Air bubbles in the LHM gives harsh ride?

This is the Forum for all your Citroen Technical Questions, Problems or Advice.

Moderator: RichardW

Post Reply
pprado
Posts: 57
Joined: 02 Oct 2006, 02:29
Location: Campinas - SP - Brazil
My Cars:

Post by pprado »

<brainstorm mode>
My "bubble rate" was quite high when turning the wheel before the "hack".

I think that the power steering works by directing pressure to one or other side of an actuator who has a piston inside of it.

When straight ahead, both sides of the actuator are connected to the reservoir.

When you turn, say, to the right - pressure is connected to the left side of the actuator and the right side is connected to the reservoir. When to the left the opposite happens.

Say a tiny hole exists in some place on the actuators. Isn't it easier for air to pass on a tiny hole than LHM? That way, when you get "negative pressure", air could be sucked in...

I am really not concluding anything as I don't know the correct designs, but I thought someone could know and comment.
</brainstorm mode>
-----
Pedro Prado
User avatar
Mandrake
Posts: 8615
Joined: 10 Apr 2005, 17:23
Location: North Lanarkshire, UK
My Cars:
x 664

Post by Mandrake »

jorgy wrote:Hi all

quick intro, I'm fighting the same problem 2 years now.
Hi George,

Thanks for joining the frey... the more heads working on this problem the closer we might get to solving it. :D
Hi Steeve, as u know I drive an XM Hydra II an it has the problem. Agree with Prado that tyres are not the problem. My XMs behavior is volatile -as described by others-. I've had it working perfectly (as it should) for a day and was on same tyres.
Yes, its definately not tyres. I've had the ride go from great to awful or the other way within a single half hour drive before, so it can't be tyres.
Now I've eliminated all electronic/electirc possibilites, and tuned pressure regulator to correct pressure range 145-170 bar (with gauge), and even fitted a one-way valve to feed the front e-valve to make sure it doesn't suffer from pressure loss from elsewhere in the system.
Great - that saves me from trying that, I was planning to try a one way valve on the electrovalve feed, but couldn't find a suitable one.
All spheres equally measured + they r the correct ones. Height also checked and adjusted, also changed bronze bit on linkage that gets sticky on XMs-.

My thoughts curently are: major probability, this problem is related to an abnormal loss of pressure somewhere (which possibly creates the bublles in LHM and also quickly destroys its mechanical qualities as oil overheats at the point of leakage).
I guess you're taking a cue from the second article you linked to ? The problem I can see with this is its part of the normal function of the suspension to release pressure from the suspension rapidly through small pipes and valves. (Height correctors, electrovalves etc) Even if it was causing trouble it would be in the form of accelerated degradation of the oil over weeks/months rather than a day to day on again off again ride quality. (IMHO)
My reg. clicks every 13secs so I have a loss somewhere -I'll check returns and find out soon, and replace relevant component-. Question: are all sufferers' cars happening to suffer from low cycling time?
No! Mine has long regulator cycle times.

Short regulator cycle times (assuming the accumulator sphere is good) are an indicator of leakage somewhere in the systems coming off the main pressure supply line. This additional fault may or may not induce additional ride problems over and above this fault.

(There is nothing to say that you have only ONE fault! :? )

A prime candidate (but not the only possible culprit) is the Hydractive 2 Electrovalves. If they are a little bit leaky they will continue to function ok but cause a short regulator cycle time, (and a few other incidental symptoms) but if they are very leaky they will start to upset the selection of soft mode.

Having said this, since trying the return pipe test, the regulator cycle time is longer, but it was never anywhere near as short as 13 seconds. It's probably improved it from about 1 minute to over 2 minutes. (Quite why it would effect it, I don't know, nor why it effects the stop light going out time dramatically)
If not *indirectly* LHM-related as above, my other bet is on sticky struts -yet undocumented on XMs-, but bit unlikely as I had my car being a magic carpet for 2-3 days over last 2 years with those same struts.
Well, again struts can be an *additional* source of poor ride over and above this fault. As can droplinks, lower arm bushes, rear arm bearings etc...

