Handwritten note by Paul Magès on the characteristics and technology of the hydropneumatic shock absorber
A handwritten study where Paul Magès defines the ideal curve of a hydropneumatic shock absorber by speed zones, details the technological choices (sintered body, crimping) and quantifies the economic advantages of the H suspension compared to the competition.

Understanding this document
In this document
I. The choice of the characteristic curve of a shock absorber
Magès first defines a reference benchmark, the "average rate" of damping: that of a shock absorber with a perfectly linear characteristic, set at 20% of critical damping at full load p. 6.
He then analyses the actual curve in three vertical wheel centre speed zones p. 6-7:
- 0 to 0.3 m/s: a zone considered "fundamental for the stability-comfort compromise". Damping that is too weak makes the car "soft" in roll and pitch, harming stability and promoting "car sickness"; damping that is too strong causes a phenomenon Magès calls the "balé", resonance of the body on the tyres between 3 and 5 Hz. He notes that the setting here is "very precise" (a difference of 0.05 mm in the jet diameter is noticeable) and concludes that it appears "impossible to reconcile perfect stability and good comfort" p. 6-7.
- 0.3 to 1 m/s: the most frequent working zone (wheel beats, small bumps), where the curve must be as close as possible to the theoretical straight line, with a point of concordance located between 0.8 and 1 m/s p. 7.
- 1 to 4 m/s: Magès seeks a degressive rate to limit the forces transmitted to the body on "steep-front" obstacles (potholes, ruts), regretting not knowing how to simply "clip" the curve beyond 2 m/s p. 7.
II. Shock absorber technology
Five criteria guide the design: minimum cost price, single body for all vehicles "manufactured or under study", precise adjustment in zone 1, good fatigue resistance of the valves, absence of adjustment loss during assembly p. 7-8.
The chosen solutions: sintered body (replacing the machined body), valves embedded by cut washer, crimping for tightening the valves and for fixing to the pneumatic block — this crimping guaranteeing a precision of "about 1%" p. 8-9.
Advantages of the hydropneumatic suspension
Magès lists several assets: insensitivity to oil loss unlike a classic telescopic shock absorber; almost constant comfort regardless of load, thanks to the variation in flexibility; improved behaviour on speed bumps/humps for the "G" vehicle due to larger stops; and a repaired leak return on the cylinders making any breakdown due to a defective seal "practically impossible" — a 300 km test without a seal on the front cylinders is cited p. 9.
A table with figures compares the cost of "G" suspension elements to that of FIAT 128 shock absorbers:
| Item ("G" vehicle) | Cost |
|---|---|
| Pneumatic blocks | 83.34 F |
| Pistons-cylinders-rods etc. | 118.30 F |
| Shock absorbers | 8.96 F |
| Labour | 27.88 F |
| Total | 238.38 F |
| FIAT 128 shock absorbers (comparison) | 237 F |
Context
"Vehicle D" notoriously refers to the Citroën DS, and "vehicle G" would likely correspond to the project that would become the GS, then in development at Citroën in the early 1960s — this connection is however not made explicit in the document and is based on general knowledge of the Citroën range.
The hydropneumatic suspension, developed under the direction of Paul Magès, remained one of the most iconic technical innovations of Citroën, adopted on the DS from 1955 and then extended to other models.
Archivist's perspective
This handwritten note offers rare access to the technical thinking of Magès himself: it is not limited to describing a finished solution, but exposes a reasoning in progress, with his assumed doubts ("it appears to us, today, impossible to reconcile..."). It thus seems to come from a work notebook rather than a document intended for distribution.
The numerical comparison with FIAT 128 shock absorbers p. 9 shows the engineer's constant concern to integrate technical innovation into an industrial and competitive cost logic.
Dated 1961, this study is probably contemporary with the work on the three-way valve with adjustable overlap, filed in May of the same year, and could shed light on the choices mentioned in the patent addendum on the anti-roll device, also from 1961.
In the same folder
Dating
1961
Document
4 pages
Source: « Scanner 5 août 2025 à 15.14.pdf », p. 6–9
Category
Archival note
Continuous handwritten text numbered (I. on p.6, points 4°/ 5°/ p.8, 5°/ p.9); date "1961" entered on p.6.






