Graphs of my test results to be added - when I work out how to add Open Office graphs!
Engine Cooling Issues
This is a wide and complex area and sadly an area in which the Jabiru engine has some well known problems. Engine overheating can lead to some significant problems which can include sudden stoppage in flight so every effort has to be made to ensure your engine doesn't have any cooling issues.
I am not an engineer and I admit quite freely that prior to building my aircraft I didn't really have any idea about engines worked. Since finishing the build, though I have done a huge amount of research, talking to numerous people including the guys at the Jabiru engine factory and to Rod Stiff (the designer). I have joined several web groups and spent countless hours playing around with my own and other Jabiru engines. So I am happy to share what I have learned but advise that what I say is my understanding of the situation and could be wrong. It is given without warranty or responsibility.
Essentially there are two areas where cooling has to be considered:
2.Cylinder Head Cooling
I haven't really had too many problems with oil cooling. When I got my kit it came with the better aviation oil cooler and the better positioning of the cooler which forced more air through it. I never had any real problems. The previous cooler was too small and positioned so air could deviate around it so the Jabiru was famous for oil cooling issues. I think that has been largely fixed now ...but there are still some things to be aware of.
What does cause me oil cooling problems?
I have had a couple of times where overfilling has caused high temps. I just throttled back for an hour or so as it dumped the excess out the overflow line and gradually was able to apply more power once the level came down.
Jabirus are particularly sensitive to the oil level. Jabiru Co. now advises that you only fill to near the bottom of the hashed zone on the dip stick. I have been told there are two reasons why too much oil can cause it to overheat. I am not sure which is the real one so I outline both. You decide what you want to believe.
If you put too much oil in it can froth. Froth doesn't pump so you get failure of the oil to carry the heat away.
Limitation of capacity of cooling system:
The cooling system can only dissipate a fixed amount of heat. If you add too much oil it carries more heat than the system can dissipate and heat is returned to the engine. This heat adds to the heat being produced so you get an overall build up of heat.
Either way, take it as read that too much oil is a good way to overheat.
This is where your instruments tell you you are overheating but you aren't really. I have had this problem more than real overheating. It may be helpful to be aware of this possibility.
It is worthwhile being aware that the oil pressure and the oil temp senders in Jabirus are prone to failure. I am on my third oil pressure sensor because in the factory the sensor is attached to the engine (despite the sensor manufacturer advising NOT to install it on the engine itself) – the vibration causes eventual failure. If you notice your oil pressures suddenly going high while all else is normal – you may have a failed (or failing) pressure sensor. I found that both times mine failed it had high swings that would swing for some time and eventually go onto to fail at constant high.
I am on my second temperature sensor. Although they are more robust than the pressure sensor they still can fail. When they fail they usually fail into a high resistance which will read high on the gauge.
If you install a Dynon EFIS/EMS (Electronic Flight Instrument System/ Engine Management System) you can get all sorts of weird interactions where poor earth contacts of other sensors can give high temp warnings.
Dynon make multiple mentions that the oil temperature sensor system is very sensitive to poor earthing issues.
I have also found that intermittently I get sudden high temp warnings when I activate the autopilot. I have installed both a Dynon EMS and the old analogue gauge and if I get a sudden high temp on the Dynon I cross check it against the other.
Cylinder Head Cooling:
This is the area where Jabirus have had most problems. The fix seems to be highly variable and sometimes complex.
How Many Sensors?
Part of the cause is that the Jabiru is VERY prone to complete variability from one cylinder to the next and from one side of the engine to the other.
This means that you absolutely MUST have sensors to ALL cylinders. (Cylinder Heat Temps (CHTs) as a minimum but preferably Exhaust Gas Temps (EGTs) as well )
Jabiru sell the kit with just a single CHT sensor which they say to install on cylinder head 6, as this is usually the hottest. But in mine the hottest for a long time was cylinder 2. If I relied on the single sensor I would be measuring my second hottest cylinder.
So why do they supply just one? They are trying to keep costs down to make it affordable. I think that's fair enough but they should tell customers they are better off if they get the full sensor suite. To be fair to them, they do, to some degree. They will sell you a Dynon EFIS/EMS.
Do you need EGT?
You will see below that CHT is the final outcome of the balance between heat production and cylinder cooling. If you only know the CHT you are missing part of the information needed, so I think you need both.
Problems with Jabiru CHTs:
A word of warning – Both Dynon and Jabiru use a metal ring sensor that sits under the spark plug. Attached to the lug of the ring is a two wire sensor lead. These wires work by the principle of “Dissimilar Metals”. They are two different metals and are simply welded together at the tip. The weld line then forms a known resistance (because that's a characteristic of dissimilar metals when they are fused together along a weld line. This resistance however changes with temperature rise (usually increases). The temperature gauge or the EMS is calibrated to read the resistance and to convert the changes to temperature.
