Hydraulic Leak in Citroen SM

[PsychDr]

Greetings, friends.

I have 1973 Citroen SM that I have professionally restored/refurbished including mechanical and hydraulic system. The car drives beautifully and the system works as intended. The brake accumulator, pump, front and rear suspension spheres and the steering rack are all new including all hoses.

Recently, however, I discovered an LHM leak with fluid pooling directly under the pump plate. The pump cycle time clicking interval when in normal drive mode is every 9 to 10 seconds and when in highest position (when I examined volume and replenished to max line) is at 20 seconds. Other than the leak, I am not experiencing any problems with the suspension -- Suspension rises appropriately at startup, with both braking and steering working well with no problems during drives.

I would appreciate any insight on what might be the cause and specific location of leak and the remedy.

Thank you.

[Hell Razor5543]

Hi, and to the Forum. Wow, what a stable!

I personally cannot offer any advice, as I have very limited knowledge of the green blooded Citroens, but if you wait a while I feel certain somebody with plenty of knowledge will answer soon, and point you in the right direction.

[CitroJim]

Welcome

Sounds like there's a leak from the pump itself or from the pipes running to/from it...

The pump itself and indeed the overall hydraulic system sounds health enough...

I'd uncover the pump as much as possible and clean it very thoroughly... Then run the engine and observe - the source of the leak should soon become apparent. This was always my go-to method when confronted by such leaks. Clutch and brake cleaner works well for the clean-up job...

One possible cause is a tiny hairline crack in the outlet pipe around the flare under the union nut... I've known this occur on a BX and is caused by metal fatigue over many years... You'd be surprised how strongly the pipe can pulse when the pump is working to pressurise the system.

The other possibility is a leak from the rubber inlet pipe from the LHM reservoir...

Else that it's likely a pump body gasket has failed... I don't know if you have specialists across your side of the pond who may be able to overhaul your pump but there remains at least one specialist in the UK...

[aerodynamica]

Sounds like you've done good work with the car! I agree with Jim about the possible hairline crack on the high pressure pipe from the pump. There used to be stories around the later 90s of sudden failure of that pipe where a restorer had replaced it with a corrosion-resistant cupro-nickel pipe. If I recall rightly, the pipe failed in a clean break in line with the 12mm pipe union nut and was deemed to be vibration induced fatigue. The thought was that the use of Cupro-Nickel replacement pipes for this location wasn't suitable (although it's good everywhere else on the car) because of the different property of the material. I think the SM has a vulnerability with this pipe that's relatively short compared with many other hydraulic Citroens. I think it actually does coil one time to increase its length.
Your car sounds in great shape but I feel 8-10 seconds of cycle time is too frequent with the engine at idle and the hydraulics doing no work - should get at least 20 seconds and ideally 50+. My '68 ID19B had a really frequent cycle time of around 8 seconds even after all new spheres, filters and LHM and a new brake valve. It turned out to be the common wear in the steering rack and a reconditioned rack immediately fixed it and the regulator cycle was nearly 2 minutes. The suspension could stay up for over a day and a half too. Don't underestimate how much reserve pressure a worn steering rack causes!

[xantia_v6]

An SM will have DIRAVI which which is constantly using high pressure LHM, so thar regulator tick rate may be normal, it is certainly not alarmingly fast, but if it gets faster, it would be worth changing the accumulator sphere(s?).

[aerodynamica]

That's true, it does have DIRAVI but although it is under pressure, if it isn't moving the steering (or the steering speed governor) then it should be static and not 'using' pressure. Only internal leakage would cause increased tick rates while stationary. I've had it with Xantia brake valve, D steering rack and on at least two of my 4 CXs with DIRAVI. The components can wear eventually after decades of use. It only takes a single component to cause the effect. But if it's not too bad then it can probably be dismissed

[Jay-Bruce]

Welcome aboard, that's a pretty lush sounding collection you have there. Not wanting to rain on your parade, but I'm inclined to agree that the tick is too frequent, from memory, on a BX, that frequent a tick would suggest the accumulator is flat? However, that doesn't solve your leak, the good news is that as it's relatively minor, it's not likely to be expensive, and as your car is recently restored, everything will be clean making it easy to find the leak. Clean with brake cleaner or another evaporating solvent, tape vertical faces with masking tape, and cover larger areas with white paper, and run it, the tape and or paper will absorb the oil and map your leak path for you. I'm not familiar with the SM, but when you say the leak is pooling directly beneath the pump plate, I'm guessing you mean wondering it's leaking from the input shaft of the pump?

