Air bubbles in the LHM gives harsh ride?

[pprado]

It will look like magic product selling but... it worked for me!

I know this thread is not new, but I loved reading it some time ago.
The clean pipe idea was great and when I did it on my car (who also has a harsh ride), it showed tiny bubbles.
By fiddling with the steering wheel I could see they are from there - turning the wheel makes them appear.
Regulator clicks were kinda short, about 10 seconds.

With that in mind, I noticed that the design of the return filter inside the reservoir takes that bubbles in account - if you pay attention you will see that it prevents direct communication from the return oil to the oil sucked by the pump.
But it appears that the design is not sufficient by looking at the bubbles appearing on the path to the pump...


I thought the solution to separate oil and bubbles were to make the return go over the oil surface - that way bubbles would keep at the surface, at most. So I removed the return filter and fitted an U-shaped tube at the end of the internal return tube, sending the oil back to the reservoir cover.

Now the part that will sound crazy.

Clicking has increased to *more* than 4 minutes,
Steering is REALLY light as anyone would expect on this kind of car,
STOP light disappears BEFORE the oil light,
The boot-sitting test WORKS on my car...



I am still incredulous of all those things. As a friend of mine said - Could Citroen miss this that way?

One thing I liked: When the sphere needs to be charged, the pump sound is much softer, and its time is also very short.

Of course I did citaerobics several times - and I DID notice a ride improvement. I still need to fix ride height, change the front HA sphere, and probably change the fluid. So I do not have yet a floaty ride, but it did improve noticeably!

All I can say is that it is dead easy to try out. Just remove the reservoir cover, remove the return filter (the "D" one), fit a piece of tubing on that downpipe, and direct that tubing to the top of the reservoir. (by the way, I used a section of the same clear pipe I used on the path to the pump)

If that proves to be a real cure, I plan to order another return filter and to adapt the U-tube on it in some way.

Hope this HELPS!

[AndersDK]

Thanks Pedro -

Interesting reading indeed.
Can you please add exactly what model of Citroen you tried it on

[pprado]

Sure...

My Xantia is a VSX one. Hidractive, 2.0 16v, 1997.

BTW, today clicks are short again... When i get home I'll see if the tube got disconnected. If not I think the magic is gone :-/

3 days running like that. If it is gone the car was surely waiting for me to post here

[Mandrake]

Hi Pedro,

Interesting idea about allowing the oil to return at the top of the resoviour instead of the bottom - I too have noticed that if there are significant bubbles of air returning to the tank that despite going through two fine gauze filters, they get sucked back up by the pump again.

I am concerned about the long term effects of running with one filter removed however....

Yes, the results from what you did do seem too good to be true, and as soon as I read the bit where you said you had done Citroerobics as well, I knew that this was what had really caused the improvement, as I have been down this road myself MANY MANY times myself

I still have not conclusively tracked down what causes the intermitant harsh ride that some Hydractive 2 Xantia's have, each problem I find and fix makes an incremental improvement, but the basic problem is still there lurking in the background...whatever the cause of this problem is, it is often fixed temporarily by Citreorobics, only to return days or weeks later.

In your particular case, assuming your accumulator sphere is good, the fact that your regulator cycle time varies wildly from very short to very long and back again, suggests that you have an intermitant source of leakage in the hydraulic system somewhere bleeding the system pressure.

The number one culprit for this on a HA2 model is a faulty Hydractive electrovalve - they can develop a fault where they won't seal completely in either the on or off mode, or both. This can bleed the main pressure quite rapidly causing very frequent regulator clicking, and intermitant operation of the hard/soft switching of the suspension.

In my particular case it was the front electrovalve, and it only leaked in the "on" (soft) mode, and it was bringing the regulator cycle time right down to 8 seconds with a new accumulator sphere.

This can be tested - depressurize the car, disconnect the rubber return hose on top of the electrovalve near the radiator and quickly block the hose before you loose too much oil. Now connect a clear hose to the stalk on the electrovalve and lead that hose back to the filler hole in the LHM tank.

Now start and pressurize the car and watch for continous flows of oil through that pipe - there should only be a very slight "squirt" each time the suspension switches from hard to soft or back, but no continous flow.

If you see a continous flow the electrovalve is faulty. Note: Often a faulty electrovalve does not always appear to be faulty - it is caused by poor sealing/seating of the needle valve, so you may have to let it switch on and off many times before it misbehaves.

You can do the same test at the rear into a container but you would need a pit/hoist/ramps to do so safely.

