Odd Sensor Design

the

detection.

of

You say this is turbocharged with an intercooler, so you must be running some serious boost. So my question is does the turbo still work when it starts leaking oil? If so, you could install a small tube that rests on the bottom of manifold or aftercooler, and have the tube drain at the top in a tall, skinny container with a float switch. As long as the tube is small enough, or long enough, very little of the boost pressures would escape. I would also be looking for some way to make it drain, even if it requires modifying the intake/intercooler (remember under boost it can be forced to go in any direction). Also, what about measuring the oil level in the sump? If it drops significantly in a short time, something is wrong. You could also supply oil to the turbocharger from it's own source, with a electric pump. The oil supply could have limited amounts of *pumpable* oil, say 250 ml, or an accurate level gauge. This gives the advantage of running special oils in the turbo bearings separately from the engine, and being able to keep the oil for the turbo relatively pure and easy to change. And, being able to keep oiling to the turbo while it's still spinning, long after the engine is shut down. A pressure switch can be incorporated to prevent turbo seizure if the oil stops flowing by shutting the engine down.

Lots of great feedback and I appreciate each and every one of them.

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----

---= East/West-Coast Server Farms - Total Privacy via Encryption =---

I like this idea. But maybe you can help clarify. Since the engine manifold temperature can vary so much, wouldn't that also screw up our known resistance or cause a large variance as well? Or is the self heating temp higher than what we would expect to see in the manifold? Manifold would vary from say 90F to 180F.

If so, then would that require two sensors. Measure the one in raws air/gas, versus the one that may be covered in oil?

Then last, if the oil level may only get 1/8" to 1/4" deep, do some of these sensor measure from the Tip only, and would that be accurate enough?

Again, thanks for the ideas. Richard

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----

---= East/West-Coast Server Farms - Total Privacy via Encryption =---

Automatic oil level checking in aircraft gearboxes, using a screw-in thermistor assembly, is quite common. In this case though the relatively thin film of oil might require the sensor to be a film or mat, exposed to the oil but insulated from the thermal inertia of the manifold.

I supose you've considered using opto electronics?

Rocky

the

actually

it

Maybe talk to an aircraft maintainence engineer about how they do the lockout on a turbine engine after a seal failure. Pat

the

"when it fails a turbo charger" - what does this mean - does the turbo break?

If it does break, you could detect that, couldn't you? Put some kind of interlock so if that happens the thing won't start.

They do! a PT6 wont start without someone reseting a mechanical lock if the turbin play is out of spec!

Remember startup != takeoff

fuselage

You are correct the engine wont shutdown if flight but it will not start after certain turbine faults. Pat

One of my old cars (can't remember which) had an oil level sensor built into the dipstick.

I searched around the web/usenet looking for comments about oil level warning lights and dipstick and found almost nothing - so I guess they are either pretty rare or pretty reliable :).

I don't know how it worked but suspect the only realistic options are hot wire and optical.

I have seen the odd reference to optical oil level sensors using a prism. I suspect when you immerse a prism in a liquid with about the same refractive index it isn't a prism any more and you can detect the lack of refraction or something like that.

Either way I suspect you will have to mount the sensor in a bit of a well to collect oil and give you more than 1/4" maximum depth.

No, just the one resistive temperature sensor.

Measure the resistance with a low current stimulus to start with. Then increase the current through it for a few seconds (to self-heat) and measure the resistance again.

If the sensor is in air then the temperature rise will be quite high, but if it is immersed in a liquid which has thermal mass then the temperature rise will be much lower.

Because the start and end resistances are both measured then the ambient temperature doesn't matter.

Yes, that's a problem. A thin film of oil needs a sensor with a very low thermal mass and less than 1/8" in height at the tip. Could be done though.

BTW: This is probably a bespoke microcontroller job.

I think he should run a thin aluminum tube to wherever he was thinking of putting the sensor and have a pump pump out any oil that has collected. Have it run for X amount of time as part of the startup sequence.

No, you measure the differential temp diference which compensates for the ambient.

The 1 I made would easily measure 0.5 mm.

Your Nichrome wires are doing the same job as thermistors, but with much lower sensitivity to the temeprature differences you are trying to detect. Thermocouples don't normally dissipate any power at all, so they would be perfectly useless.

You don't need two sensors - if you measure the the self-heating of a single sensor at two different current levels in quick succession (the time constant of a bare thermistor is about a second) you've got all the information you need.

---------------- Bill Sloman, Nijmegen

Hmmmm -

My long-sold Renault Espace 2.1 TDi had a separate oil pump for the Turbo to keep the oil flowing for minutes after the engine switched off to cool those Turbo bearings. Cars usually epitomize "Cheap".

They Dont!!!

They measure Oil Flow and Bearing Temperature e.t.c. - there is *no* shutdown. The Pilot decides. On a takeoff it might be better to blow the engine and get enough speed to turn back and land, than not getting the speed and run off the end of the runway. Especially in Frankfurt (Large appartment blocks at the end!!)

Aircraft use very regular and ridgid schedules for maintenance and inspection - supposedly - Scandinaivian Airlines had a near-fatality a few years ago when a bearing seized, causing the drive shaft to break and the end-stage of the turbine took off in little pieces, penetrating the fuselage almost cutting the APU off where all the 4-way redundant hudralics meet.

This was a known risk with that version of engine + seals, However management did not care to do the upgrade recommended by Pratt & Whitney - maybe they still do not care (SAS was in financial trouble then as it is now).

I did some rough calculations and verified it with a quick bench test using a light oil.

With only about 2.5mm immersion I can get plenty of mV of signal for an instrumentation amplifier and a bit of other circuitry/programming to deal with it. I did a similar sort of thing (quite a bit fussier) for an anemometer probe. The probe would have to be mounted a bit off of the bottom, perhaps spaced with an insulator.

Best regards, Spehro Pefhany

Time constant in air of a thermistor can be pretty damn slow.

Best regards, Spehro Pefhany

"Pat Ford" skrev i en meddelelse news:cslqtv$sm5$ snipped-for-privacy@nrc-news.nrc.ca...

Thanks for the correction - I am not an aircraft engineer ;-)

temps

routine

other

and

I remember seeing some liquid sensors in the past that consisted of a LED/Photodiode pair on the end of a piece of plastic rod with a "wedge" at the other end arranged so the light from the led went down the rod and was reflected by the two sides of the wedge back to the photodiode. When the end of the probe was immersed in a liquid with a similar index of refraction, the light quit reflecting back and could be detected as "fluid present". Perhaps something along that line ??

mikey