Workaround for bubbles in LHM return flow

[aerodynamica]

Hello, I've had some experience with this on 2 Xantias and 2 XMs. On one XM it was caused by a near invisible crack on the LHM tank plastic pipes/filers housing. I replaced the HP pump suction pipe with a clear one (yachting bilge pipe) that clearly showed the bubbles flowing to the HP pump. I think this was a rare one off though.

The other XM and 2 Xantias had no such air leak but did have bubbles in the LHM tank seen looking into the LHM when the suspension was lowered and/ or the 12mm bleed on the PRV opened. It foamed up with bubbles. I attempted to flush and bleed and re-prime the HP pump suction over several weeks and had changed the spheres all to no avail - bubbles returned. I did not understand it! Then one day I was replacing the LHM in the XM and noticed how much bubbling appeared from simply topping up the LHM fluid level. The action of the LHM stream hitting the LHM surface causes bubbles despite there probably being anti foaming agent in the LHM. Anyway, it gave me an idea- after reading on here or one other Citroen forum about a member who added the pipe to the return circuit filter housing of the LHM tank and reported success, I realised the problem was caused by the return LHM coming back into the tank.

The main LHM return from the PRV has a long, deep nozzle that reaches straight down through the surface of the LHM in the tank (albeit with a long slit down most of its length). This I reckoned was to prevent churning, foaming etc. But the 3 return connections next to this main return from the brake valve, suspension, and PAS etc don't. These 3 have no such extension down through the susface and the return LHM falls straight down to the surface and causes bubbles that then get sucked up by the HP pump suction despite the baffles in the LHM tank.

I think the guy who had modified the return filter block had fitted a flexi tube to the return from the main return from the PRV. I don't think this alone works, I fitted an extension from the suspension return next to it by fitting a push-fit white nylon windscreen washer tube-sized pipe connector into the return pipe outlet on the underside of the filter block (nylon return filter removed) and then cut a length of windscreen washer tubing, about 50mm and pressed it on the end. It has to be short enough to let the nylon filter be refitted and the new tube is stays inside the filter. All the tube does is extends the return outlet to below the surface of the LHM in the tank. It makes the return LHM flow from these little return pipes flow out below the LHM surface.

Result - no more bubbles at all! It was an instant change in performance. Lowering the suspension from maximum height to minimum sees the LHM level softly rise in level without a single bubble. I dont have the PRV return modified at all - just this suspension return and the PAS return and there are no bubbles now

Try this it's quite easy!

Beware of pushing the nylon pipe connector too far into the return port outlet as it can block the return flow and causes an LHM explosion... (how did I find this out .... )

Graeme

[xantos]

So we might be catching our tail at the wrong side

It makes sense that the tube that extends to the bottom of the LHM filter disperses the bubbles before they reach the bottom. Might be worth trying as it is not a difficult mod.

So regarding my mod... I've done this.



I've put a pipe directly into the reservoir so that I've ruled out bad connection and broken plastic where the suction pipe is connected to the LHM reservoir. I've cut the end of the pipe four times "V" shape so it can't stuck on the bottom of the filter. Had to also modify the filter and done "10 min mode". Also had to improvise with a bit of wire that the filter doesn't fall down. I know it doesn't look nice







But... No real improvement. Bubbles still remain in the reservoir... So the next thing to investigate is the connection to the HP pump and the HP pump itself... The pipe in the picture has a 10mm diameter and the HP pump connection has a 9mm diameter. So it's not a snug fit. And also noticed that the HP pump itself is a bit LHM wet. Not drastically but made me think that it probably sucks air into the system.

Sooooo first thing I will try this mod that @aerodynamica mentioned and then reseal the HP pump. And then I will probably catch my tail (no probably not )

[white exec]

Hi Graeme,

I remember your earlier report of a hairline crack in the reservoir unit. As you say, hopefully a one-off.

Was intrigued by your observation of bubbles forming simply by topping up the reservoir with fresh LHM, which suggested that the pouring action alone was sufficient to produce bubbles. Implication is that a similar thing might be going on when the system returns fluid to the reservoir.

