Revised Summary of CitroŽn Hydraulic Fluids

Conceived and Compiled by Tony Jackson
Edited by Mark L. Bardenwerper, Sr.
Updated March, 2007

Fluids for LHS Cars
Original Fluids For LHS Cars
Alternate Fluids For LHS Cars
Specifications For Fluids Used In LHS Cars

Fluid for LHM Cars
Original Fluids For LHM Cars
Alternate Fluids For LHM Cars
Specifications For Fluids Used In LHM Cars

Late Breaking Information
Seal Compatibility
Fluid Changes and Flushing

Editors Note:
Tony Jackson first began supplying the CitroŽn virtual community with this summary several years ago. Up to this time, however, it has only been available in text form. I have chosen HTML as the ideal universal format for widespread assimilation and easy modification of this information. With the Internet as a conduit, it has become easier to track the many experiences of owners of these unusual and fascinating cars. CitroŽn lovers are a resourceful lot, given the complexity of their cars and the diversity of their geographic locations. Unlike many popular collectable cars, older hydropneumatic CitroŽns were scattered thinly into the remotest corners of every continent. Even in places where they were once frequently sighted, they now are getting more and more difficult to find. Keeping old cars in running condition is challenge enough, but these CitroŽns offer the extra complication of hydropneumatic systems. This radical approach to vehicle design is just one of the innovations that make these cars endearing and a fascination to those who prize them.

Given the increasing difficulty in finding the original specification hydraulic fluids, owners have often relied on ingenuity to devise alternate sources. The Fluid Summary is an attempt to provide an easy to modify source of information on these ongoing trials.

It must be stressed that the authors do not in any way wish to take responsibility for consequences arising from the use of any of this information. Readers are hereby informed that all of this work is fully experimental in nature, and any actions taken by the readers will be construed as being taken on their own volition. In other words USE THIS INFORMATION AT YOUR OWN RISK.

We invite the CitroŽn community to feel free to offer us information on your efforts. This document is by no means complete or final. As you can see by the following chapters, charts and articles, there are many gaps in our knowledge. Please send your contributions to the authors.

Fluids for LHS Cars

Original Fluids For LHS Cars Top
Early production cars used red colored LHS fluid. Their main hydraulic components were painted black. This fluid proved to be problematic because it was hygroscopic and anything less than frequent fluid changes resulted in corrosion and damage to the system, particularly in humid climates. LHS was apparently manufactured only by Eugene Kuhlmann in France and Deutsche Pentosin Werke in Germany. All other suppliers bought from one or other of these and then packaged it under their own labels: Castrol, Lockheed, Total, BP, Shell, Esso and Bendix. LHS originally was castor-based. In 1963, synthetic-based LHS2 supplanted the old formula. Some problems subsided, but hygroscopy was still an issue.

Everywhere except in the US production changed in 1966 to use a green-dyed mineral fluid, LHM, which did not take up water, and which has proven highly successful ever since. LHM was not compatible with the seals used in cars built previously, so the hydraulic component color was changed to green as a warning. Cars sold in the US changed to LHM during 1969 due to delays in governmental certification, so consequently, early 1969 model cars in the U.S. still use LHS. Again, the color of the hydraulic components is the tell tale; black is for LHS and green signifies that the car uses LHM.

LHM was introduced at the following serial numbers for all except U.S. export models LHM for US export models was introduced at the following serial numbers
DS19A-H 4 316 000 DS19A-H 4 330 000
DS19A-M 4 442 000 DS20-H 4 332 001
DS21-H 4 376 200 DS20-M 4 451 001
DS21-M 4 473 100 DS21-H (including cabriolet) 4 621 000
Cabriolet DS21-H 4 376 050 DS21-M (including. cabriolet) 4 490 001
Cabriolet DS21-M 4 473 020 ID19B 3 794 600
Cabriolet Chapron DS21-H 4 376 000 ID20 3 820 001
Cabriolet Chapron DS21-M 4 473 000 Break ID19FA-M 3 546 800
ID19B 3 710 001 Break ID19FHA 3 980 380
Familiale ID19FA 3 535 000 Break ID20F 3 980 501
Break, Ambulance, etc. ID19FA 3 536 000 Break ID20FH 3 985 001
Familiale ID21F 3 554 000 Break ID21F 3 561 600
Break, Ambulance, etc. 21F 3 554 500 Break ID21FH 3 575 350