[Hell Razor5543]

May I also suggest that, as part of your maintenance routines, every six months or so you loosen off all of the spheres of all of your cars by quarter of a turn or so, and then do them back up hand tight (no more than that is required). Spheres that are left too long can get biblically tight, to the point where extreme measures are required to replace them (I genuinely have had to use a hammer and chisel to get spheres to start moving). I was told this trick by a Citroen engineer who did this on his green blooded Citroens. Make certain the vehicle is properly supported before doing this, as it is known that people have been killed when the suspension has failed, dropping the car down on them.

[PsychDr]

Thank you all for your invaluable perspectives and advice. Indeed, I have replaced everything hydraulic including the main accumulator and the steering rack - and as mentioned here, the clicking pump interval of 10 seconds with the minor leak under the accumulator is what triggered my concern. I will continue to be mindful of additional changes and follow-up with my mechanic accompanied with your excellent advice! This SM has been in our family for over 4 decades and the refurbishment has been especially meaningful for me and my dad. All the best!

[aerodynamica]

Hi again, well with the steering rack having been replaced it does sound like you've done 'the works' on the system. I hope you manage to get the leak fixed without to much trouble.
I was thinking there that any leak from the pump to the regulator is effectively 'upstream' from the regulator and accumulator and really this is the only part of the system when a leak or any other fault would actually NOT cause rapid cycling of the regulator. You would actually see the opposite effect if there was a leak or any other loss e.g slipping drive belt, at the pump (SM's is driven by a cardan shaft though). In those cases you'd get long, slow build up to the cycle as the pump struggled to meet the regulator's demand. Any fast cycle times have to be caused FROM the regulator - including the regulator itself: It can happen that the spring inside the regulator that force the shuttle valve against its seat can tire over decades and cause the cut in pressure to be lower than 140 bar. Similarly for the cut out spring. There are shims fitted under these springs in the factory and indeed I added shims to both springs in my CX 2400 that raised the pressure of the system. If I recall a 0.3mm shim adds 1bar but that could be wrong. Citroën actually added a second spring to the cut out valve from a certain year of XM and Xantia models, I believe- to mitigate the relaxing of the springs over years. It's not likely to affect it that much though I certainly noticed a difference in my CX's steering at that time. Certainly if the cut in and cut out pressures were closer together it could cause the cycles to be faster. It may of course simply be a characteristic of the SM's particular system.

An easy way of eliminating some aspects of the system is to time the cycles over a minute and take an average and then repeat the same with the suspension on low. This takes pressure out of the suspension and rear brake circuit. Any increase in cycle time might show that the loss is via the suspension or brakes. If you then press and hold the brakes hard, any improvement to the cycle times suggests the brake valve (the SM has two) might have sufficient wear to allow leak through and that holding the brakes causes the supplied pressure to be dead ended to the callipers and hence, not leaking through. This last thing is exactly what happened on my last Xantia- rapid cycle times that improved to near normal level with my foot holding the brake. A replacement brake doseur valve fixed it in one go.
Anyway, if you're car raises strongly, can steer lock to lock without running out of pressure causing the stop lamp to illuminate, then you're probably fine and once that pump leak is found you can confidently hit the road in style.

[PsychDr]

Good day, friends, and thank you Aerodynamica, for your comprehensive assessment and lead towards the regulator. Indeed, as I noted, I sourced and replaced all spheres, accumulator, pump as new and my steering rack was a professionally rebuilt one. The only variable is that my regulator was rebuilt by my mechanic using a new rebuild kit sourced from a reputable Citroen SM supplier. While I understand that these kits are fairly straightforward, it is plausible that something went awry with the rebuilding of the regulator -- to your point about the spring and shims. I will continue to review based on your feedback and share with my mechanics as well. Again, appreciate the group's substantial advice!