That little car had a great heater that worked well even in the frigid city of Saskatoon.

Access to the engine would be through an insulated hatch in the rear of the platform. Undoing quick disconnects along with removal of the bumper would allow you to roll the engine, transmission, and all of their ancillaries out the back of the car for heavy-duty maintenance.

The spare tire would move to the front of the car and would be used to provide some impact resistance.  The tire would be placed in such a way that it would slide under the car in a collision.

The seats would be split benches on 4-inch legs attached at the edges of the platform. This would allow you to slide boards under the seats the full length of the car.  However, I would put a guard plate ahead of the driver’s seat so boards would not slide under the pedals. For big loads, the rear seat could be removed entirely by unscrewing 4 big fasteners that attach it to the platform

The roll-back roof and big vents would be a must to avoid the need for air conditioning. However, I would design the vents to deflect and drain water away and prevent passengers getting wet in a rainstorm like they do in the deuche.  I’d also add a second roll roof below the first.  This one would be a 60% screen to allow circulation without having the sun burn a hole in the top of your head.

Rear brakes would be inboard and the front would be a fully shielded but vented design as found on aircraft to keep guck and water out of the brakes and prolong brake life.

The poor old deuche was actually surprisingly good in a front impact because of the strong passenger platform and door frame hoops, but not so good in side impacts.  This would have to be corrected with reinforced doors.

Another heresy – no frontal air bags. As an ex-firefighter I can affirm that a direct front collision is rare.  Most fatal and injurious collisions occur from 15 to 95 degrees off the long axis of the car.  Second, occupants always hit the inside of the car more than once, by which time the air bags are on the way to deflation.  The biggest problem is associated with the 3-point harness.  In a collision, the driver’s unrestrained right shoulder lets his body rotate so his right temple whacks the windshield.  I can still recall the large number of windshields with a big bump at the center where the head hit. So, 4-point harnesses for this car.  God, Mr and Mrs Uckumslutch would hate this one!   The next danger area is where the side of the head hits the b-pillar or the side window.  I don’t know of anything else except the use of side air-bag to help with this problem.

Now, about the windows. No expensive window crank mechanisms to interfere with the reinforcements of the doors and add complexity. The windows would tip up like those in the 2CV (but with a better clip to prevent bashed elbows) or, God help us, sliding windows a la Mini.

Because the car has a platform, it would no longer be necessary to have high door sills. This means that the entire floor could be flat and you could therefore wash slush and crud out of the car with a hose, especially if the floor mats were dimpled rubber instead of carpet.

The gearbox would have an extra cog in it, but the shifter would still stick out of the dash near the steering wheel, where it is easy to hand, and yes it is possible to do this even with a rear engine without a complex and spongy linkage.  Maybe a centrifugal clutch like that once offered in the 2cv would make life bearable for creeping through traffic. I think it may also be possible to incorporate a motorcycle-style constant-mesh gearbox with 2 clutches for those who absolutely insist on automatic transmission.

A feature I absolutely insist on is the addition of heating ducts to the side windows, front and back, and to the rear window.  No expensive electric heater elements to rupture.  I’d also add small ducts to each side mirror to keep them clear on frosty days. While we are on the subject of winter, the washer bottle needs to be readily accessible, heated, and able to contain more than one gallon of fluid.

When the deuche was designed it was so slow that aerodynamics could be ignored.  Not so it’s modern replacement.  A Mazda 5 shape and a smooth underbody provided by that all-important platform chassis would keep drag in check. Naturally, flush headlights and hide-away wipers would be included in the design to keep drag and wind noise down to a minimum.  The front windshield would have a deep curve to stop airflow detaching at the sides of the car and inducing turbulence and yet more wind noise.

To keep winter dirt off the rear window, I would add a deflector to the rear of the car that would direct air down across the rear window to keep it clean.  The alternative would be an expensive wiper.  To keep drag low I would allow it to stow out of the way on dry days.

So ends my daydream. Now back to harsh reality.

One of these features is a sandwich platform chassis, which has only recently appeared on cars like the Mercedes A and B class. Yes, rust is a source of frustration with the deuche implementation but it can easily be prevented with today’s modern materials and coatings. 

This platform chassis provides great strength for its weight and is structurally efficient. It is essentially a wide flange beam with plenty of load-carrying material in the upper and lower fibers.  The suspension and spring mounts at each end of the beam are located very close together (approx 6’6”) unlike many other cars where the mounts are located way out in the wheel wells. Because the beam is so short, it can be much lighter to carry the working loads.  To boot, the suspension design routes the spring and shock loads horizontally along the platform where they help to counteract passenger loads on the platform.  For example, when the car goes over a bump the lower fibers of the platform are in tension.  However, at the same time, the springs are in compression and tend to counteract the tension.

In addition, this strong platform which forms the floor of the passenger compartment helps to keep the compartment intact during a frontal collision with the overhanging front spars of the chassis collapsing to absorb impact energy.   Pierre Boulanger did not have this in mind when his team designed the 2CV, but there you go, sometimes there is serendipity. Frontal crash tests support this thesis.

Another novel feature of the deuche is the soft springs and huge suspension travel. Colin Chapman, designer of Lotus cars, proved that soft springs and long suspension travel are key ingredients for good handling. The deuche has both soft springing and incredible amounts of travel.   However, Colin also included careful and progressive damping as one of the ingredients, which are marginal on the deuche.  Two out of three ain’t bad.