Alternate Fluids For LHS Cars Top
Lockheed 70R1 grade fluid was initially specified by the factory for cars sold in the US. By 1966 the factory approved any fluid to SAE 70R3, including Mobil Super HD, Delco Super 11, Lockheed Wagner 21B, Mopar Hi Temp. Specification SAE 70R3 was later replaced with DOT 3.

Standard glycol-based DOT 3, DOT 4 or DOT 5.1 brake fluid works but allegedly causes problems. It is very much less viscous, especially at high temperatures, and has much poorer lubricant properties than LHS. Rolls Royce used DOT 3 in bottles labeled RR 363 for their application of CitroŽn patents, but theirs was a much more restricted use of hydraulics. DOT 5.1 is a newer category for glycol/glycol ester based fluids that essentially matches the silicone based DOT 5 fluid for dry boiling point. However, silicone fluid should not be mixed with glycol based fluids as they are incompatible and unmiscible.

The main difference between DOT 3,  4 and 5.1 ratings has to with the dry and wet boiling temperatures of the fluids. The designation "heavy duty" that one sees attached to these ratings means that the fluid in question has a higher rating for the wet/dry boiling point that the specification for that category but not high enough to elevate it to the next category.

Some believe that silicone based DOT 5 fluid would be an improvement as it does not take up water, but I think it would be prudent to replace all seals before doing so. In the light of experience in cars with non-powered hydraulic brakes and clutches, where seals have already deteriorated (probably due to moisture in the glycol fluid), they quickly fail when the change is made to silicone. DOT 5 Silicone fluid has a tendency to foaming. Dale Ice recommends the use of an anti-foaming agent made by Dow Chemical Co. (phone 1-800-FOAMFREE).

It should be further noted that silicone DOT 5 does not absorb moisture, it displaces it. Water entering the system will settle in low places or accumulate in points of low flow. Because of a problem with trapped air bubbles, bleeding the system is more difficult. The air bubbles will cause altered braking action because a tendency toward spongy feel. Further, DOT 5 does not attack paint as glycol based fluid can, but it leaves a residue behind that can cause problems with subsequent paint application.

Several owners have run cars satisfactorily on silicone for years. Some use a filter between the pump and tank to collect sloughed off material before it blocks anything. They report problems caused by its very high electrical insulating tendency. Weeping around the brake mushroom can cause trouble with the brake switch. Steps must be taken to isolate this part. Also, trouble with contamination of ignition points has been reported. Many owners convert their systems to electronic, thereby eliminating the points. One owner has eliminated the brake light switch problem by adding a pressure switch to the braking circuit on one of the front calipers.

Other owners have suggested that the addition of castor oil to the glycol-based fluids improves lubrication and ride without causing harm. This theory is supported by the discovery that the original specification LHS was castor oil based. At one time, CitroŽn was recommending the use of castor oil as an additive, but troubles cropped up in very cold climates as the castor oil thickened. Castrol R racing oil (used by some racing motorcycles) is one source. Model engine or aircraft suppliers are another. Castor oil is also available in health stores. In addition to its effect on lubricity the castor oil raises the viscosity to around 35 mPa.sec at 75 deg. Fahrenheit, the same as LHS (plain glycol brake fluid of DOT 3 or 4 has a viscosity at this temperature of 23.1 mPa.sec). This mixture has been used for some time without problems, except in cold climates. If the car is to be used while ambient temperatures are below the freezing point, the castor oil fluid should be drained and replaced with DOT 4 brake fluid. For more information of this experiment, please go to Mark Bardenwerper's web site .