Good luck tracking down this problem, because as you probably noticed from reading this thread, nobody has yet cracked the true cause of the problem.

Regards,
Simon

[pprado]

Hey Simon thanks for the reply and tips. I will surely try that valve check out as soon as I get a small enough clear tube - I have some large ones only now.
I hope I am contributing in any small way to get near to the root cause - if not at least I am having fun...

Anyway, it was time for a second try.

I opened the reservoir to find out that the tube I did put had bent. Because of that, all flow was escaping through the original pipe's side slit.

So I obtained 2 3/4 L-shaped tube junctions and a tube that I could put to cover the pipe slit - so I made a square "U" pointing the end tube, again, to the top of the reservoir.

Fitting it to the reservoir was a pain... basically I need to disassemble, insert each half, and reassemble. Try assembling parts submerged in oil - fun guaranteed for whoever is watching you

Everything fitted, time to start the car.
Erratic clicking at first, then it stabilized on more than 2 minutes (I just got tired of waiting to click). Seems good - I did no citaerobics this time just to make sure I knew what changed...

Trying the steering, I felt it a little lighter, but I could also feel a little trepidation while getting to the locks. Ride was good too, same as on last experience.

Back from the ride the clicking was something more than 3 minutes. (I need a start/stop chronometer).

The bad news (for me) was that I pointed the tube to the LHM filler cap and forgot to push it firmly down... it seems that sometimes the steering return gets a reasonable flow. At least by looking at the way it spread over the reservoir and on the floor

Now let's see if this lasts. For anyone wanting to try, I did use L's made for heated water tubing (holds up to 80 celsius). But I advise to use 1/2 ones, but using a converter so the end is 3/4. This is to slow down the return speed - no one wants squirts on the top of the reservoir...

The other news is that it is possible to attach a piece of a thin fabric to the end of the tube, to act as a filter. I am seeking for one appropriate.

I did a photo of the monster, but it appears there is no way to attach it here. When I upload it in some place I'll post here.

[AndersDK]

... it seems that sometimes the steering return gets a reasonable flow. At least by looking at the way it spread over the reservoir and on the floor ...

The power steering needs LOTS of fluid - and hence returns LOTS of fluid

This is to slow down the return speed - no one wants squirts on the top of the reservoir...

You may try if that will work, but I would NOT recommand to slow down the return flow. Then the flow becomes a pressure - in soft rubber return hoses. Either the rubber hoses pops off, balloon up or burst - as they are certainly not designed to withstand pressure.

[pprado]

Thanks about the steering tip... I think I heard that somewhere...

About the slowing down process - it's accomplished just by increasing the tube width. I'm not holding the fluid in any way, as the original reservoir's pipe is like only 1/2...

[JohnCKL]

Today I went to a workshop with a Proxia computer which diagnose my XTD having 4 faults, among them, steering sensor out, hydractive I out, speed sensor out and one more which I forgot. The reason for all these faults are that they've been taken out during overhaul, replaced or have been moved. The steering sensor was easy to rectify, the rest just needed to be reset. Once the Proxia computer reset the faults, my front suspension which was previously quite hard, now is softer. Seems that hydractive system need the Proxia to reset faults to get back to original settings.

[pprado]

Interesting about the ECU not recovering from diagnosing a faulty sensor. When one sensor is detected as faulty the ECU will set a default value (for example, 100 Km/h on the speed sensor). The defaults will always point to safety and that means more probability of the hard mode...


Back to the bubbles to the top approach, I replaced the 3/4 sections with 1/2 sections that did fit better and allowed me to direct the flow to the side instead of the LHM filler cap. When removing the 3/4 sections I then realized that a piece of the tube was limiting the flow...

With only the 1/2 sections in place, the car feels like on the first try. Long click times, STOP light going out very fast.

After turning the car on, In 1-2 seconds the STOP light goes out and almost immediately the rear of the car drops a little. Right after, like 0,5 seconds, the front drops also. Then it stabilizes and that's it...

At this point, I am very confident that this U-section is bringing large improvements. The system is really degraded by having air on the pipes, as it seems.

Now where the bubbles come FROM? Although I do have a minor LHM leak on the power steering, I would expect LHM going OUT but not air coming IN.

Assuming no air can come in on the power steering, that air must be being sucked up by the pump - what points me to the tube between the reservoir and the pump. I'll try replacing it with a clear pipe, as a next step.


Anyway, being at the power steering or being on the path to the pump... the original design allows for realimentation of bubbles into the system, thus any minor leak that could cause bubbles to appear will get amplified.


