I've sent the following question to Opieoils (Oilman - member of this forum I think) about relying on dielectric constant for oil changes:

quote:

Do you happen to know the dielectric constant of Castrol RS 10W-60? I have an instrument that’s giving a value of 6, but I think that this is a little high for an oil.

I’m planning to do oil changes to my car based in both km and dielectric constant measurement. The idea is to do the oil change between 8K to 12K km provided that the dielectric constant doesn’t increase in a certain percentage. I have some questions that I hope you might help me with:

- What percentage should I consider for the dielectric constant change to conclude that something is wrong?

- What anomalies can I detect by measuring the dielectric constant? What makes it change? (i.e. oxidation, additives depletion, water, wear debris?)

- Is it a good (safe) idea to do this?...better than consider just km.?

- What if I’m getting no changes in the dielectric constant after more than 20K km.? Can I safely extend oil change interval, relying only in dielectric constant?

He passed the question to a Silkolene expert. This was his answer:

quote:

Over the years there have been various patent oil testers or oil condition monitors based on conductivity or dielectric constant; all very ingenious, all a complete waste of time. With a very pure, clean mineral oil such as transformer oil, oxidation leads to the generation of long-chain carboxylic acids which, being polar, cause the conductivity and dielectric constant to increase.

The trouble with modern engine oils (ie post -1955!) is that they contain chemical compounds which inhibit acid generation, but some of these are polar themselves, so even new oils will give a high reading. (As with the Castrol RS 10W/60) Also, they allow very high levels of solid particles to be held in suspension. So a used oil can have high conductivity even when in good condition. Needless to say, a trace of water will also make a big difference.

So, unless you are testing transformer oils, the best place for dielectric testers is the skip. (I’m perfectly serious; I’ve actually junked a couple of these useless devices myself.)

Regular oil changes are good for an engine, so running oil until it’s ‘on its last legs’ is false economy. If there is any concern about oil condition the best advice is to submit a sample to an oil monitoring service.

What do you think about this answer?

I've read that some car's diagnostic devices measures the dielectric constant to tell when to change the oil. Considering the answer given by the Silkolene expert, are this devices totally useless

...any recommendation?

Thanks
Tomás

mmm...anybody?

thanks...

Tomás

We supplied a sensor suite for some seeded fault diesel engine tests a few years back. The oil samples obtained during the engine tests were analyzed and compared to Army condemning limits to determine when each test should have been ended for soot, oxidation, coolant and fuel. The test were continued well past the Army condemnation levels to evaluate the responses of all the sensors. The dielectric sensors (two different manufacturers)increased at a gradual,constant rate for the soot accumulation and accelerated oxidation tests. The coolant contamination and fuel dilution tests caused only small changes in the dielectric sensors outputs. Fact that sensors were in a flowing, 160F oil stream didn't give coolant/water chance to settle out as in table top sensor. On-line soot sensor was erratic - problem never identified. However table top version of soot sensor worked very well at monitoring soot accumulation -used by Army as condemning test. Viscosity measurements made off-line were too late - oil badly degraded by soot or oxidation - well past other condemning levels, before viscosity condemning level reached. Worked well for fuel dilution only. Only oil sensors that worked on-line were two low frequency sensors we designed (of course) which showed a distinct breakpoint when the polar additives were depleted by soot or oxidation - sensor readings decrease then become constant with soot accumulation and decrease, flaten, then increase with accelerated oxidation. Sensors increase rapidly with fuel dilution and diverge in presence of coolant - one sensor increase, other decreases. A hand-held instrument the Navigator - no longer manufactured by Predict due to some cleaning issues with screen - showed similar trends as low frequency on-line sensors
Bottom-line: for an on-line sensor to be useful it has to show a distinctive breakpoint that correlates well with accepted criteria- starting point will vary with manufacturer, new specs (increase soot carrying, oxidation life, lower phopshorus), etc. so going off of absolute value impractical. Second the oil condition being monitored should be important to equipment health - would be more important to detect corrosive coolant leak immediately than moment oil starts to oxidize at an acclerated rate. Finally for aftermarket sensor shouldn't require engine modification for use - work on/replace dipstick, oil plug