LHS cars can be converted to LHM if all the seals are replaced with LHM-type seals. This sounds like a better use of the labour of changing all the seals than going for DOT 5. The metal components were not changed when CitroŽn changed to LHM, but the color changed from black to green. This step should not be omitted to avoid fluid mixups.

Texaco Biostar is an interesting alternative which has the advantage of being safe to handle and is biodegradable. It is also apparently compatible with both types of seals, being based on rapeseed. But I worry about Biostar in low temperatures. A pour point of -20C could pose problems in some parts of the world in winter - Texaco say it is for use between -15 and +80 degrees C. Presumably farmers must face this problem - I wonder whether there is an agricultural solution to this? Also, I would imagine there is a risk that some parts of the system might exceed 80 degrees C. From the figures it is clear that Biostar is a good bit more viscous than LHS or LHM at the temperatures for which I have information. I guess this would have the effect of stiffening the damping, which some might like, and of increasing the work required of the system in various ways. The main worry may be increased loads on the pump, which might reduce its life expectancy a little.

Biostar comes in two grades, 32 and 46, both rather more viscous than the LHM or LHS, and both with an importantly lower viscosity index than LHM. I have chosen Biostar 32, which is the version nearest to the CitroŽn specs. Biostar is the only oil I have checked for which Texaco make particular claims about wear resistance. It sounds as though this might be worth trying in a system which has suffered wear - e.g. from the use of brake fluid. However, its rather high pour point and low viscosity index might give one pause before using it in very cold conditions, and it would be a comfort to know its specified upper temperature for use (80 C) is safe.

Yet another possibility is over the counter rapeseed oil, or "Canola", as it is now more commonly known. It is compatible with modern synthetic rubbers used in all but the earliest cars. It is fatal, however, to natural rubber, and it also has been found to be harmful to rubber used in adjacent components on older cars. It has a rather low viscosity index, meaning it will be thicker than recommended at low temperatures. But several owners have had success with it in moderate climates. One might be concerned about the possibility of decomposition of vegatable oils, so regular changes would be a must. Rapeseed oil is the only known vegetable oil that is not harmful to LHS seals. Never use any other type.

Specifications For Fluids Used In LHS Cars Top
These specifications are harvested from other sites. They are not intended for use as MSDS or true data sheets!
Characteristics Unit LHM+
For Comparison Only. Do Not Use!
LHS 2 DOT 4 DOT 5.1 DOT 5 Silicone based brake fluid Texaco Biostar 32 Mark's Mix (18.5% Castor Oil/DOT 4)
Colour - Green Red Amber Amber Blue/Purple Clear Yellow
Density at 15C Kg/l 0.830 1.007 ? ? ? .92 ~1
Viscosity at -40C cSt <1200 ? 1800 ? 120 ? ?
Viscosity at 20C cSt ? 32.4 ? ? ? ? 34.7
Viscosity at 40C cSt 18 14.5-16.5 8.59 ? ? 32.6 ?
Viscosity at 100C cSt 6.3 4.5-5 1.5 ? 7 7.53 ?
Viscosity index - 355 256 30 ? ? 210 ?
Pour point Deg C -62 ? ? ? ? -28.9 ?
Boiling point Deg C 255 ? ~230
(lowers with age)
(lowers with age)
(does not lower with age)
Flash point Deg C 135 99 149 ? 204 210 ?

Fluid for LHM Cars

Original Fluids For LHM Cars Top

Best is LHM. No other readily available fluid has the same viscosity, and any replacement runs the risk of giving rise to behaviour other than that which was originally intended, most particularly in conditions of extreme temperature.

Properties of LHM+ (the latest version, fully compatible with original LHM). Recommended by CitroŽn.

