Citroën 2cv crash test

© Copyright: J.Cats.

"That 2cv of yours is too dangerous to drive! As soon as you hit a tree with 20km/h it will be totalled".

"I know a guy that drove his 2cv into a Volvo. The Volvo had a damaged license plate and the 2cv was totalled".

These statements are often heard by people driving 2cvs. For some reason everybody thinks a Mercedes is a safe car and a 2cv isn't. (Well at least a 2cv doesn't fall over when trying to avoid certain obstacles!). Are these believes based on facts or is a 2cv perceived as a flimsy car and thus it has to do bad in case of an accident. Where are the hard facts to prove these allegations?

In a German magazine dating from 1976 I found the only crash test I have ever seen with a 2cv participating. In this test small cars were compared on front end accident safety. This is the picture on the front cover:

The test cars involved are the Citroën 2cv, Fiat 126, Fiat 127, Renault 4 Renault 5, VW beetle and VW Polo. The crash test consists of driving the 7 cars into a non deformable barrier at 40km/h. By today's standards this is outdated. The test is not very representative of everyday life because very few people drive their car into a concrete wall. Today, crash testing is done with deformable barriers. The main difference is that on modern tests, light cars will do better then on the old tests and heavy cars will perform worse. On this test and in general, the 2cv is one of the lightest vehicles around.

Here is the sequence of crash pictures:

Click here to see an animated version (95kb).

Notice that although the front end deforms badly, the rest including the passenger area deforms very little. The only visible deformation on the body is near the top of the windscreen. The side window stays open during the crash! And here are the figures:

Crash speed (km/h)40.139.540.239.840.139.939.8
Max. deceleration (G)33403640464937
Static deformation (mm)375297265335225385280
Steering column 41-4545336515
Head injury criteria (HIC)340-253249329567220
Max. deceleration head (G)85-3975638932
Chest injury criteria (SI)68-185172139196133
Max. deceleration chest (G)23-3342343928
Max. seatbelt force shoulder (kg)6762909116767931073897

Some explanation. On the Fiat 126, the seatbelt actually broke off. So on this car some data was not available. The same happened on the VW beetle. This car was delivered to the test with an automatic roll up seatbelt. This was found not to be standard and so a static seatbelt was fitted. On this static seatbelt the lock let go. For some strange reason this doesn't show up in the numbers.

The 2cv on the test is fitted with 3 point static seatbelts. From 1980 onwards automatic roll up seatbelts are standard on any 2cv.
Only on the VW Polo and the Fiat 127 did the dummy not hit it's head on the steering wheel. With automatic roll up seatbelts the risk of this happening is highly reduced and probably the test would have turned out very differently.

Now, I'm not a crash expert but I would like to interpret the results. For example, how much better is a "chest injury criteria" of 68 when compared to one of 133? Are both really low? Is 133 really high and 68 just about acceptable? Is any value above 50 fatal?

If a car was 100mm shorter after the crash, it was also 0.1m shorter. 100 looks bad, 0.1 looks neglectable. So as long as you don't know what exactly the numbers stand for, it's very hard to interpret them.

Therefore the figures are transformed into something everybody can understand. Each car is ranked on each criteria. This will remove the ability to see how much better one car is compared to the other. But since we don't know what the figures exactly are, this isn't necessarily a bad thing. This is why ranking is very common on tests nowadays. It makes things clear for everybody. Here, the best gets 1 point, worst gets 7 points, the car with the least points in the end is the winner. Minimum total is 8 points, maximum is 54 points.
Now the same results look like this:

Crash speed (km/h)2.57152.545
Max. deceleration (G)14.524.5673
Static deformation (mm)*******
Steering column intrusion (mm)37**4.54.5261
Head injury criteria (HIC)57**32461
Max. deceleration head (G)57**24361
Chest injury criteria (SI)17**54362
Max. deceleration chest (G)17**36452
Max. seatbelt force on shoulder (kg)1.57**51.5364
Total # points2053.525.531.527.54619

* Static deformation is not ranked. The article states that deformation of the body is a good thing as long as it's controlled deformation. The information given doesn't say if this deformation is controlled. On one hand deformation is good because it absorbs energy which won't "reach" the driver and/or passengers. On the other hand, too much is bad because the driver could get stuck or injured. Either way, the number doesn't say if this is good or bad deformation so it's omitted.
** The seatbelt on the Fiat 126 broke off, explaining it's last place in this criteria.

Intrusion of the gearbox/front axle on the 2cv


The Fiat 126 and VW beetle have come off worst. Both had safety belt failures and they ended up last on the test. The Fiat 127, R4 and R5 are average, they are not consistently last on each criteria although especially the R5 doesn't score good on any criteria either. The 2cv and VW Polo do quite well. The Polo is the better car especially when comparing the absolute figures.

One thing is clear, the 2cv certainly isn't the most dangerous car here.

But to say the 2cv is a safe car... Please see below:
Real Media 362kb
MPEG 2411kb
7.5 ton truck driving at 70km/h.

Citroën 2cv Index
Cats Citroën Net Index