Air in power steering pump

[Benz Dr.]

Also Hydraulics 101:  If a pump creating fluid flow moves fluid that has nowhere to go then the resultant is fluid under pressure acting upon all surrounding surface areas within the system.  I never said or eluded to any suggestive commentary that would lead one to believe that a pump creates pressure, directly or indirectly. 

You stated earlier, "I cannot see, how moving the piston in the gear will "suck" fluid out of the reservoir".  If after explaining to you why moving the piston DOES take fluid out of the reservoir you still can't comprehend then I guess there isn't much we can do to convince you.  It may not take the fluid out of the reservoir on the high pressure side.....it likely takes it in on the return side with the down stroke of the piston.  The fact remains that no matter how the fluid is taken in it still displaces air and achieves the intended goal.  With your knowledge you should be able to make sense of this, especially after your having further elaborating on what I had already stated.  I feel that if you'll keep an open mind and just try one of the aforementioned methods you'll see that what was suggested does work....  So well, in fact, that if done properly you can literally start the car an go with 100% power assist right away.  Still don't believe it?  Well, again, try it!  I learned a long time ago that the physical properties learned in the mechanics of fluids classes is pretty black and white, but the manipulation of fluids is not.....and this is exactly why there is more than one way to bleed this particular hydraulic system, be they conventional or not. 

And yes, this post is about bleeding the power steering system.  The original poster is probably very lost right about now.  So I'd like to apologize to the original poster for having hijacked his thread.   Back you you, Abudhabi.

Just imagine Aaron. 10 hour drive to get here and all the snow you can eat.

[450sl]

Thanks Mike, will do the engine off -stop to stop procedure, and check my tubings.

Cheers
Mark

[w113abudhabi]

Thank you for all the responses and apologies if I have started a difficult dialogue.

I have shared all comments with my mechanics who have successfully removed all air from the power steering system. The pump has been in and out the car several times in the past weeks and issues remain. I will summaries below some key points:

1. The power steering was functioning well but the pump had fluid leaks
2. An original mercedes seal kit has been used to prevent the leaks
3. There are no visible signs of wear or damage to the pump whatsoever
4. Following advice from this forum it is believed that all air has been purged from the system
5. Whilst removed from the car the pump has been pressure tested and holds good pressure

However, there is no power assisted steering at below 1200 rpm. Once the revs exceed that level the steering is power assisted.

Can anyone please shed any light or suggestions?

[MikeSimon]

There are obviously  several more people here who know more about the power steering system than I do, and I am sure, they will chime in. In the meantime, I will make a comment.

The flow output of the power steering pump, as I said, is direct proportional to the rpm. That is the typical characteristic of the balanced vane pump.
If one would plot the output of the pump over rpm, you would see a rising flow until a certain rpm when the flow control point is reached. From there up in rpm, the flow remains basically constant, regardless of the back pressure.
In the lower rpm part, the "non-controlled flow" part, the output is sensitive to back pressure. If the back pressure increases (steering the wheels) the flow goes down. When designing a p/s system, the "dry park" condition is the key for picking the displacement size of the pump. At engine idle and back pressure at the parking maneuver, the pump has to put out enough flow to provide the torque in the gear to steer.
If everything is in proper condition as you describe, there could be two reasons the car will not steer;
1- the pump displacement is too small ( I want to rule this out as you said the pump was rebuilt and it is probably unlikely they put the wrong size rotating group in) -OR-
2- the internal leakage of the pump is too large. This could be caused by excessive wear on the sides of the rotating group, a missing vane, or excessive wear in the valve bore.

The proper diagnosis for the pump will be to test it for flow at low rpm and a back pressure of, let's say, 500-600 PSI (dry park condition)
If memory serves me well, you should get around 8-10L per minute (2.5-3 gpm). The pump building pressure at high rpm is not a sign for a good pump.
Most service dealers when hydraulic p/s was popular in cars had a simple test set-up for the pump. It consisted of an extra line with a flow control meter, pressure meter and a shut-off valve.
This line was connected in the pump output line between pump and gear. Then the back pressure was increased by slowly closing the shut-off valve until the pressure meter would indicate the desired back pressure. The reading of the flow meter would give an indication of pump condition.
If you do not get the required flow, the pump needs to be opened up and checked again.
Any hydraulic service shop can test that pump for you. Probably cheaper than the Mercedes dealer

Now take all this with a grain of salt and maybe wait for better explanation because as stated in previous posts, I do not have the right comprehension regarding these issues.
Also, we can take this off the forum, if you like to at mike.simon70@yahoo.com

[Cees Klumper]

At this point I would try installing a good used replacement pump.  They can't be that expensive, off Ebay or what have you.

[Aaron h]

Mike, it wasn't the intent of any of us to imply that you hadn't the ability to comprehend anything regarding the material in this post.....you're more than qualified, and we all recognize that.   All we were trying to do is prove to you that it is possible to rid the system of air without having to run the engine and pump.  The method Dr.Benz made mention of is a tried and true method that Mercedes dealers and private shops across the world have used for 50+ years without fault.  All we had hoped for was a little understanding and for you to give it a try.  Nothing more, nothing less. 