Alternate Fluids For LHM Cars Top
Next best is what is known in the aviation community as "red oil" (MIL-H-5606). It is cheap to buy and is just slightly lower in viscosity. Having existed before LHM, it's reasonable to believe that CitroŽn would have allowed it if they thought it delivered what they required. But the demands of the CitroŽn system are unique and very specific. It is also conceivable that they were concerned with the color similarity of LHS, a potential disaster for older car owners, should incompatible fluids be accidentally intermixed.

The actual specifications for LHM differed from the earlier version of "red oil" in one important aspect, VI (viscosity index). The VI represents how much a fluid's viscosity changes with temperature. The higher the number, the more constant the viscosity will remain across a given temperature range. As of 2/97, MilSpec 5606(F) was supplanted by 5606(G). The major improvement was in the area of low temperature viscosity. Since then, the specification has again changed and is now (H). The VI for LHM is over 350, while the older 5606(F) red oil was around 300. The VI of the new MIL-H-5606(H) is now above 370, surpassing LHM. For comparison, the VI for Dexron (regardless of type) is only about 200. However,  the viscosity of red oil is lower than LHM at all temperatures.

A comparison of MilSpec 5606(H) and LHM+:

Third best is Dexron automatic transmission fluid. More than twice as viscous as LHM at low temperatures, Dexron may be problematic in cold conditions. The basic problem with Dexron is that it has relatively poor lubrication qualities. Keep in mind that a number of critical hydraulic components in your car rely on very close tolerances. Long term use of Dexron in a car may lead to excessive wear in these areas. As internal leakage increases, system cycling increases, which in turn puts more stress on the source/supply for the system. By long term we are not talking about weeks or a few months, but wear can start in as little as a year in a heavily used car.

<>Automatic transmission fluids are highly specialized due to their dual purpose. They are specially engineered to provide proper lubrication to the transmission's bearings while providing enough friction so that the transmission's clutches work as designed. Their requirements conflict with the specifications of LHM in two critical areas, viscosity and frictional drag.  The latest versions of these oils are less harmful than they used to be - some synthetic oils manage good wear control while still having quite high friction for brake bands, or the cones in synchromesh gearboxes Modern ATF is thus an acceptable fluid to use temporarily, but I think I'd prefer to see a more specialised oil in Citroen hydraulic systems in the long term. There are, after all, several easily obtainable and reasonably priced options which are a good deal better.

A car with a Dexron transfusion that has lasted for a number of years and has high mileage may not take kindly to a charge of standard LHM or LHM +. The only cure for these cars is to use a slightly thicker fluid such as AirCraft #15. Just because a fluid does not destroy the seals does not mean that it is not causing other problems that may take years to surface.

Owners are further warned not to use type F or Mercon. These fluids diverge excessively from requirements. They use a lot more friction enhancers and cause much more wear than Dexron. Ford type F or Mercon compatible transmissions will not shift properly unless the specific fluid is used because of specific clutch design.

Motor oils are mentioned in the manuals, but they are suitable only for emergency situations and should be flushed as soon as practical.

A comparison of Dexron and LHM+:

Late Breaking Information Top
At least one owner is using a hydraulic fluid made by Kendall called Hyken Glacial Blue. Others have used Exxon's Univis 13. Some Mercedes and BMW cars use a fluid called Pentosin CHF. We've found two versions of this, and both look to be suitable for use in later D series suspensions (they are even green) according to the information we have. They are usually quite expensive! Mercedes uses a fluid called ZH-M in some of their cars for power steering and self-leveling rear suspensions. This fluid, while almost certainly harmless to seals in CitroŽn cars, has a lower viscosity index and its viscosity is generally lower than that of LHM. This would demand more work of the high pressure pump and would have some effect on suspension behaviour. It might also be a bit marginal in high ambient temperatures, or when the brakes were really punished, as when descending a long mountain pass. It is, however, probably a better option that ATF (see below). It is cheaper than the Pentosin CHF fluids but costlier than ATF. We have added that data to the chart below, though so far there are no confirmations of owners' experiences. The only conclusion I have right now is that it is likely that there may be several hydraulic oils that could work in LHM cars, though there can be adverse effects on performance under varying conditions. I have added Shell Tellus to the list below for comparison purposes. You can see that common hydraulic fluids will probably work, but their viscosities are not within approvable range. For the most part, they thicken excessively at low temperatures and therefore would be unacceptable except in mild climates.