This is how it looks:



This is how it fits:



A section of pipe must be put over the 1/2 downpipe to cover its slit. Then the "L" fits directly to the downpipe.
This 1/2-sections assembly do not force you to reassembly it inside the reservoir...

Anyone wanting to give it a try?

[JohnCKL]

Yes, anything to improve my Xantia. Will get a u tube, maybe one which can fit into the filter, no need to remove the filter which is essential to filter out all the gunk. After hydraflushing 2nd time in 4 years, system quite clean after a year of driving, just a fine layer of dark particles on filter when I opened up to clean last week.

[Mandrake]

Interesting about the ECU not recovering from diagnosing a faulty sensor. When one sensor is detected as faulty the ECU will set a default value (for example, 100 Km/h on the speed sensor). The defaults will always point to safety and that means more probability of the hard mode...

Hi - I was surprised about this too, everything I have read in the HA2 training manual says that the ECU is self resetting - if a fault is repaired that sensor will start working again with no resetting required. That's been my experience too.

With only the 1/2 sections in place, the car feels like on the first try. Long click times, STOP light going out very fast.

After turning the car on, In 1-2 seconds the STOP light goes out and almost immediately the rear of the car drops a little. Right after, like 0,5 seconds, the front drops also. Then it stabilizes and that's it...

I hate to be a wet blanket, but these are exactly the kinds of things I notice after doing just about any work on the suspension that involves depressurizing - for a short time (between hours and a few days) the suspension behaves extremely well - good ride, stop light going out quickly, etc.

For example I just fitted new droplinks on the weekend - and immeidately the ride was much better, and stayed that way for 2-3 days, now the harshness is coming back into the ride again...

At this point, I am very confident that this U-section is bringing large improvements. The system is really degraded by having air on the pipes, as it seems.

Yep, theres no doubt that air bubbles drawn into the hydraulic system cause all manner of minor malfunctions and probably ride harshness too.

Now where the bubbles come FROM? Although I do have a minor LHM leak on the power steering, I would expect LHM going OUT but not air coming IN.

Assuming no air can come in on the power steering, that air must be being sucked up by the pump - what points me to the tube between the reservoir and the pump. I'll try replacing it with a clear pipe, as a next step.

Been there done that. By all means try it, and there is no doubt that a faulty pipe would allow air to be sucked in, but for the really stubbon cars which have this problem and can't seem to be fixed, (such as mine and bernie's ) the pipe and pump have both been replaced and on my car although the problem is greatly reduced (suggesting it was a part of the problem) it still remains to some extent!

Anyway, being at the power steering or being on the path to the pump... the original design allows for realimentation of bubbles into the system, thus any minor leak that could cause bubbles to appear will get amplified.

I've been giving a lot of thought to your ideas and thinking back to my own efforts to stamp out these bubbles and I've thought of an alternative way bubbles can be sucked into the pump even when the pump and inlet hose are ok.

As you have noted (and I've noticed before too) if a significant amount of bubbles return to the tank via the return hose, despite two filters they are often sucked back in by the pump.

Where could these bubbles come from ? Perhaps these returning bubbles are not bubbles that have previously been sucked in by the pump, pumped into the system, and returned, but perhaps these are bubbles that are initially drawn in by an air leak somewhere in the RETURN piping in the car ?

The entire return system runs at very low pressure (much like the inlet pipe to the pump) and any tiny pinhole leak or poor joint (of which there are lots of press type joints in the return system) could result in air being drawn into the flow, but not allow oil to leak out when the system was off.

How many Citroen's have slightly wet return piping junctions ? (Lots!)

If the return system is the initial entry point of air then improvements from your idea now make sense - the bubbles that get into the return system harmlessly disperse in the top of the LHM tank instead of being drawn in by the pump.

Perhaps when the cars are brand new and all the return piping is in good condition the return system is airtight, but as the car ages and the rubber return pipes and fittings start to age they start to admit air well before they are leaky enough to allow LHM to leak out onto the ground...

It might also explain why I have noticed a temperature related aspect on my car - cold weather it rides a lot better and there are a lot less bubbles. Hot weather and I see lots more bubbles. Perhaps it is temperature affecting the sealing of the rubber return lines...

This is how it looks:



This is how it fits:



A section of pipe must be put over the 1/2 downpipe to cover its slit. Then the "L" fits directly to the downpipe.
This 1/2-sections assembly do not force you to reassembly it inside the reservoir...

Anyone wanting to give it a try?

I'm definately willing to give something a try!! I like Johns idea of a 180 degree U-turn which would allow the original filter to be left in, as I'm not keen on the idea of leaving one filter out permanently.