This leads to a couple of things you should know.
If a wire breaks (which they are prone to do) you should replace the whole sensor wire and ring. I have done some experiments with a broken one and just by crimping them together you “can” get an accurate reading for a while but you tend to get a drift in resistance due to oxidation of the wires and so will get a drift in reading. Eventually it will give totally wrong readings – it is not worth trying to save money.
Also both CHT and EGT leads are cut to a known length at the factory. The length also affects the resistance of the wire itself so you can't change the wire length either by shortening or by adding any more of the same type of high resistance wire – You CAN lengthen using normal wire, because it's of such low resistance it is effectively invisible to the gauge.
The heat has to travel up the ring lug and transfer to the wire. But there can be problems in installing the sensor which can lead to false temperature readings.
The Jabiru supplied ring is a 14 mm ring which is the wrong size. The instructions say to acquire a 12 mm ring and carefully lift the crimps and release the wire. Then replace with the 12 mm ring.
Dynon have Jabiru sized leads, if you ask for them.
Releasing the crimp:
This is easier said then done. I destroyed several trying to lift the crimps. In the end I found the safest was to secure the ring in a vice and use a dremel cutting tool to carefully grind off the sides of the crimp till it falls off easily. A hand file would do the same, probably even a bit safer!
The “12 mm” rings are still a little to big (in fact if you measure them they are really almost 13 mm despite the code saying they are 12 mm!) but if you are going to stay with ring sensors then you are stuck with them. There are a number of problems with rings under the spark plugs and some builders have removed them and installed screw on tabs in other places.
What problems are there?
1.Practical one of fitting the ring in the recess.
2.Problem of getting the ring to sit in a proper position
3.Problem of distorting & damaging the ring with later removal and replacement of the plug
4.Problem of inconvenience with replacement
The rings and wires take up space and can stop the spark plug socket spanner from fitting down in the recess. The ring can be difficult to get to sit flush. To get the best seal of the plug requires the plugs own crush washer being removed and then the ring installed and the crush washer replaced. This is not as easy as it sounds as the crush washer has inner 3 lugs which can make it very difficult to remove and replace. They are fiddly, worse if the plug is hot, they tend to jam in the threads and if you use a pair of pliers – one slip and the thread of the spark plug gets burred and ruined.
Getting the ring to sit in the right place is absolutely essential. If the ring is off centre or distorted then gas can escape out from under the ring and if it blows across to the wire tip will register a false high temperature. If you are getting a crazy higher temp on one cylinder – check for correct position of the ring. This especially true if you have rings that are too big.
The spark plugs need removal many times. Whether it is to replace them, check the colour of the residue on them or to conduct compression or Leak Down tests, you will need to remove them. When you do the final tightening of the plug it is almost standard that the ring will try to follow the direction of the turn and will distort the lug, distort the ring itself or worse break the wires. This is so much of a problem that many builders (myself included) resort to grinding out a groove and cutting off a small section of fin to allow the wire to fit into and snug up when tightening the plug. Make sure you tape over the plug hole beforehand so you don't get swarf down into the cylinder!
I find I still need to insert a small screw driver and jam the lug from turning as I put the last tightening on the plug. However I have still succeeded in breaking a number of sensor wires doing this.
The plugs are so cheap that I change them all over every 50 hours but the fact that the rings make it so difficult make it tempting to leave them on longer.
So what can you do about it?
Some builders have resorted to moving the sensor out from under the plugs. After all many automotive sensors are not under the plugs! This requires acquiring smaller rings, just big enough to take a screw. But the ring should be of a heavy copper type not a light electrical type as the electrical ones don't transmit heat as well.
Two places I have heard of are in the small hole in the ridge between the spark plugs and under one of the cylinder head bolts. The thing you have to be concerned about if you choose to shift the ring is it must be to a place whose temperature is the same as the cylinder head temperature where it is currently measured (under the spark plug). So what you will need to do is use one sensor lead as a trial lead and try your selected spot and compare it to the standard one under the spark plug. If they are the same then it will be OK.
I did a trial on mine and tapped a thread into the centre hole between the plugs and attached a lead. Trouble was when I compared the temps the new position was no where near the same temp as the under plug temp. It was the same till the CHT got to about 85 degrees and then it lagged more and more behind till at operating temps it was measuring 15 to 20 degrees too low.
Basic Cooling Tutorial – Air Cooling & Oil Cooling Vs Heat Production
The final measured CHT is the result of two processes. Heat production from burning Air:Fuel mixture and removal of the heat by the oil and by air entering the engine bay, and blowing over the cylinders and out through the lower cowl opening.