Specifications For Fluids Used In LHM Cars Top
These specifications are harvested from other sites. They are not intended for use as MSDS or true data sheets!
Characteristics Units LHM+ Pentosin
CHF 7.1
ATF+3 Texaco
MIL-H-5606 (H)
ARAL Vitamol ZH-M Shell Tellus 22 Exxon Univis  13 Kendall  Hyken  Glacial Blue
Colour - Green Green Green Red Red Green ? Red
Density at 15C Kg/L 0.830 0.857 0.825 0.825 0.86 .861 .866 ? .855
Viscosity at -40C cSt <1200 1050 <1100 1500 600 6000 (?) ? 371 2840
Viscosity at -20C cSt ? ? 230 ? ? ? ? ? 240
Viscosity at 0C cSt ? 75 ? ? ? ? 180 338 ?
Viscosity at 20C cSt ? 32 ? ? ? ~32 ? ? ?
Viscosity at 40C cSt 18 18 18.6 36.8 13.2 16 22 13.5 14.9
Viscosity at 50C cSt ? 14.3 ? ? ? ? ? ? ?
Viscosity at 100C cSt 6.3 6.0 >6 7.65 5.0 ~4.2 4.3 5.3 4.4
Viscosity index - 355 326 320 185 370 181 100 404 233
Pour point Deg C -62 -62 <-62 -45 -60 -40 -30 -60 -60
Boiling point Deg C 255 ? ? ? ? ? ? ? ?
Flash point Deg C 135 ? ? ? 82 140 204 100 170

Seal Compatibility Top
A seal's compatibility is determined by its durability in the fluid that it must operate in. The systems that we are concerned with use glycols (normal brake fluid) or petroleum (paraffin /mineral oil/others) or synthetics fundamentally designed to supplant them.

Seals come in many shapes. Automatic transmissions have "o" and square rings, flanged, or lip seals and gaskets. Primarily, the hydraulic systems in our cars use o-rings and protective covers such as those found on the ends of the height correctors and in the suspension ram boots. But seals can also be made of other elastomers, metal, paper or other fiber products, or specialized plastics such as teflon.

For glycol based fluids the material of choice is ethylene propylene (epm, pdm). Introduced in 1961, it still has the best resistance to brake fluid. A new compound that has been recently introduced and shows promise as being suitable for both glycol and petroleum based fluid is Aflas (TFE Propylene/trademarked 3M).

For Petroleum based fluids (LHM, Dexron, 5606 Spec) the following materials are the most widely used: Fluorocarbon based, (Vinylidene fluoride - hexafluoropropylene) also know under the trade name Viton (and others) and Nitrile (NBR or Buna N, Acrylonitrile-Butadiene Copolymers). Of the two Viton has the best mechanical strength/temperature resistance and is much more expensive compared to Buna N. While there are others, the above two are the most common.

The seals in our cars are of two types - static and dynamic. Static seals are those where the sealing faces do not move. Dynamic seals are those where one or more of the sealing faces moves relative to the other. To list a few, the power steering rack, suspension cylinders, brake pistons, clutch engagement control/steering speed control (in SM's), height control valves, rear brake articulating joints on ID/DS series are all examples of dynamic sealing points. The high pressure pump has one dynamic seal, though it is a metal to metal seal at its driveshaft. The suspension sphere diaphragm is a special kind of seal and presents real problems from a design standpoint. Not only must it be resistant to the fluid in use, it also has to have extremely low gas permeability, excellent flexibility and tear resistance over a wide range of temperatures and pressures.

