advanced engine instrumentation ?

[RFguy]

Is anyone doing advanced engine instrumentation ?

I am inexperienced (and new) with aviation engine mechanics. In that I don't know details about pistons designs, crankshaft loads on direct drive props etc

 

But from my signal processing and instrumentation life, there is alot of scope for real time and off line diagnostics of aviation engines to improve reliability and safety. It's something I thought I can bring to the RA community. Of course it might be useless, but I have always found in my industrial work that instrumentation that detects problems before they really are is oh so worthwhile.

 

And it need not cost alot.

 

Aviation engines in GA seems to be back in the 19th century compared to some other industries that have devices with high stress loads.

 

I'm in the process of getting into RA, will buy a old Jab and get new Gen4 engine in it, and instrumentate the hell out of it. Time line is 6-12 months. If anyone is interested in discussing this here, I'll be happy to write down a few of my ideas, hoping that seasoned engine (re)builders and designers can put the 5$ opinion in on what is worth while, what is not, what would be useful etc.

 

and I am not talking about a tacho. I mean real time all engine vibration/mode analysis, (so you might see a change in the structure of the metal in the cylinder jacket before it is a problem) , real time gas & oil & exhaust spectral analysis, real time thermal imaging analysis of cylinder heads, so comparing the thermal imagery to a template , all by machine learning which puts the workshop expert into the chip . etc etc etc

 

glen

[marshallarts]

I like your line of thinking. I owned an aircraft (built in 1967) back in the 1990s and was quite taken aback by the instrumentation - it looked more like plumbing! If it ever happened that I owned an aircraft again (extremely unlikely), I'd want electronic instrumentation in it for sure. For a start I assume it is lighter than all those big mechanical things, but I'm happy to be corrected if that's a wrong perception. And I'd rather have the engine concentrating on driving the propeller, not a vacuum pump for some of the instruments - one more spinning/wearing thing to go wrong. As far as aviation engines themselves are concerned, it seems to me that they have stayed as they are mainly because they WORK, and perform the required task quite well. But adding a layer of instrumentation to monitor an engine in all sorts of ways doesn't change that (or shouldn't), so could be quite interesting. I will watch progress with interest.

[Bruce Tuncks]

Good luck to you RF guy. I have an Arduino-based cht and egt on all 4 cylinders and it cost very little and works great. My programming expert mate says I should add a recorder so I can follow temperature trends better. And I like the idea of measuring vibration so you can be sure if rough running is real or not.

From your posting, it seems that you are planning a more sophisticated system than I have by far. Please keep us posted!

[Blueadventures]

Is anyone doing advanced engine instrumentation ?

I am inexperienced (and new) with aviation engine mechanics. In that I don't know details about pistons designs, crankshaft loads on direct drive props etc

 

But from my signal processing and instrumentation life, there is alot of scope for real time and off line diagnostics of aviation engines to improve reliability and safety. It's something I thought I can bring to the RA community. Of course it might be useless, but I have always found in my industrial work that instrumentation that detects problems before they really are is oh so worthwhile.

 

And it need not cost alot.

 

Aviation engines in GA seems to be back in the 19th century compared to some other industries that have devices with high stress loads.

 

I'm in the process of getting into RA, will buy a old Jab and get new Gen4 engine in it, and instrumentate the hell out of it. Time line is 6-12 months. If anyone is interested in discussing this here, I'll be happy to write down a few of my ideas, hoping that seasoned engine (re)builders and designers can put the 5$ opinion in on what is worth while, what is not, what would be useful etc.

 

and I am not talking about a tacho. I mean real time all engine vibration/mode analysis, (so you might see a change in the structure of the metal in the cylinder jacket before it is a problem) , real time gas & oil & exhaust spectral analysis, real time thermal imaging analysis of cylinder heads, so comparing the thermal imagery to a template , all by machine learning which puts the workshop expert into the chip . etc etc etc

 

glen

Mark Kyle on this forum does a lot in the electronics and monitoring area with others. Mainly in the Rotax area though. Perhaps you might like a hybrid rotax powered Jab airframe and fit nice electronic monitoring. Cheers and look forward to your progress. Also must remover any changes to most factory aircraft require costly certification therefore a financial hurdle to development improvements.

[facthunter]

Look after your engine by all means but flying the PLANE and keeping a good watch OUTSIDE is paramount in U/L flying. Your own senses can be attuned to the engine, it's rumbles and noises. . Nev

[RFguy]

Hi Nev and others, thanks for the comments.

