Best Place to Take Oil Sample from Hydraulic Circuit

Oil or hydraulics is not my specialty. That having been said, being that this is a test stand you should have a supply and return line that you connect to the equipment being tested.

It makes sense to me that you would want to protect whatever it is you are testing.

I would grab a sample at the supply line.

I know there are others on CR4 who specialize in this area and will shed more light on this subject.

_{Just my two pennies worth.}

If you are sampling for particulate contamination, take it before the filter. If you are sampling for chemical contamination, I don't think it matters.

My background is not oil analysis, but I've built lots of systems that included filtration ranging from sock filters to Ultra and RO filtration.

You should NOT pull oil directly from the reservoir.

Can you elaborate on why this should not be done?

Thanks

The oil in the reservoir is not a good place to take an oil sample from. You should always take an oil sample from an active part of the circuit to get results for the condition of the oil in the operating system. Reservoirs are designed to release entrapped air, and drop out heavy contamination to the bottom of the tank before slowly making it's way around to the suction side of the tank. There is always a baffle inside the reservoir to separate the returning oil, and the pumps inlet oil. Usually (but not always) you will filter the oil through a return filter before dumping the oil back to tank where it should sit for a while before returning to system. When you test the oil in the system you should be testing before any filters, and preferably after your pump and before your most sensitive component. In reality, you can sample oil anywhere you want, but you need to keep in mind why you are sampling, and what the allowable micron level is for the most sensitive piece in the system. If you manage to keep your oil as clean as the most sensitive piece in the system your system will last a long time (keeping temperature within range as well is also critical).

If you want the most accurate results, sample from a pressure line before any filters.

BE CAREFUL! just two weeks ago, one guy was killed and two others sent to the hospital via life flight about 10 minutes from here. They were working with a hydraulic system, and did not have the proper training. Oil under pressure can (and many times WILL) kill you if you do not know what you are doing.

This is the lab I use, they offer many different ways to take a sample with their recommendations. Do whatever makes the most sense for your application.

Thank you for the explanation.

I still don't quite see a rationale for not sampling from the reservoir. I can see that sampling from the bottom of the reservoir should be avoided as you may get an indication of excessively dirty oil due to the contaminates that have dropped out of the oil. But if your sample comes from higher up in the reservoir, shouldn't it be reprentative of what your pump inlet is going to see?

Not at all.

First, the pump inlet should only be a couple inches from the bottom of the reservoir, opposite the return line end, preferably on the other side of the baffle.

There are many different readings that can be taken when simply taking a sample from the reservoir. You could have horribly dirty oil (even when filtering your return line oil down to 3 micron) floating around in your reservoir on one end, and relatively clean oil in other areas. You will also get varying temperatures withing the reservoir(extremely important when dealing with hundreds of gallons of oil). The best way to obtain an accurate sample of the oil in your system, is to test the oil in your system, not the oil in the tank, or the oil that's been sitting in a hose, or any other place in your system. Sample the oil that is actively moving, and working in the system, preferably on the main pressure line.

If you wanted to test the water quality of a stream would you go take a sample from a backwash eddy full of stagnant water, or would you want to take it from the part of the stream which is actively flowing? I think that may be an easy answer, and is exactly what we are looking at here as well.

Okay...I understand. Thanks.

For this, it sounds as though the best sampling point would be in the return line just before it enters the tank. (This may be the same as you were describing as the discharge line.)

If you want to monitor contamination that the pump under test may be putting off, the return line before the filter would be a good place. If you are trying to monitor your test stand itself (the oil the tested pump is subject to), take the sample after the test stands pump (not tested pump).

All in all, it depends on what you are trying to monitor. overall system, pump, valves, tested pump, filter condition etc etc etc.

Thats a judgment call, depending on your particular situation. I pull oil after the pump before the pressure filter on my test stand, and it works very well. We always know that we are sending clean oil to the pump under test. We need to document that the oil we are using to test a customers pump is clean BEFORE it enters their tested pump. This reduces our liability that our test stand may have given contamination to the customers pump therefore causing damage. If you sample after the pump under test, you will be measuring the contaminants that the tested pump is producing, not the oil your test stand is providing to the pump.

what is purpose of the test stand ? and what checks are done on oil sample ?

Are you chaing machanical properties of pump ? volume pumped at certain presure ? in such case you need to know oil viscosity only along with inlet restriction discharge pr and volume.

Take two samples: one before tank discharge line, an another one just after the check valve placed on the pressurizad line (Pump discharge).

You will have good information.

Whitephone

Thanks all for your pertinent answers.

After the check valve, that would mean that we'd have to increase pressure to cracking pressure to take our sample? I guess we rather take a small hand throttle valve and let the flow go back directly to the reservoir for 10 mins before sampling to avoid contamination due to the valve.