When CitroŽn introduced LHM in 1966, they ran into serious high temperature problems with the diaphragm material during the first couple of years, primarily with gas permeation. This problem has been almost completely eliminated in the latest cars, such as the C5. The diaphragms are now 2 ply.

More complete information on seal compatibility can be found at Engineering Fundamentals and Marco Rubber .

Fluid Changes and Flushing Top
Before I close I'd like to address the issue of fluid changes. Though owners of LHS cars should already know that they must make frequent changes because of hygroscopy, the fluids used in later cars also degrade with use. The long chain polymers that are used to improve viscosity and provide lubrication under extreme pressure begin to break down. This is caused by the physical shearing of these chains as the fluid is pumped under high pressure throughout the system. Furthermore, corrosion can occur in later cars, too. Anti-oxidant additives have a finite life. Once they become ineffective, the polymers oxidize and change their characteristics. Moisture trapped in the system will cause corrosion. Dirt and other particles collect in the system and accelerate wear. Keep in mind that just because the fluid looks "clean" does not mean that it is providing optimum protection. The factory recommendation of a change every 24,000 miles/40,000km is not overly restrictive for LHM cars..

LHM systems use a flushing agent called Hydraurincage. It can be used full strength for full effect, or it can be mixed. Hydraurincage can be left in the system for as long as 3,000 miles/5,000km  before it needs to be removed and fresh fluid installed.

Those of us with older ID/DS cars using glycol-based fluid need to be more diligent regarding fluid changes. When the car rises, fluid moves out of the reservoir, drawing in air from which water is absorbed. This problem is aggravated in moist climates and lessened in dry. High moisture content can greatly increase corrosion problems in the system - especially on parts in areas where the fluid is static, as in the brake cylinders. High moisture content drastically lowers the boiling point of your fluid. This can make braking dangerous as trapped water boils under extreme pressure and becomes compressible vapor. We have shown that several owners have used alternative fluids to counteract these problems. Yet just like LHM, the viscosity modifiers and lubrication additives all degrade over time. This will happen faster than in non CitroŽn systems because of the much higher working pressures, the constant circulation and the influx of moisture laden air through the vent. The factory recommended change interval is  18,000 miles/30,000 km. I would recommend every 2 years under normal conditions, 1 year under very humid conditions, regardless of mileage in cars that are not used frequently.

The flushing agent specified by Citroen for LHS cars is hexylene glycol. It should only be left in for 20 miles/30km if used full strangth.  One should never use Hydaurincage, as it is not compatible with LHS seals.  I (Mark) have used hex. glycol. It is very expensive and hard to find. I drained and added only about a quart instead of a complete change and left it in for double the time. Results were favorable though I did have to clean my filter a few extra times. My steering seemed to have better feel and power, for one thing. Some owners used pure brake fluid and a little alcohol. I would not recommend using alcohol as it will not lubricate at all and could cause damage. Alcohol or soapy water can also be used to clean LHS components. They must be thoroughly dried before reassembly.

Flushing instructions can be found here. This old brochure was written before LHM was put into use. For later cars, the procedures are the same. Just substitute the proper fluids for those stated.

Credits Top
Image courtesy Pomini Paolo
Special thanks to Stan George, Alastair Macintosh, Adam Reif, Shane Leviston, Bob Alexander, Brad Putchat, Maurice Gunderson, Steve Hammond, Ulf Petermann, Dale Ice, Jack Shotton, Jint Nijman, Carter Willey, Nils Oehler and the selfless contributions of the CitroŽn List communities found at Egroups Dseries-L and citroen-dsid . This was truly a world class effort!

Copyright 2006, Tony Jackson and Mark L. Bardenwerper, Sr. Please send suggestions to Mark or Tony .

It must be stressed that the authors do not in any way wish to take responsibility for consequences arising from the use of any of this information. Readers are hereby informed that all of this work is fully experimental in nature, and any actions taken by the readers will be construed as being taken on their own volition. In other words USE THIS INFORMATION AT YOUR OWN RISK!.