Thoughts bernie ?

Regards,
Simon

[AndersDK]

We had experienced from the swedish forum - years ago with a BX that was 'impossible' to get a good ride on - that the problem was caused by a torn filter gauze.

This allowed the air bubbles (foam) from the return lines to get directly into the pump feed - and thus the problem with air was kind of self-amplifying during drive.
It immediately showed that the filter gauze helps bursting the air bubbles, thus diminishing the air problem.
This is also a known problem from industrial hydraulic systems, where return flow spurting out in free cavity forms hydraulic oil foam.
The problem is solved by using additives in the hydraulic oil that decreases the surface tension of the fluid to prevent foam build up.

Exactly the principle used in dishwashers to make the water easy flow off your drinkglasses while drying up after cleaning !

Bet you have one of those small bottles under the zink in your kitchen !
2 drops (no more) in the LHM - and watch the difference 8)

[Mandrake]

A few more thoughts:

Firstly if a U tube is being used to redirect the flow back upwards within the confines of the return filter, we'd have to avoid directing the flow at the leakback returns - these are the smaller ones which have very short pipes and exit above the LHM level, otherwise we could pressurize the leakback returns enough to cause one of their pressfit junctions to blow off...

Also, how many high volume return flow pipes are there ? I'm pretty sure that on mine the power steering return, pressure regulator overflow return, and suspension height corrector returns all agreggate into a single pipe where it goes into the LHM tank, but I don't believe this is the case for all models and years - I think some varients have TWO pipes which have large volume return. This would need to be checked out.

The other thing is I probably hear people crying - how can there be an air leak in the return lines that wouldn't leak LHM on the ground ?

A number of ways actually, the obvious one is if the air leak is high up above the tank, but this is unlikely as the only pipes that qualify for that are the ones actually connecting to the tank, and most return lines are low down.

But there are other ways with high volume low pressure flows. Anyone who has done highschool physics must have seen a water tap vacuum generator - basically a flexible hose attached to a tap with the end open, and half way along a hypodermic needle is inserted into the pipe facing in the direction of the flow, with another small hose attached to that.

Provided the end of the main pipe is not blocked, (generating pressure) water will not come out of the side hose and in fact a vacuum is generated by the flow of water pulling air in through the small opening.

The same thing could easily happen in a press fit junction in a return line in the hydraulic system.

If the hole is small enough (pinhole sized) surface tension of the oil prevents it from leaking out when oil is not flowing or under pressure, however air can still be drawn in through the hole when the oil is flowing since air doesn't have surface tension...

Regards,
Simon

[Mandrake]

We had experienced from the swedish forum - years ago with a BX that was 'impossible' to get a good ride on - that the problem was caused by a torn filter gauze.

This allowed the air bubbles (foam) from the return lines to get directly into the pump feed - and thus the problem with air was kind of self-amplifying during drive.
It immediately showed that the filter gauze helps bursting the air bubbles, thus diminishing the air problem.

Hi Anders,

Thats very interesting.

I think you have hit the nail on the head here - perhaps even in a new system small amounts of bubbles returning from the return lines are normal and the filter gauze can deal with small amounts. But, as return pipes age and more air gets into the return lines, perhaps a point is reached where the filters can no longer deal with the volume of air returning and some starts to pass to the pump inlet ?

This seems to be what is happening.

This would also be affected by the shape and design of the LHM tank and its filters - which is different between a BX and Xantia.

So if we accept that this is happening and that it is impractical to locate and fix all the small cumulative sources of air leakage in the return lines (dozens of joints) does modifying the internal return line in the tank so that the oil returns above LHM level instead of in the bottom of the tank provide an acceptable workaround ?

Regards,
Simon

[AndersDK]

Simon I really dont think its a question of small cumulative air leaks in the return lines.

The fact that hydraulic oil escapes from a high pressure point out into free air, makes for oil foam build up.
Remember that in this context good and leakfree return lines are to be considerd as 'free air' - compared to the situation from where the oil escapes.
Best example is a HC, where the HC rod (valve) has moved to release pressure from the suspension. If you try find one of these return lines, then lead it separately to a jar, you will recognize this oil foam ...

The foam is air bubbles trapped in the oil by its surface tension.
If this foam is allowed to recycle - even any number of times - through the pump - then you can imagine what happens after some driving : no hydraulic component will act quite as predictable as it should ...
Because hydraulic oil mixed with that amount of air does no longer act as a pure fluid !

The 'fluid' is then compressable - because of the air