The pump we are using to take our sample isn't the pump we are testing. It is in fact an oil filter pump but I can't find what type it is. I've been told that cavitation is not as a big issue as in other types of pumps. What we need to take as oil supply is the oil that is going through the pump, so the only way I can get away with this problem is to take the sample at the suction line.

RVZ717 proposed using a Bernouilli type device which I'm not familiar with. I shall make some research about that. Some pertinent links would be appreciated.

The question of my thread is now clarified to: How do we sample out of a vacuum line? other than using gravity once pump has stopped...

Something simple, like a syringe porteded into the suction line would do. You just need to apply more suction to the line than the pump is producing to draw your sample.

Good luck.

I would wager your pump is a gear pump, the most forgiving of the hydraulic pumps, they die a slow painful death as opposed to a catastrophic failure, as you would see in a piston pump. Therefore you can have the gear pump survive even with some cavitation... but it should still be avoided at all cost.

As for the Bernoulli device, Volvo used to make one a long time ago, but I'm not sure if you'll be able to find anyone still making the device, as no one really does that anymore for one reason or another.

Do NOT take a sample of your oil if you are not running your system, the data WILL be meaningless, you need working oil!

Get yourself a good needle valve to crack the test point line open, let this circulate back to tank for a while as you already know. After a period of time that you feel comfortable with (don't worry if you let it run for hours, no harm done) take your sample, using a sterile container supplied by your lab.

You may see if the lab running your sample has any recommendations for the sampling procedure.

I work in the Industry, and actually just finished designing a test bench system which will test nearly every pump on the market, with graphical analysis of every aspect of the pump in test. We also have a dedicated circuit to simply filter and cool the oil, this is called a kidney loop, and it can be a reasonable place to test for overall oil quality (before filters).

In the end, you should keep in mind that you need to look for working oil, not stagnant oil, the faster it's moving through your system the better for contamination sampling, your system should be in full operation when taking your sample.

It could probably be done with a catchpot connected to a venturi type pump. The problem is that a liter has to be taken before the sample to take away valve contaminants. A catchpot design with a moveable graduated cylinder inside it to take the sample could be interesting. Anywayz, I'll keep thinking about this. Thanks again!

Hello,

The system needs to be running and the oil warm.

Sampling from the return line is good place except with tank mounted filters as this means that you are sampling before it does anything so you have no idea how effective it is.

To establish this you will need to do either a sample from the tank itself or the pressure line.

Either are acceptable but need different equipment for each type of sample. I do not know your location but either in the UK or USA you should be able to obtain suitable sampling kit.

The tank is easier as it probably won't be under any pressure so a suction sampling device into a clean Mason Jar( A jar where you vacuum the jar and the fluid enters without entering the vacuum pump).hand vacuum pump is fine.

In the pressure line you need a test point and sampling hose. Hydrotechnik make excellent units in Europe. Again bleed the system into a clean sample jar.

Does the filter have any form of clogging indicator? Is it showing any signs of blocking. (Must be viewed when the unit is running and the system delivering fluid as most indicators are not of the memory type. Alas many filters do not have indicators and rely on changes being based on time, which never happens!

You say it is a test rig so knowing what you are testing should give you some indication as to what type and level of debris you might expect to find.

Your next problem is to judge the results! nearly all the major manufacturers, Rexroth, Eaton, Parker etc give ISO or Nas figures for acceptable levels for each type of equipment. i suggest you look at the pump as this is often the most critical item.

Good Luck

Oliver Dunthorne

Thanks all!

I have a question about your reply M. RVZ717.

"Get yourself a good needle valve to crack the test point line open, let this circulate back to tank for a while as you already know. After a period of time that you feel comfortable with (don't worry if you let it run for hours, no harm done) take your sample, using a sterile container supplied by your lab."

Don't forget that the test point line is at the suction, so the oil won't circulate in any bypass that would be installed there back to the tank. As I said first, there might be something I am missing hehe!

About the fact that the sample mustn't be taken with stagnant oil, would it be a good compromise to run the oil for a while, and then stop the pump, open a valve and let gravity do its work to take oil directly from the tank outlet (in which oil would have been mixed for a while just before taking the sample)??

"Don't forget that the test point line is at the suction"

What?? Why on earth would you do this? no one ever does this for many good reasons. Take your sample on the pressure side of the pump, with the pump RUNNING. You want somewhere between 10-1000 PSI.

No, that would not be a good compromise if you want an accurate sample. You should always take a sample out of the working fluid stream, the more turbulent the better.

You should completely forget about ever thinking about pulling an oil sample from the "suction" (inlet) side of the pump.

Is there not a sump or a storage point were you can extract oil from,the oil must rest somewere,this is not my field but you may try that good luck.

"Resting" oil is no good for an oil sample. A proper sample is taken with the fluid flowing through the lines.