Odd suspension issue ?

[Northern_Mike]

From what I have seen, Citroen had almost as little appreciation of its own suspension as does the public at large, by the time the Xantia was being built. What makes me say this? - no reference to how to change a wheel using the suspension's movement, steady alteration of the sphere design which has led to increasingly jarring low speed ride/poor smaller amplitute bump absorption and then the decision to abandon the time-proven design and use seperate electric motors for each corner on the C5 (at a stroke deleting the wonderful interconnection of rear brakes and suspension) on the advice of a bureaucrat in Brussels in the interests of safety. So, unwittingly building in a design fault would hardly be surprising.

The good bit is that people on here are aware of this recurrent problem and if it is air, should be fairly easy to eradicate.

As an afterthought, would those running comfort spheres experience just as much harshness?

My HDi Saloon has Estate rear spheres on, they're MUCH softer. It wallows like a drunk pig when driving it hard on twisty roads unless in sport mode. It still jars over sharp bumps at the rear though.

As for changing the wheel with the suspension movement - I know how to do it, did it on a GSA (Stepmum had one in Greece) and did it a couple of times on the first Xantia, but honestly, it's much less faff (and you can leave it in gear for extra security) just to do it the way you would with any normal car..

[Mandrake]

Just a thought about air bubbles in the system, wouldn't that have the opposite effect i.e. soften the ride, air being compressible?

No, air bubbles mixed in with the oil don't make the ride softer, they make it harsher and also cause some loss of damping control, the reason being that the bubbles of air in the oil are on the wrong side of the damper valve, thus they don't simply serve to increase the total compressible volume of gas, but instead add another component with its own spring rate in "series" with the existing ones.

Normal operation of the suspension is that you have all your compressible medium - the nitrogen gas - together in one volume within the sphere, behind a damper valve situated in the neck of the sphere.

When the wheel moves in response to a bump the oil transmits the movement to the damper valve hydraulically without any compressible medium in between, the damper controls the flow of oil into the sphere before it can then compress the gas.

Because there is nothing compressible between the piston attached to the wheel movement and the damper valve this gives excellent damping control even of small movements.

Conventional coil spring and external shock absorbers don't damp small movements that well because the shock absorbers are externally mounted with isolating rubber bushes. It takes a certain amount of movement of the suspension before the shock absorber even starts to move and thus damp the movement, meaning small (a few millimetre) movements aren't damped much.

Anyone who has driven a hydro Citroen that is working properly will know that ultra stable feeling that the car has in response to small undulations, and I believe this ability to damp very small movements is thanks to the incompressible coupling between the wheel movement and damper valve - in the Citroen the damper and spring are in series, (the "spring" can't move at all until the oil has passed through the damper) while in conventional suspension they're in parallel. (The spring can move a bit before the damper starts to move and exert control)

If you get a large amount of compressible air between the piston and the damper valve the result is that small movements and oscillations can occur without oil passing through the damper - these small movements are no longer damped. In my experience with my previous Xantia, when the ride got really bad the damping seemed to suffer - the car would not feel rock steady but would feel a bit unstable and tend to undulate up and down a bit on uneven surfaces.

That's the first problem. The second is that now that you have a compressible blob of air on the wrong side of the damper valve the volume of gas does not simply add to the volume already in the sphere - when you hit a bump the small pocket of air in the oil will compress rapidly first due to the damper limiting the flow rate into the sphere, causing a pressure spike that results in a jolt.

Most of the time when a wheel hits a sharp bump it doesn't just follow the shape of the bump, instead the suspension oscillates up and down at the wheel hop frequency (about 10-15Hz) for as much as 1/10th of a second or more, (you'll sometimes see this if you watch the wheel of the car in front of you when it crosses a pothole) but you don't notice it because the suspension isolates you from this oscillation - the wheel oscillates but the car body doesn't. However with the air bubbles it causes hydraulic hammering as the air bubble rapidly compresses to almost nothing and expands again 10 times a second or so, because the damper valve is restricting the flow into the sphere... this is a source of harshness on small but sharp bumps.