 

Yes agreed. Regular pilots of their own aircraft become well tuned to any changes. and yes there are lots of things that force airplanes down, and I see the engine as being the least developed item.

 

The idea is to collect lots of data to track changes. and post analysis using machine learning. When something is up, there is often speculation and guesswork on the ground.

 

There are many low hanging fruit to pick. The idea with the Jab is that I am confident that instrumentation can be such that I can see a problem coming.

 

As for modifications. I'll have to interpret the regs. Adding instrumentation to the engine I would have thought would be no different to adding an extra cabin thermometer. IE the installation of the instrumentation does not in any way affect the system being observed. I guess its all arguable. And if there is a demonstrable improvement to safety......

There would be things that would certainly interfere, if they were done , with the certification. Like real time chem analysis of the oil, and fuel, exhaust etc. requires instrumentation to be inserted into the line. Well there are ways but lets try and keep it simple on the sensors.

 

There is quite a bit in the vibration arena. Not only are there the fundamental , dominant notes, but there is much information in the echos (reflections) from discontinuities (crack/fatigue, change in crystal structure ?) in the medium (IE a block of Ally) . Sort of like sub miniature acoustic geo strata surveys. The speed of sound in hard metals is fast so the sensors have to be pretty wide band.

 

I mean you can really go to town on things you can monitor. acoustic and thermal sensors aside- that's the low hanging fruit. attaching high bandwidth high temperature vibration sensors to the engine components might be a no no, but that's also possible by looking at the light reflected off the component, it has all the same information as a attached sensor, maybe with exception of being able to differentiate between various types of waves in the medium. I am sure a geologist geophone person could add loads here.

 

and there are other techniques used in industrial robots like continuous strain and torque sensors on actuating rods. there are plenty of those in many craft... (non engine components I am referring to)

 

anyway, plenty of ideas. My journey in aviation will be about personally developing myself in my own technical field, as much as the actual flying and all the techniques, experience and judgement that goes with it.

 

Sounds like I should have a sandwich with Mark (Kyle) . Oh and if anyone feels I am really f***ing mad, please let me know, I take on all opinions. .

[old man emu]

Oh and if anyone feels I am really f***ing mad, please let me know, I take on all opinions.

 

Not mad, but over-zealous. At the moment you are going at the whole game like a bull at a gate.

 

You obviously have either experience or interest in structural matters and material sciences, as well as your keenness to get into the air in your own aircraft. However there is a big difference between the highly technical aspects of structural analysis and going to buy a $100 hamburger, which is a main aim of recreational flying.

 

As Facthunter, an very experienced member of our community has said, " flying the PLANE and keeping a good watch OUTSIDE is paramount in U/L flying " . In the next few months you are going to be pounding the forum with questions related to getting your certificate, and there will be a crowd of experienced pilots and maintenance people to put you right. How about concentrating on getting your certificate first before your start diving into the minutiae of aerodynamics and aircraft systems design?

[RFguy]

indeed ! and all true.

yes, I will be pounding the forum with questions relating to getting my certificate.

 

cheers

[turboplanner]

First job, find a good instructor; he is the one who is constantly assessing you and teaching you to be a good pilot in real time.

After maybe 3 or 4 hours you'll be in a good position to judge your priorities, and I'll bet by then you'll be totally focused on how you are going to control the aircraft to achieve landings, not get knocked around by crosswinds etc.

[RFguy]

The best advice I got I think was that instructors should be matched to the student. A couple I spoke to said they changed instructors twice... I'll move back into that forum topic for the rest of that topic.

[Kyle Communications]

Glen I like your enthusiasm and obvious knowledge and experience in the commercial end of embedded and a sensor areas. The biggest thing about aircraft is to KISS. There is a very wide line on costs as it is with any sort of aircraft ownership and operation. Keeping your eyes out of the cockpit is the main thing to worry about. Yes you can log all sorts of everything now with the array of sensors available and depending on how good you are with software you can have some really funky stuff. Then of course there is all the time AFTER the flying to be pouring through logs and graphs to be looking for something that may have a effect in the long term. But in the bigger scheme of things the cost to implement this stuff is far out of reach of most with sensors your talking about and the cost of the hardware design and implementation. Most private aircraft rarely get to 2000 hrs over its life so this is why you dont see all the gear your talking about.