However I replaced the struts (the actual rams) on mine fairly early on when I got the car because I was convinced that they were a major source of ride harshness, and - surprise surprise - the ride was really good for a week or two then deteriorated, and later deteriorated to such an extent that I realised the struts were in fact nothing to do with the ride harshness problem. :evil:
FIY I found out that best practice in hydraulic systems is oil return being under oil, not above it -any spilalge creats foaming-. I append the links.

http://www.mcnallyinstitute.com/10-html/10-10.html

http://www.lube-tips.com/challenge/2003-01-22.asp

XM tank looks well designed according to above references, important point seems easy LHM sunction from pump ie: clean and good tank filters, airproof LHM sunction pipe, tank breather (filler) cap not stuck (subpressure will be created and pump may tend to suck from other points except of LHM tank).
All good points, except we've been there done that, got the t-shirt and still looking for the problem :?

Over the last two years I've replaced the suction hose, the pump itself, and the filters and tank have been cleaned twice with new LHM both times. The problem persisted.

What it now seem to be is that SOMEHOW air is being drawn into the various return lines by the flow of oil returning to the tank. Normally a small quantity of air will be trapped by the return filter and prevented from recirculating, however it seems that if a large enough quantity of air is sucked in it is able to pass straight through the filters and be sucked back in by the pump. Once this happens it tends to recirculate over and over, and won't clear itself out.

Having the oil return to the tank under oil level is fine if the returning oil doesnt already have any air in it, but if it does, you now have the returning oil going into the tank at the same depth as where the pump draws oil from and the bubbles and froth are sucked straight across to the pump.

What we're trying to do is arrange things so that any air bubbles returning can't find their way to the pump, by letting the oil return at the surface but (hopefully) without causing too much additional frothing on the surface. If a few bubbles break at the surface it doesn't matter as long as the oil deep below is bubble free.
Relevant cases I found all over the English and French net: a CX where a guy had similar (?) problem and he opened up struts to find a "syrup" substance which he cleaned and car became soft as it should;
Interesting considering the CX doesn't use struts :wink: It uses double wishbones and a seperate hydraulic cylinder connected by a pushrod. The cylinder on the CX is fully self lubricating, so there is no direct comparison with a McPherson strut as used on an XM or Xantia.
also a BX guy found an LHM filter had a whole and putting new filter sorted him out.
This sounds like the story that Anders described with a torn filter - this one does fit in with what we are discovering - the filter plays a key role in preventing air recirculation, and a torn filter will let the bubbles straight through.
That's all very quickly, I'll keep you posted as I intend to fix components that need fixing on mine (prob new brake valve) and buy new tank filters and put LHM (I m on hydraflush now). Will even remove and clean hydractive blocks if I get no result -suspect the valves+ball in there might be sticky as CX case above, but difficult for hydraflush to clean this-.
Let us know how you get on.

Regards,
Simon
Simon

1997 Xantia S1 3.0 V6 Auto Exclusive in Silex Grey
2016 Nissan Leaf Tekna 30kWh in White

2011 Peugeot Ion Full Electric in Silver
1977 G Special 1129cc LHD
1978 CX 2400
1997 Xantia S1 2.0i Auto VSX
1998 Xantia S2 3.0 V6 Auto Exclusive
bernie
Posts: 882
Joined: 10 Apr 2001, 02:25
Location: Southampton United Kingdom
My Cars:

Post by bernie »

jorgy wrote:
ps. Bernie your tube's "nerves" instinctively don't look best for flow?