Regarding abudhabi's issue, we first need to make some corrections to some of the numbers and devices used for testing.  We'll first touch on flow rates for the Vickers winged cell pumps.  (Refer to pictures below) The pump used on Pagodas had a flow rate of 5.6 liters per minute (about 1.47 gallons per minute) at 500 RPM.   8-10L per minute seems really high.....or was that assuming there was no flow control or counter pressure in which the pump was moving fluid at it's maximum RPM?.   I realize that in the reference picture it denotes a W108.015, W109.015, and a W109.016. However, the superscript "1" is a foot note that states "And all other types optionally equipped with DB power steering.".  Obviously, that would include the W113.

For counter pressures, we need to know if the pump is on a test stand or operating on the car....two very different operating/test conditions.  The early BBB notes a TSB from Vickers that says "At a counter pressure of 50 ATM (735 psi) and at 500 RPM the minimum delivery of the pump should be 5.7 liters per minute (1.5 gallons per minute). However, these values can only be measured on a test stand."  In the later BBB Mercedes took this into account and revised the circulating pressure testing procedure to reflect the tests of the pump being done with the pump on the car.  Though, in the early book Mercedes still shows how to perform tests on the pump with it still on the engine and connected to the gearbox.  This brings us to the next item of interest......the test jig. 

The testing device only consisted of a throttle valve and a pressure gauge.  The pressure gauge gives us the pressure readings, and the throttle valve acts as the the counter pressure. There was an early and a late set up of the test jig. (see pictures below).  Both achieved the same result.  However, there was no flow control meter, as it wasn't necessary due to the pressure gauge being present and in use throughout the entire test, as short as the test was.  Though, a flow control gauge would be nice to integrate for peace of mind.  Regardless, it was a simple set up.  The hydraulic line going between the pump and the gear box was removed at the pump.  The gauge side of the jig was connected to the pump, and the throttle valve side of the jig was connected to the end of the hydraulic hose that was removed from the pump. 

After the oil was topped off, the throttle valve was opened, the engine was started and circulating pressures were noted from the pressure gauge at idle speed (700-800 RPM).  Circulating pressure should NOT exceed 4 atü (71.1 psi).  During this time the steering wheel is not to be turned, as the control valve in the gear box would be displaced causing pressure to increase.  Next, the throttle valve was closed to measure the maximum pressure of the pump.  This was done by revving the engine up to 1,000 RPM and holding it there while noting the reading on the pressure gauge.  The maximum pressure should be at LEAST 55 atü (796.5 psi), but should NOT exceed 65 atü (938.7 psi).  Alternatively, maximum pressure can also be measured by turning the steering against left hand and right hand lock.  Regardless of which test for maximum pressure is chosen, make the test as quickly as possible, as running the pump too long with the throttle valve closed or steering locked in one direction or the other can cause the oil to get too hot and cause potential damage to the pump itself.  Also, note in the pictures below that some of the pumps have the maximum pressure printed on their identification tag. 

So, to review all of the aforementioned in short:

-Flow rate of pump is 5.6 liters per minute (1.47 gallons per minute)
-Circulating pressure should not exceed 4 atü (71.1 psi)
-Maximum pressure should be at LEAST 55 atü (796.5 psi), but should NOT exceed 65 atü (938.7 psi)

Oddly, none of the literature gives us a minimum circulating pressure value.  Mike, in your professional opinion, what would you consider a minimum circulating pressure value to be?  Or, what minimum value do you think would deem the pump faulty? 

[MikeSimon]

Aaron: The flow rate for the pump at 500rpm is a Vickers pump spec and was given to determine the correct displacement. It is used on a test stand. The engine idle in a car is not 500 rpm. I based my flow rate on an engine idle of 850-1200 rpm to test the pump at real world conditions. The VT27 used by Mercedes Benz in early p/s was, I believe, a 13cc unit. As the flow is direct proportional to rpm, following the Vickers spec at 1000 rpm, it would be 11.2 liters. (Theoretical flow would have been 13L/min)
Pumps have an internal leakage, so mathematical calculations are not always 100% correct. As I said, back pressure affects the pump performance below flow control greatly.
The required flow rate of the pump for a steering system is based on the flow required to achieve 1-1.5 steering wheel turns per second. It is not the same as the 5.6L given as the spec number.
As far as the circuit pressure at closed steering valve (straight ahead) is concerned, for layout purposes we always assumed it to be between 30-50 PSI, depending on the length of the hose routing and the size of the steering gear.
Vickers always stamped the maximum pump relief pressure on the nametag of the pump. This is the pressure value the internal relief valve is set at. It can vary for the same pump model depending on the application.
As far as the p/s test kits are concerned there are a few different ones around. I am not familiar with the one MB used at their dealers, but we used one with a dual meter, flow and pressure.

Interesting side note, as we have the "different world words thread" elsewhere: The pump in the Mercedes Benz literature is called a "winged  cell pump". This is a literal translation of the German term "Flügelzellenpumpe". Although Vickers - being a US company - used the proper English term "Vane Pump"

[w113abudhabi]

Problem solved. Thank you to all for their help and recommendations.

It turned out that the control valve had been reinstalled in the pump the wrong way around resulting in power steering only being available when the revs exceeded 1000rpm.

The following linked thread and attached file provided the necessary information.

https://www.sl113.org/forums/index.php?topic=20219.0

Thanks again!

[Cees Klumper]

Congratulations, a very simple cause. And we got to watch an intellectual debate between experts to boot!

Must say when I read a post in the thread you referenced, where Johnny quoted me as having contributed probably 10+ years ago on this subject of bleeding the power steering pump, I have zero recollection of that.