So, have just taken a fresh bottle of LHM, unshaken, and slowly poured about 1/4 litre of it into a clean glass jar. Apart from a few large bubbles appearing at the surface, which quickly burst and disappeared, I was surprised to see hundreds of very tiny bubbles throughout the fluid. Being extremely tiny, these only rose very slowly - so slowly, they were, in effect, hanging stationary in the fluid. I repeated the test, and saw the same thing.

Then took the 1-litre LHM bottle, and shook it half-a-dozen times, to roughly simulate fluid squirting into the reservoir. When this was poured into a clean jar, the number of micro-bubbles had increased many times over, and again just hung there.

Do wonder, then, whether such aeration might be happening in the car. Might it explain the repeated reports of suspensions becoming less comfortable as journeys progress?

Such very small bubbles might also be difficult to see if an inspection section of less-than-perfectly-clear plastic tubing is being used to monitor the fluid flow. Larger bubbles (possibly the result of air ingress through a loose joint) will be spotted, but these tiny fellas need a clear container, and a bright light behind, to be seen.

A better test for their presence might be to take a careful dip-sample of reservoir fluid with a small clear container, and check this.

Also wonder whether poured LDS does the same thing?

[CitroJim]

White exec, that's incredibly interesting!

Great research

[aerodynamica]

But after the mod of extending the small returns down I am no longer getting any bubbles! It's as if the only cause of the bubbles was due to the fluid returning from the suspension and PAS returns.

I think the baffles in the LHM tank do prevent the fluid from sloshing hard during driving.

One caveat though: I had found this phenomenon while I still had my CX, I looked for the same think in the CX LHM tank and...... nothing! there was never any sign of bubbles and I don't know why! As far as I could see the CX has the same return pipe/ filters plastic block in the LHM tank albeit in white plastic not black. The CX LHM tank is of course, metal and circular but I don't think it has the baffles in it like on the XM/ Xantia.

Anyway, it's worth a try to do the mod. ( I think of the three small returns next to the big return from the PRV, the F+R suspension returns is the closest one and the PAS is the furthest from it) I didn't add the pipe to the brake return as I felt it wouldn't be returning much LHM or very often. Also, oddly, the diameter of one of the three return ports is slightly larger and so the push-fit connector doesn't locate. Beats me as to why the brake one would be like, 0.25 mm wider...

[xantia_v6]

I have not seen the purpose of the "filters" discussed in the context of this issue. It is my belief that (despite inferences from various sources, including Citroen) the "filters" in the LHM tank were not designed to catch dirt or debris, but were intented to act as "bubble breakers" to prevent foaming of the LHM.

A bubble will not generally travel through a mesh screen, unless the mesh size is bigger than the bubble, any bubbles thus caught should accumulate on the front face of the mesh, merge and rise to the top of the filter.

LHM taking a round trip through the tank should be passing through 2 layers of bubble breaker mesh, so should be bubble free by the time it gets back to the pickup tube.

[Stickyfinger]

I have not seen the purpose of the "filters" discussed in the context of this issue. It is my belief that (despite inferences from various sources, including Citroen) the "filters" in the LHM tank were not designed to catch dirt or debris, but were intented to act as "bubble breakers" to prevent foaming of the LHM.

A bubble will not generally travel through a mesh screen, unless the mesh size is bigger than the bubble, any bubbles thus caught should accumulate on the front face of the mesh, merge and rise to the top of the filter.

LHM taking a round trip through the tank should be passing through 2 layers of bubble breaker mesh, so should be bubble free by the time it gets back to the pickup tube.

Totally agree with this

[white exec]

Having done the pouring test earlier today, and observing the microbubbles created, I then wondered whether they could be contained by the very fine mesh screens in the tank. Thought I had my old mesh filters left from when I renewed them a year or so back, but they must have been thrown out.

Next logocal step would have been a clear jar test with a mesh filter in it, to see whether bubbles were contained.

Was always slightly puzzled by the LHM tank needing two filters; would have thought that one, on the pick-up side, would have been sufficient to catch dirt. Interesting, too, that the 'return' filter has a very large surface area - maybe to allow a large volume of bubbles to be held back and disintegrate.