Inboard front disk brakes like those on the deuche appeared on formula 1 cars for a few years. The idea was to reduce unsprung weight and improve adhesion on bumpy surfaces. These disappeared after a while because of the difficulty of cooling the disks under racing conditions – definitely not a problem with the slow deuche under daily driving conditions.  In addition to the low unsprung weight, the inboard disks obviate the need for flexible hoses which can contribute to a mushy pedal feeling.  But the big advantage is that the front brakes are out of the wheel spray and guck which contributes to longer brake life.

The deuche was designed before air conditioning. To keep occupants cool, the deuche has a wonderful vent below the windshield coupled with a roll-up roof.  I have ridden in a deuche on a hot summer day with the vent open and the roof rolled back and been very comfortable. So, why eat up valuable gas to run an air conditioner?  The only improvement I might make would be to install a 60% shade screen like they do on lettuce farms to let the air pass through without having the sun burn a hole in the top of your head.

Now about that 2-cylinder engine.  Car manufacturers such as FIAT and Hyundai are moving to fewer cylinders and smaller engines to reduce weight, friction losses,and improve gas mileage while cruising. In fact, FIAT’s new “500” can be purchased with a 2-cylinder engine.  The 2 CV has 2 cylinders but, unlike the FIAT, requires no balance shaft to reduce vibration.  The difference with the modern cars is that they achieve much more peak horsepower for acceleration by means of turbochargers.  I am unconvinced that the extra cost and complexity of the turbocharger is worth the increased cost and reliability concerns introduced by a turbocharger.  Also, how much acceleration do you really need?

The car’s light weight and small engine contribute to excellent fuel economy that surpasses that of many modern cars, and all from an engine design that is almost 60 years old.

The rear opening and flat floor let you slide long boards into the car and close the trunk lid.  Try that with a small modern car.  That feature is only usually available in much larger mini-vans and SUV’s.

So lets hear it for the intelligent features built into the deuche and hope that modern car designs will use them.

Seriously, this is the famous Matt of Matt’s Soft Tops.  And that actually is Ian’s two-tone red and white roof being sewn.

All four stories are now available.  Enjoy!

Here is what the frame looks like.  Under the skin, there are two side rails and an internal X frame.

At first glance the lower skin of the frame looked good but a closer look showed pinholes all along the x-frame that lies beneath the skin , a rotten rear flange, and pieces rusted off the edges of the frame.  Because the lower skin was in such poor condition  I was afraid of what lurked underneath.  So I opted to remove the rear half of the lower skin.  The standard practice for this is to locate the spot welds and drill them out.  But  no luck here because the welds were not visible even after wire-brushing the lower skin.  Usually, this shows their location. Also  the edges of the lower skin had been welded as part of a previous repair.  So this task turned into to very fine surgery 

I have to say that whoever did the welds on the previous repair is a damned fine welder.  The welds are of the highest quality without so much as even one blemish.

So, first I cut triangles out of the lower skin  so I could locate the x-frame and the frame rails beneath the lower skin. I goofed in a couple of spots and nicked the rails and the x-frame, but those spots were rotten and had to be replaced anyway.  So once the interior bits were located, I drew them on the outside of the skin and cut out more of the skin.  Then there was only one way to remove what remained welded to the innards of the frame- that was to grind away the skin inch by inch until I had removed all of the skin stuck to the frame.  What a job!  Two wheels for the angle-grinder and 7 hours later, the skin was off.

With the lower skin off, I found the rear flange beyond repair, parts of the flange on the x-frame were pooched, as were 40% of edges of the frame, and the floor of the box section at the front of the frame was toast. However the top skin of the frame and the upper surfaces of the frame rails and internal x-frame are just fine.

At this point, I felt like giving up.  Option 1 was to buy a new frame.  But buying and shipping a frame to Canada is beyond my means.  After paying for shipping, customs duties, the total is over $2,500 Canadian.  Option 2 was to replicate the frame but this requires access to an 8-foot sheet metal break and a similar sized shear. Since I have access to neither it was Option 3 – fix the frame or bust.

First, I cut and welded into place most of the rear flange.  Here is the new one  About 3/4 of the old flange was pinholed or missing, so most of what you see is new.  This entire work requires a very very delicate hand with the MIG.
The key is  low voltage, low feed, angling the gun steeply and stitch welding very slowly, allowing the weld to cool between each stitch. In places, the metal of the vertical section is now paper-thin, so I did blow through a coupe of times and had to weld holes shut.

The dogleg that joins the frame rails to the x-frame inside the chassis had a chunk missing out of it.  Here is the new dogleg welded into place.

Part of the flanges of the inner x-frame were totally done, so I cut 14″ of new flanges and welded them in place.  Here is the result.

 The next thing was to reskin the underside of the front box section.  This is solid now, but I am not happy with the result.  The upper surface of the skin has lumps along the edge and when I weld the new skin over it, it may ripple.  At this point, I am contemplating ripping the new skin off, bending up a new u-shaped hunk of 18 AWG, then welding it into place. Here is the current repair to the box section.

Stay tuned there is much more about the repairs to come.