Oil condition monitoring? Just not needed by you?

[M61A1]

M61A1 those gas turbines that you work on, are they Siemens gas turbines? Are they near Dalby?

Vibration analysis is simple. Get a set of vibration traces, use a program to convert the vibration to frequency and amplitude (fast Fourier transform maths) using an ap freely available. Then look for things that happen at the largest frequency, how fast the machine is going, meshing speed of gears (all in hz). Finding a match in frequency will usually lead you to the problem.

They are sort of near Dalby but they are MTR. The vibration analysis I’m interested in is to do with tracking and balancing helicopter rotor blades. It’s one thing to static balance them, another to dynamically balance them and that all changes once you get forward flight at different speeds.

some of the balancing kits have a settingto do a “spectrum” analysis, where it gives you a graph showing all the different frequencies and their amplitudes through the airframe. Once you know that you can match it to a rotating component because you know what rpm that component spins at.

[Bruce Tuncks]

That was interesting Allan. I would have guessed that most of the ferrous wear particles I catch on the magnets come from the rings and cylinder wall.

So this is wrong... can you tell us some more please?

[Geoff_H]

Valve guides, timing gears (could be fibrous, or belt though)

[pluessy]

I have been taking oil samples on my Rotax 912 since I got it at 180h (engine). The initial results showed 2 problems:

1. avgas is loading up the oil with lead, big time
   
2. the oil temps were too low (lower end of the green range), visible from fuel dilution and some water
   

I'm cutting the oil filter and keeping a section for my records, and take a picture of the mag plug at the gearbox before cleaning it.

 

I'm trying to avoid avgas like the pest now and I have added a barrier to the oil cooler to increase the oil temp (size varies with season). Rotax recommends to run the oil temp to the top of the green at least once per flight, to evaporate water and fuel. The Bings are running rich and fuel gets into the oil.

 

Professionally, I deal with oil samples every day, interpreting the results from the lab and preparing work instructions if actions are required. At the same time, I have removed a number of oil sample points on mining gear and replaced with a photographic record of the mag plus. Much more telling and immediate (3 compartments with oil mixing).

 

Oil samples need to be taken at regular intervals, one-off samples have little value. It doesn't have to be every oil change, it can be every 2nd/4th or what ever is suitable. Just make it consistent and provide the lab with the correct information (critical is oil hours and oil make/type/grade). The only problem here in Aus is that no lab has the personnel familiar with aircraft engines, hence the "we can't comment".

[Markdun]

Here is an oil analysis on my Jab 2200 by ALS in Australia. I do it every 50 hours. Around $35. I agree you need a baseline. Seems easy enough and not that expensive. I've found filter chops less satisfying; only ever saw a few aluminium millings in the break-in period. A surprising result is that even after hours of running with oil temps less than 70C, still no water in the oil!

1592132544241_Mark Dunst_Unit193393_Petrol Eng_Severe_34971739.pdf

[Yenn]

Checking the filter contents may seem like a waste of time, but you will most likely see a sudden increase in metal particles, which is what you are really looking for. Just because you don't see particles, only tells you that so far the wear is not really bad, As soon as you see them you know you have a problem.

Relying on visually checking a magnetic drain plug is only the same as opening a filter, except you will not see non ferrous particles.

[pluessy]

You really need all 3, filter, magplug (if fitted) and oil sample to get the full picture:

• the filter and mag plug show the accumulation of the wear over the time frame since the last change/clean
  
• oil samples are a snapshot at the time of taking the sample, they are subject to variations due to incorrect sample taking processes
  

The oil sample results are also limited in the size of the particles that are detected. All elemental analysis (Fe, Cr, Al, Si etc) is limited to about 8 micron in size. The PQ is ferrous material in the 10-200micron range. All the bigger stuff is not reported unless it is visible in the sample. The filter catches everything bigger than about 10-20micron (chewed-up O-rings, piston pin plugs, brass thrust washers etc) and the mag plug catches the magnetic particles that are floating past (more important in engines with gearboxes).

 

They are like an insurance, you hope that you never see anything but you will be glad you checked when you do see stuff.

 

Oil analysis, filter inspection and mag plugs can only detect failures and abnormal conditions if they happen over time. A rod bolt failure, dropped valve, conrod/bearing failure etc cannot be predicted with this type of condition monitoring, unless it is preceeded by abnormal running conditions that do show up in oil sample results (excessive oil temps, fuel dilution).

 

The one value of the oil sample is the oil condition indicators. The ALS report is pretty poor in that respect, there are very few oil condition indicators (only fuel and viscosity). I like to see the viscosity at 40 and 100deg C (standard for engine oils is 100, not 40), the oxidation, nitration and sulphation. The ALS report also doesn't contain any of the additive package elements (Ca, K, Mo, Mg, Zn, B). Additives can change from batch to batch or over time, potentially causing unintended results in some engines (leaching of Cu/Al from oil coolers is a common problem). Or show an incorrect oil used.

I'm using KOWA, same price, twice the results:

Query.pdf

[M61A1]

Our work oil samples go to Intertek, I don't know how much they cost, but the analysis goes way further. They can tell us what type of ferrous material it is to determine whether its a gear, bearing or seal. We have a table with all the different compositions of metal in the subject component. Particles get sent away for analysis as well.

[Markdun]

You really need all 3, filter, magplug (if fitted) and oil sample to get the full picture:

• the filter and mag plug show the accumulation of the wear over the time frame since the last change/clean
  
• oil samples are a snapshot at the time of taking the sample, they are subject to variations due to incorrect sample taking processes
  

The oil sample results are also limited in the size of the particles that are detected. All elemental analysis (Fe, Cr, Al, Si etc) is limited to about 8 micron in size. The PQ is ferrous material in the 10-200micron range. All the bigger stuff is not reported unless it is visible in the sample. The filter catches everything bigger than about 10-20micron (chewed-up O-rings, piston pin plugs, brass thrust washers etc) and the mag plug catches the magnetic particles that are floating past (more important in engines with gearboxes).

 

They are like an insurance, you hope that you never see anything but you will be glad you checked when you do see stuff.

 

Oil analysis, filter inspection and mag plugs can only detect failures and abnormal conditions if they happen over time. A rod bolt failure, dropped valve, conrod/bearing failure etc cannot be predicted with this type of condition monitoring, unless it is preceeded by abnormal running conditions that do show up in oil sample results (excessive oil temps, fuel dilution).

 

The one value of the oil sample is the oil condition indicators. The ALS report is pretty poor in that respect, there are very few oil condition indicators (only fuel and viscosity). I like to see the viscosity at 40 and 100deg C (standard for engine oils is 100, not 40), the oxidation, nitration and sulphation. The ALS report also doesn't contain any of the additive package elements (Ca, K, Mo, Mg, Zn, B). Additives can change from batch to batch or over time, potentially causing unintended results in some engines (leaching of Cu/Al from oil coolers is a common problem). Or show an incorrect oil used.

I'm using KOWA, same price, twice the results:

Thanks for that, I'll check them out.

[Downunder]

There is also the "context" of oil sampling.

If a sample comes back with a problem, what are you going to do?

It is important to think about that before getting involved and paying for analysis.

Are you willing to put time, effort and money into investigation?