Could someone with a spare LHM filter test this out? Shake up a bottle of LHM, and pour into an upturned filter standing inside a clean glass jar . . .

Mesh size probably critical. Interesting.

[lcwin]

Sorry Guys I don't know how post pic here
It is Very Important to make 100% sure there is no air leaking from the Tank inlet to Pump inlet pipe.
If you don't do these you will be on a wild bubble chase....HaHaHa...
make sure inlet tube at main tank is not damage inside ( most of the time this is damaged but you won't see it ).
Each time you plug in a new inlet tube a very deep tear is gouge into the inside and as you pull the pipe out for maintenance the raised edge of the tank pipe cuts the rubber inlet hose again. This is the main culprit that cause all the bubbles issues !
As I said earlier carefully file down the edge so that the inlet tube can glide in with some effort thats all . Then get a " proper" round spring clip to fasten it.
Same goes for the pump side but you need not file the edge as it is not very sharp.

On the issues about intake bottom sucking up bottom particles
Citroen put it very low to make sure during cornering and low LHM level the hydraulic pressure is still there - priority to safety
If your filter have lots of black particles this is a symptom of broken plastic parts inside the front strut dampers
It quite easy to tell if the filter is clogged your steering will feels heavier that usual.

Fillter Mesh size is about 50 um and don't filter well. I nowadays use a 5 um filter with handpump to clean the LHM when I feel like it.
Hope my explanation helps

[myglaren]

There is a guide to posting pictures here.

[lcwin]

Thks myglaren.
Go here for pic of the tank inlet pipe.
look closely you will notice I have filed down the top edge to only about 1mm or less.
The fasterner on the left should never be used on the inlet hose. The one on the right is the reccommended one.

[Mandrake]

One caveat though: I had found this phenomenon while I still had my CX, I looked for the same think in the CX LHM tank and...... nothing! there was never any sign of bubbles and I don't know why! As far as I could see the CX has the same return pipe/ filters plastic block in the LHM tank albeit in white plastic not black. The CX LHM tank is of course, metal and circular but I don't think it has the baffles in it like on the XM/ Xantia.

I have a possible theory for that - different power steering design.

On a Xantia the vast majority (3/4) of the return flow back to the tank comes via the power steering. Xantia power steering is of "conventional" design where there is maximum flow when the steering is doing nothing and flow is restricted when you turn the wheel and put a load on the steering.

CX power steering on the other hand does not use any constant flow of oil when it is not doing anything. Both sides of the ram are closed and the steering rack hydraulically locked in place. When you turn the wheel oil is admitted to one side of the ram and some return flow comes back from the opposite side - but it is not much, and as soon as the ram stops moving both sides are shut off again. So most of the time you are driving there is no return flow from the steering.

And in relation to this general problem of bubbles, the GS didn't have any power steering at all...

[Mandrake]

I have not seen the purpose of the "filters" discussed in the context of this issue. It is my belief that (despite inferences from various sources, including Citroen) the "filters" in the LHM tank were not designed to catch dirt or debris, but were intented to act as "bubble breakers" to prevent foaming of the LHM.

A bubble will not generally travel through a mesh screen, unless the mesh size is bigger than the bubble, any bubbles thus caught should accumulate on the front face of the mesh, merge and rise to the top of the filter.

LHM taking a round trip through the tank should be passing through 2 layers of bubble breaker mesh, so should be bubble free by the time it gets back to the pickup tube.

I would agree in general - the large filter is there specifically to be a "bubble breaker", and in the normal operation of the system it does its job sufficiently well. (Otherwise they would misbehave from new)

Bubbles that come back through return lines (leaky low pressure plastic/rubber return lines etc) and bubbles that are introduced by the pouring action of some of the returns depositing their contents at the surface should all be contained by the filter and rise harmlessly to the surface and not make their way into the tank proper, outside of the return filter, and certainly not to the feed filter.