Another way to look at it is that the air bubbles cause small but sharp discontinuities in the pressure versus oil displacement curve of the suspension. Instead of a smooth variation in pressure with wheel movement there are sudden abrupt changes in pressure with the wheel movement over the bump which is felt as a harsh vibration.

[Mandrake]

I think the most significant difference between the GS and CX is that they are mechanical systems compared to the electronically controlled Hydractive system.

Don't forget a standard Hydropneumatic Xantia has no ECU for the suspension, its entirely mechanical as well, but still seems to suffer from intermittently harsh ride. To be fair though, it doesn't seem to be nearly as bad as Hydractive 2 in this regard, but I think thats as much about the absence of long, large diameter piping to the struts as absence of an ECU.

I am less convinced by air in the fluid because as the fluid is returned to the reservoir it becomes de-aireated, like a dry sump oil system.

That depends. It's true that there is a filter in the hydraulic tank between the overflow return and the pump pickup which will de-airate most of the bubbles that come back from the return lines, but in my testing I found that the filter can only cope with a certain quantity of bubbles - a few large bubbles here or there will just sit inside the filter at the surface and pop, but a large quantity of fine bubbles overwhelms the filter and passes through it.

The return oil going into the tank is fed through a narrow slot down near the bottom of the tank inside the filter, and in my testing I found that this tends to froth up the oil a lot with fine bubbles which then pass through the filter into the main tank where they are then sucked back into the pump - you can clearly see frothing on the surface of the oil outside the filter that shouldn't be there.

The pump pickup is near the bottom of the tank outside the main filter, so if bubbles are forced down into the bottom of the tank the pump can suck them up.

In an effort to combat this I attached a small curved hose to the return outlet inside the filter which redirects the incoming flow upwards - this allows the airated oil to enter the tank at the surface where the bubbles can disperse more quickly and prevents the bubbles being driven down into the bottom of the tank.

Although it didn't cure the intermitent harsh ride completely, it did seem to make quite a difference and the ride became a lot more consistent - so much so that I left that pipe in the filter permanently, even when I sold the car. (When the new owner, who is Citroen savvy, goes to change the LHM he will be thinking what the heck is this doing in here )

I've attached a couple of pictures to show what I'm referring to: (click to view full size)

Just this small change suggests to me that the design of the tank, filters, and piping in the tank (all of which is very different to older Citroens) could be part of the problem, failing to de-airate as well as it should, compounded by a high continuous return flow from the PAS.

Also, most of the return fluid is from power steering anyway, as far as I understand it anyway.

Yes it is, that was one of my points. CX/GS didn't have large volume constant flows returning from power steering frothing the oil up. They just had the return from the pressure regulator. In a Xantia the PAS return flow is about 6 times greater than the regulator cut-out return flow.

It doesn't matter where in the system the bubbles come from, if they get sucked into the pump they will cause trouble. If the quantity of bubbles being sucked in varies from day to day the ride quality will also vary.

[Old-Guy]

Thanks Simon, fascinating and illuminating stuff. But, for me, the $64,0000 question is "How is all this air getting into the LHM in the first place?". All of a Citroen hydraulic system upstream of the two pumps is under positive pressure, so it seems to me that if significant amounts of air (not nitrogen) are getting (being sucked) into it, there must be a leak into the negative pressure (pump pickup) section of the system - just like the common air-in-fuel problem of XUD/Bosch engines. The prime candidate for me would be the reservoir head unit as the reservoir is well above the pumps.

According to my theory, the pump would suck up a steady stream of small bubbles, which would be broken up, compressed and dissolved in passing through the pumps. The air then comes out of solution as the pressure is released resulting in an aerated return flow to the reservoir. Dare I suggest that the apparent improvement from your modification could be co-incidental by disturbing the pickup pipes?