 

Your better off trying to improve what you have than design what you want....unless you have lots of money and time to burn. The stuff I play with for moding my aircraft really is to improve what I consider inferior to make for more reliable and safe operation. I am not saying dont do what you want but you really have to look at it in a business like environment and its the bang for the buck that is most important. Owning and flying your own aircraft has a large cost attached to it and for most the less you have to spend the better. You only spend what you think with be a big benefit to either performance or usability of the aircraft.

[RFguy]

agreed and taken on board..

we should have a sandwich sometime., when cross border sandwiches are permitted.

cheers

[Kyle Communications]

Yes certainly... I just came back today from up at Murgon here in Qld. We had a Savannah muster at Angelfield. There were going to be quite a few coming from over the border but of course 1 am on saturday morning killed that. We still had about 20 Savannahs show up and another 10 aircraft types come in the Murgon breakfast was a great turn out and a couple of great days. Even got to see "The Girlfriend", she was flown in from Chinchilla where she now resides. If you read enough of this forum you will see she is the first aircraft I built about 8 years ago and I sold her off some 7 weeks ago as I have another savannah wreck I am not to far from finishing rebuilding and I have another kit here a Rans S-21 that I will start here at after the rebuild moves out to the hangar for assembly and fitout.

[Yenn]

I think you need to decide what you need to have info on. Then work out how to obtain that info. You don't want too much info when flying, so have it recorded for examination later.

Definitely make a list of what you want first and also why you want it. most of your needs will be fairly easily met, except for such as vibration recording.

[facthunter]

They removed over 1/2 of the warnings on the B 747 that affect take off because they were a danger and they have 3 people to monitor that and it's in controlled airspace. There ARE times when a warning may be an expensive distraction and some of them are inevitably FALSE.. You are trying to MANAGE a FLIGHT primarily and as a priority. Engines and even the structure is expendable sometimes in order to keep the lives of people more safe.. I look after engines probably more than most always during my flying career (start with little throttle observe warm up shut down from slow rpm, climb at higher airspeed if temps going up, etc). Your engines no good if you crash the plane it's in and you are taught how to handle engine failures wherever they happen and NO engine is guaranteed not to fail or run out of fuel because you are lost, or didn't notice some one pinched some of the fuel last night. Nev

[Yenn]

My thoughts on instrumentation are that extra to the usual tacho and oil pressure and temp., it is nice to have CHT and EGT because these not only tell you what is happening now, but they are extremely good diagnostic tools. manifold pressure or inlet vacuum are necessary with constant speed props, but with fixed pitch, they can tell you a lot. The other tool is a fuel flow gauge. You have to know how to interpret what you see from these gauges.

If you have anything more than magneto ignition or fuel injection some of them become essential.

[Nobags]

Very nice to have CHT & EGT for all cylinders, only if you’re flying behind a machine that has manual mixture control. IMHO

[Bruce Tuncks]

Good point Nobags. I reckon the EGT was a once-only benefit to the Jabiru. And the reason is that the carby on the Jabiru is so automatic that there is nothing you can do short of changing things like the feed-ins to each cylinder in the plenum chamber. Well the EGT on all cylinders sure helped with this, but these days it is useless as Nobags suggests.

[RFguy]

anyone using the Rotec TBI fuel system on their Jab ?

[facthunter]

I've seen it used on a VW conversion. Good way to start a grass fire. May be some way to tame it. I'm becoming more of the opinion that a mechanical drip injection system is the best with fine tuning of each injector flow rate. Like GAMI injectors. Less chance of damage from the effects of a backfire which has more chance of severely damaging an engine than most realise.. Nev

[RFguy]

OK just looked at the GAMI website. interesting. My Datsun injection system (1988 - age 17) I used a single injector in a body just down stream of the carburettor. The Carburettor ONLY functioned as a convenient butterfly and intake tube. I used a mass flow sensor on the inlet. O2 sensor on the other, very rudimentary but it did work. But not as well as an SU for reasons I couldn't figure out without much instrumentation at that age.

 

Backfires on the Datsun could occasionally jam the butterfly shut. Complete stop ! That was the extent I knew of problems at that point in time

[Jim188]

RFGuy,

 

I hear what you are saying RFGuy and see what you are thinking and I am enthusiastic to hear you are wanting and willing to do bring some of your expertise from another field.