However it's been my experience that the mesh can only cope with so much - if the quantity of return bubbles becomes excessive and those bubbles become large, whatever the reason (and there are multiple possible causes) it starts to overwhelm the filter and some bubbles do start getting through, and yes they do get sucked up by the pump. As there is a constant circulation from the pump via both the pressure regulator return (most of the time) and the power steering its easy to see how bubbles that are already in the system can start to build up into a runaway state where the filter gets overwhelmed and the quantity of bubbles increases until it gets out of hand.

The basis of the 10 minute mod is that the main return is too far down at the bottom of the filter - and although there are two filters in between, the exit of the main high volume return flow is too close to the intake - it's only about an inch or two at most. In my opinion one mistake Citroen made here is there should have been a flat baffle between the two filters that is most of the width of the tank so that the oil has to pass a long way around the baffle rather than just travelling from one filter directly to the other. A kind of silly mistake in hindsight.

Cutting the main return pipe shorter brings the turbulence level higher up in the tank and seems to prevent those turbulence related bubbles from going down through the bottom of the filter, straight into the suction vortex of the pickup filter. Instead they swirl around a bit and come back up to the surface within the filter. The exit of the pipe still needs to be below the surface though, if you cut it too short you just create more "pouring turbulence". The 10 minute mod does not fix the problem at its source (excessive bubbles coming back through return lines and insufficient baffling between return and intake in the tank) but it does work around it to some degree.

Unfortunately I don't think there is any one silver bullet for this bubble problem - the causes of bubbles are numerous and some cars may have some problems (air leak in the pump feed line) some may have others, and some particularly stubborn cars may have multiple problems. So there is always going to be debate about what works and what doesn't (or even that nothing works) because although the end result - bubbles - is the same, the causes can be very different.

Some people have found what works for their car but that doesn't necessarily mean it will translate to other cars. On my current Xantia V6 (which I have not done the 10 minute mod on) I thought I'd try putting a little bit of black silicon sealant around the junction of the pipe and the spigot of the tank on the feed pipe to the pump, and also around the "join" in the pipe 50mm from the tank that the V6 models have. I didn't notice any immediate difference even after some Citrerobics but over a period of a few days the car has got significantly softer and more cushioned feeling and has stayed that way. I wouldn't say that it has completely solved the intermittent harsh ride I've been seeing but it has helped quite a bit.

[white exec]

Have to agree, Mandrake, with all you say - bubbles can arise from multiple sources, and can be slight or overwhelming.

In order to observe the bubble-blocking action of the filters, I have today ordered up a couple of new ones from Citroen, and will test and post when they arrive.

The hydraulics on the XM are working perfectly, and today we've driven about 90km on motorway, urban and winding mountain road. Dipped a small sample (from outside of the filters, obviously) of the LHM tank in a clear tube when we got home. Not a single bubble of any size to be seen.

What is essential, I think we all agree, is that the suction pipe between reservoir and pump is completely leak-free, and the advice on clip types and hose seating is really useful.

[lcwin]

In a proper functioning hydraulic system no bubbles should go into the pump and system. In fact a filter at the outlet to pump is not a recommended practice .
Any air bubble entering will leads to rapid pitting of hydraulic components. If your car hydraulic have small amount of bubbling during idling you should find the cause of it as soon as possible as it will leads to rapid wear of hydraulic component in a short time.
Air leakages at low pressure return pipes will usually have lhm seeping in the joints and this includes the front struts as well .
Same goes for power steering around the inlet/outlet pipes and steering ram. Wipe clean the joints and observe after a day of driving to see if it is wet . All high pressure pipes and component can be check by this method.
Like I have said earlier 90% of the time it is the inlet pipe to pump that is the culprit. You won't see wetting here as this essentially is a low vacuum hydraulic hose. When hydraulic pump is sucking hard at the inlet any restriction like partial clogged filter will build a higher vacuum situation at the intake filter silo tube which will pull in air from outside through very tight openings in the clamped hose. Another area air could go in is the round flat cap on top of the tank pipe going to pump. If you pries it open you will find an O ring there. Replace it with a new one.
If you have bubble issues go get some petrol or hydraulic hose and fit it in placed of the original pipe, replace the O ring and see if it helps. You got nothing to loose and it cost peanuts to try.