Sorry jorgy but I don't understand what your saying. :?
3 Fiat 124 Sport 1969x2, 1968
2 Fiat 124 Spider 1976, 1971
1 Fiat 20VT Coupe Plus
BUT maybe moving to France
User avatar
Mandrake
Posts: 8615
Joined: 10 Apr 2005, 17:23
Location: North Lanarkshire, UK
My Cars:
x 664

Post by Mandrake »

I think he's refering to the effect the ribs on the inside of the tube might have on smooth non-turbulent oil flow ? (Personally I don't think it would be an issue)

Regards,
Simon
Simon

1997 Xantia S1 3.0 V6 Auto Exclusive in Silex Grey
2016 Nissan Leaf Tekna 30kWh in White

2011 Peugeot Ion Full Electric in Silver
1977 G Special 1129cc LHD
1978 CX 2400
1997 Xantia S1 2.0i Auto VSX
1998 Xantia S2 3.0 V6 Auto Exclusive
pprado
Posts: 57
Joined: 02 Oct 2006, 02:29
Location: Campinas - SP - Brazil
My Cars:

Post by pprado »

I´m thinking now in doing my "v2" that would be a "J-tube".

I would keep both L-joints and would leave a small section of the last tube, just to make sure the oil flows upwards - but it would still be submerged on oil.
That way I hope the flow itself will take the bubbles to the surface top.

I am however more inclined to build bernie´s approach as it is certainly nearer from a final non-beta version... it keeps the filters in place :-)

I still have to see if only blocking the lower portions of the return filter also does the same trick. I noticed that the reservoir cover has bypass holes over the filter - most probably to be used only if the filter gets clogged.
-----
Pedro Prado
MikeT
Posts: 4809
Joined: 11 Jun 2007, 16:17
Location: Christchurch, Dorset. UK
My Cars: 2005 C5restyle 1.6HDI 16v 110hp VTR Estate
2008 C5 X7 1.6HDI VTR+ Saloon
x 231

Post by MikeT »

I'm spitting in the wind here and admit I have no idea what principles you're working to and what design restrictions you're facing but the air in fluid got me thinking about a free floating pipe. (Sorry if I'm way off the mark...)

If there will be air in the return fluid, given a light enough piece of pipe, if left unfixed, would it not regulate itself to float to just below the surface regardless of level? Additionally, could the air be teased out from a series of angled/vented/baffled holes near the end or through a small foam/sponge filter?
User avatar
Mandrake
Posts: 8615
Joined: 10 Apr 2005, 17:23
Location: North Lanarkshire, UK
My Cars:
x 664

Post by Mandrake »

pprado wrote:I´m thinking now in doing my "v2" that would be a "J-tube".

I would keep both L-joints and would leave a small section of the last tube, just to make sure the oil flows upwards - but it would still be submerged on oil.
That way I hope the flow itself will take the bubbles to the surface top.
The problem with this idea is that the oil level in the tank is not constant - and varies from very low to very high depending on the suspension height. :?
I am however more inclined to build bernie´s approach as it is certainly nearer from a final non-beta version... it keeps the filters in place :-)
Yes, I like bernies approach, as soon as I can find a piece of suitable pipe, (and a fine weekend!!!) I'll be trying it. I'm unhappy about running mine for too much longer with one of the filters effectively bypassed.
I still have to see if only blocking the lower portions of the return filter also does the same trick. I noticed that the reservoir cover has bypass holes over the filter - most probably to be used only if the filter gets clogged.
I don't think it would be a good idea to block any of the filter - it wouldn't be able to cope with the return flow rate at higher engine revs, even with a clean full sized filter the oil level inside the filter is 10-20mm higher than the tank level at idling speed.

Good point about the bypass holes at the top - that could be a problem if an upwards firing pipe pointed at the holes - not all of the oil would be passing through the filter. :(

Regards,
Simon
Simon

1997 Xantia S1 3.0 V6 Auto Exclusive in Silex Grey
2016 Nissan Leaf Tekna 30kWh in White

2011 Peugeot Ion Full Electric in Silver
1977 G Special 1129cc LHD
1978 CX 2400
1997 Xantia S1 2.0i Auto VSX
1998 Xantia S2 3.0 V6 Auto Exclusive
jorgy
Posts: 27
Joined: 04 Jul 2007, 01:02
Location: Glasgow
My Cars:

Post by jorgy »

MikeT wrote: could the air be teased out from a series of angled/vented/baffled holes near the end or through a small foam/sponge filter?
The sponge/filter is a good idea but this is what the filters included are supposed to do. Perhaps with age though their material is getting used and they loose efficiency in removing the bubbles? That's why I 'll order a new set of them anyway, they r not that expensive. Has anybody put new filters ratehr than washing old ones?