 

I would be happy to have a yarn with you, I have an experimental Jab 430 with a Gen 4 engine with 2 full EFIS's and a spare input that can be programmed and configured. I don't feel the need to have vibration monitoring on my aircraft, but I think it would be good to confirm that you are flying your aircraft in the RPM sweet spot for your air frame and engine. I also think vibration monitoring would be just like having your oil temperature monitored, the vibration may never change at your cruise RPM, but if it does it's not just you thinking it has changed.

 

A few years a go, I was flying a Jab 230 and was heading off for a bit of a local Nav. I had flown this Jab 230 a number of times, so after taking off, something did not quite feel quite right, nothing huge, but the engine seemed to be running ok when cycling the throttle, but it had what I thought was a very small hint of laboring, with a very small hint of vibration which seemed nothing to do with the engine rpm, in all nothing was jumping out at me saying we better land in a hurry. But, I did land and lifted the top cowl to have a look, at first nothing jumped out at me, until I saw the air intake scat hose was neatly tucked down and in towards the carby. The aircraft had no recent maintenance and had been flown a few hours by a number of different pilots before me. The scat hose was clamped back onto the bottom cowl intake as it should have been and all felt nice again.

 

Another example, I had some work done to my plane and on the way home I felt the plane had a little more vibration in it than on the flight up, I talked to the maintenance guy about the work that had been done and he said pull the cowls to check nothing was touching cowls. Sure enough the positioning of the oil blow bye hose had changed slightly, so could have been touching the top cowling and adding this vibration to the air frame which I felt in my feet or when I touch the top of the instrument panel. I just re-positioned the hose and put the cowling back on and that small vibration was gone.

 

I think these types of situation may have been picked up by RFguy type vibration monitoring. As a pilot flying my own aircraft, we all hope we will feel such things, but for those flying hired aircraft, those pilots may not be able to feel those vibration changes as they happen, but RFguy vibration monitoring could be a good way to warn a maintenance crew to have a look for something.

 

 

 

Jim 188

[Kyle Communications]

You could just fit a accelerometer and have a display either LCD with a value of a row of LEDS... once calibrated you would have a baseline and could see in realtime if anything was changing or changed since the last flight. You could mount the sensor in the engine bay or on the engine mount maybe. Fairly simple to make now days and relatively cheap. Just need a low value range one 2G or under. There are lots of these devices now

 

https://core-electronics.com.au/triple-axis-accelerometer-bma220-tiny.html?utm_source=google_shopping&gclid=Cj0KCQjwsuP5BRCoARIsAPtX_wFUO6x2bmqEuS-Xz9INjyA951Z_fIOwiTRSCwDEvwdtIWGzAAyHn0caAuVnEALw_wcB

 

Another one

[spacesailor]

My grandson,s pannel is pretty good, even if it does,nt fly.

spacesailor

[RFguy]

Yeah the accelerometers Mark is pointing at are very useful, if anything as a peak G recorder just to remind you that you pulled it out of a deep dive last week.

 

I was thinking about how to monitor torsional mode crankshaft vibrations. Probably use a laser looking at the flywheel outer rim in the radial plane, and a very fast photodetector for processing the observed vibrations (kHz) . It would depend on how much got transmitted into the flywheel through the bolts, plate, and perhaps some flywheel flex might attenuation it a bit. Nev probably knows something here.

 

There is lots of vibration analysis, and presumably using machine learning for FAULT CLASSIFICATION going into machines these days. It's a hot field of engineering.

 

Looking at the time domain analysis of the noises in the cylinder head, I would expect, based on what other industries do, that it's quite easy to see micro cracks in the casing.

 

A pixel image bolometer (a thermal camera, not a pin spot reader) can see leaking gas a mile away.

 

Also, very high frequency ultrasonic noises get made by high pressure gas escaping, when using ultrasonic leak detectors on pressured feedline on radio masts, we can hear it a mile away.

 

IE hearing escaping gases where they shouldnt be is easy enough, if it is useful.

 

The firing waveform of the spark plugs can be template compared, and can tell you a little on what is going on at ignition.

 

exhaust gases can have real-time mass spectroscopy.

 

exhaust pulse noises have signatures, bad valves can be heard I am sure.

 

carburettor icing can be detected in advanced pretty easily.

 

The thing is, the electronics of all this monitoring need to be made robust, the engine bay is a very harsh environment !