regards
G
1994 XM V6 12v manual
jorgy
Posts: 27
Joined: 04 Jul 2007, 01:02
Location: Glasgow
My Cars:

Post by jorgy »

Another test would be to syphon out a full quantity of LHM that has provided good ride the first days after being put in the car, and then put it back later. If it provides again a good ride that also deteriorates again after a period, we are even more certain that the problem is LHM and bubble-related. And that its chemical/mechanical characteristics have not been permanently altered by something in our cars.

G
1994 XM V6 12v manual
jorgy
Posts: 27
Joined: 04 Jul 2007, 01:02
Location: Glasgow
My Cars:

Post by jorgy »

I'd like to share with u another characteristic of my "crashy" ride, which I can't think how it could relate to bubbles in the LHM: if a wheel of my XM falls in a hole singularly -ie only one side- then it is obvious -even to the eye- that it is the car that falls in the hole, rather than the wheel moving down and following the road surface. While I can perhaps hypothesise that air in LHM would alter the compression characteristics of the suspension in a bump, I just don't know how it could affect the motion of the wheel downwards. The way my car rides actually is absolutely similar to a car that would have a very stiff antirol bar, or one were the struts would be very sticky, or, being specific to Hydractives, one where the middle sphere is not permanently and entirely *in* circuit (when it should be). Does this sounds like your cars?

On my car, the one-way valve I've put to drive high pressure in the front hydractive block, should guarantee that the hydractive valve remains fully open, but air-in-LHM complicates things in a way that I cannot predict. I think the ultimate test left to be done is find another hydractive block, clean it, and block that valve to fully open position by means of inserting a piece of tube (can be done). I'd be amazed if the ride remains crashy then....If anybody can do this trick faster than I, please don't hesitate -I can send pictorials of what exactly to do-.

regards

G
1994 XM V6 12v manual
jorgy
Posts: 27
Joined: 04 Jul 2007, 01:02
Location: Glasgow
My Cars:

Post by jorgy »

...forgot to say there's a brand new e-valve on my front block as old was found leacking -visibly, even at idle-. So, for my front system, high-pressure being guaranteed to the evalve, only really unknown quantity is the cleanliness of the block itself -ie if the 2 ways available for LHM to cross-flow are well open-. This is assuming my front struts are not sticky, as they've offered the magic-carpet ride on a couple of occassions before. And assuming that flow to the thick pipes feeding the front struts are also well open -can't imagine them being blocked from dirt-. Correct me if there's any other point in the system's architecture that I have ommited.

G
1994 XM V6 12v manual
pprado
Posts: 57
Joined: 02 Oct 2006, 02:29
Location: Campinas - SP - Brazil
My Cars:

Post by pprado »

MikeT wrote:I'm spitting in the wind here and admit I have no idea what principles you're working to and what design restrictions you're facing but the air in fluid got me thinking about a free floating pipe. (Sorry if I'm way off the mark...)

If there will be air in the return fluid, given a light enough piece of pipe, if left unfixed, would it not regulate itself to float to just below the surface regardless of level?
Hi, we are working on a fluid reservoir on a cilyndrical section with about 20cm of diameter. The floating pipe does not seem a bad idea at all, but maybe difficult to implement on that small space.
Additionally, could the air be teased out from a series of angled/vented/baffled holes near the end or through a small foam/sponge filter?
This system has 2 filters - one for the return flows and one for the pump flow. Certainly the return filter has also this function - and it is not doing it proudly, if it was the bypass-to-air would not make appreciable effects...
-----
Pedro Prado
pprado
Posts: 57
Joined: 02 Oct 2006, 02:29
Location: Campinas - SP - Brazil
My Cars:

Post by pprado »

I don't think it would be a good idea to block any of the filter - it wouldn't be able to cope with the return flow rate at higher engine revs, even with a clean full sized filter the oil level inside the filter is 10-20mm higher than the tank level at idling speed.

Good point about the bypass holes at the top - that could be a problem if an upwards firing pipe pointed at the holes - not all of the oil would be passing through the filter. :(
Oh well... solution on this case appears to be a replacement filter that is fatter and closed on the lower portions. That is not easy to "beta test"...
-----
Pedro Prado
pprado
Posts: 57
Joined: 02 Oct 2006, 02:29
Location: Campinas - SP - Brazil
My Cars:

Post by pprado »

jorgy wrote:Another test would be to syphon out a full quantity of LHM that has provided good ride the first days after being put in the car, and then put it back later. If it provides again a good ride that also deteriorates again after a period, we are even more certain that the problem is LHM and bubble-related. And that its chemical/mechanical characteristics have not been permanently altered by something in our cars.
Jorgy I think that the fact most of us feel a large improvement after any maintenance proves the LHM has not lost its characteristics...

This test can be done but me at least am convinced the fluid properties are not the problem, at least not the major one.
I'd like to share with u another characteristic of my "crashy" ride, which I can't think how it could relate to bubbles in the LHM: if a wheel of my XM falls in a hole singularly -ie only one side- then it is obvious -even to the eye- that it is the car that falls in the hole, rather than the wheel moving down and following the road surface. While I can perhaps hypothesise that air in LHM would alter the compression characteristics of the suspension in a bump, I just don't know how it could affect the motion of the wheel downwards. The way my car rides actually is absolutely similar to a car that would have a very stiff antirol bar, or one were the struts would be very sticky, or, being specific to Hydractives, one where the middle sphere is not permanently and entirely *in* circuit (when it should be). Does this sounds like your cars?
My car does the same - when both wheels are involved the ride is good but when only one is normally I feel and see the car following the hole.

Hydractive cars have a setup to avoid too much flow from going from one side to other and they do have a stiff anti-roll bar. If we got a hydractive block that is going to hard when it shouldn't, this may happen...
On this case I am more inclined to think the air is making bad things on the HA block rather on the struts/hard spheres circuit.
On my car, the one-way valve I've put to drive high pressure in the front hydractive block, should guarantee that the hydractive valve remains fully open, but air-in-LHM complicates things in a way that I cannot predict. I think the ultimate test left to be done is find another hydractive block, clean it, and block that valve to fully open position by means of inserting a piece of tube (can be done). I'd be amazed if the ride remains crashy then...
Have you tried the bypass-to-air test? It's very simple to do Simpn's test and it's well worth doing... And on his case the symptoms appear to be lasting!
-----
Pedro Prado
User avatar
Mandrake
Posts: 8615
Joined: 10 Apr 2005, 17:23
Location: North Lanarkshire, UK
My Cars:
x 664

Post by Mandrake »

pprado wrote: Have you tried the bypass-to-air test? It's very simple to do Simpn's test and it's well worth doing... And on his case the symptoms appear to be lasting!
Yes it does appear to be lasting. Occasionally the ride is a little bit worse but then it quickly improves - usually after the car has been parked a while the next time I drive it the ride will be good again without any Citrerobics - previously it would only ever deteriorate with normal driving and you had to do Citrerobics to get any kind of improvement, and temporary at that.

Unfortunately in the last few days my car has developed what looks like a faulty Hydractive 2 steering wheel position sensor, so I may not be able to provide much feedback on the air bubble performance until I fix that. :evil: :evil: :evil:

One interesting fact is that even stuck in hard mode above 25Km/hr the ride is not "harsh" as before. It is a bit fidgety and slightly bouncy on the tyres yes, but not filling rattlingly harsh as before...

It never rains but pours with Citroen's :roll:

Regards,
Simon
Simon

1997 Xantia S1 3.0 V6 Auto Exclusive in Silex Grey
2016 Nissan Leaf Tekna 30kWh in White

2011 Peugeot Ion Full Electric in Silver
1977 G Special 1129cc LHD
1978 CX 2400
1997 Xantia S1 2.0i Auto VSX
1998 Xantia S2 3.0 V6 Auto Exclusive
Post Reply