This obviously forms from burning fuel and air. But the heat given off varies according to the Fuel:Air ratio. If we start from excessive fuel and insufficient air (over rich) the mixture won't burn at all. If we gradually decrease the amount of fuel and increase the amount of air we get burning. Basically when the mix is rich there is unburned fuel left and the heat given off is less. As the ratio decreases (ie: less fuel: more air) the heat given off increases till some ideal ratio occurs (maximum burning = maximum power production). However if you continue to lean the ratio the temperature actually decreases again (so called “lean of peak”) till it then becomes so lean it fails to burn and the engine stops.
If we had an engine made out of heat impervious material we could run the engine at the ideal mixture that gives complete burn with maximum power output. But unfortunately we are limited by the fact that high temperatures are destructive on engine components. So we have to run at some sub-power level that gives a compromise of power and temperature.
A Very Basic Carburettor Lesson:
The carburettor is where fuel gets added to the mass of air (in hopefully the right ratio). In “normal” carburettors the pilot has a control which alters this ratio to account for changes in air volume at altitude (which requires a leaner mixture). The Jabiru uses a Bing Type 94 constant depression carburettor which is supposed to compensate for changes in altitude automatically. “Constant Depression” refers to the internal workings which adjusts the position of the needle according to the ambient air pressure (and thus altitude.). Adjusting the needle position changes the amount of fuel that can be delivered thus making the Fuel:Air ratio more or less rich. The pilot is removed from the loop and cannot make any input into fuel:air ratios. Altering Fuel:air ratios manually is dependent on the size of the needle and seat that are installed. The mechanic is required to use particular sizes and settings to cover the usual range of altitudes used most commonly by the pilot.
In the ideal situation air gets pulled into the induction system by the vacuum caused by the piston going down. It would flow smoothly along the intake hose through a filter to keep out bugs, dust and stuff. As it passed through the carburettor it would be instantly mixed with the fuel which is also sucked into the flow by the low pressure (partly from the piston drop but also partly because the throttle through-tube is slightly narrowed which produces a low pressure by venturi effect.)
Ideally this fuel is instantaneously and completely mixed with the air so that each milliliter of air has the same concentration of fuel as every other milliliter. Then every bit of the fuel:Air mix will travel evenly dispersed to all the cylinders and each cylinder will get exactly the same amount.
That is the ideal. Sadly the Jabiru falls way short of ideal.
Here is the “real” outcome for the unmodified Jabiru.
Air gets entrained into the induction system and flows into the Hot Air Mixer Box where it gets swirled around. It then passes down the Scathose (In common terminology this refers to the red corrugated tubing made from a cloth tube held open by a wire spring. The tubing is often referred to by its size in one quarter inch increments. The size used in all the Jabiru applications is 2 ¼ inches or Size 9. Hence called “Scat 9” tubing. ) The air once it exits the hot air mixer box is not flowing smoothly but tends to begin to swirl/spiral down the tube into the carburettor throat. It picks up the fuel from the carburettor and it continues to swirl as it exits the carburettor. Immediately after the carburettor there is a wide chamber in the centre of which is a vertical aerofoil shaped metal divider whose role is supposed to be to stop the swirl and to direct the air in a more laminar flow. But apparently it does not do a very good job as evidenced by the common problem of rich air often going to one side and lean air going to the other.
This is the basic cause of the problem of uneven distribution of Fuel/Air and has been the focus or builders and Jabiru Factory's attempts to fix the problem.
So what has been tried by me and by others?
1.You absolutely MUST follow the manual about ensuring the inside of the hot air mixer box has no sharp edges and angles. These will cause flow disturbance and may aggravate the swirling effect.
2.The Scat hose should be as short as possible to avoid the spring corrugations bunching up on the inside of the curves and adding to the turbulence. There have been attempts to make Scat hose with an inner smooth lining but that has lead to issues with the lining coming away and blocking the tube with engine stoppage following.
3.The Scat hose should be as straight as possible.
4.Just prior to entering the carburettor the Scat hose joins onto a short curved section of fibreglass tubing called a “Cobra Head”. There are three designs for this piece. The “standard” which is about 75 mm long, about 54 mm in diameter and has a bend of about 20 or 30 degrees in it. There are apparently other designs. One with a much bigger angle for non-jabiru airframes and one with a spherical mid-section made to try to alter the airflow characteristics. I have never actually seen these two but have heard about them.
This Cobra head should be/can be modified with a number of fibreglass plates to attempt to stop the swirling. I have now built three different models. One with a single vertical plate, a second with a crossed pair of plates and finally a complete grid of plates with roughly 15 mm squares.
Just to make life difficult, on one engine the grid made a complete cure of all the issues but on the other it was slightly worse than the crossed plate version and we reverted to a crossed plate.
5.The carburettor can be tilted. This requires having six CHT gauges to work out which way to tilt it. Essentially if you tilt the carburettor with the top AWAY from the hottest side you will get a drop in temps on the opposite side. There is not much room to move it, due to the top of the carby impinging on the distributor cap but basically it only needs to go 5 – 10 degrees.
6.I am assuming at this stage that the carby is properly fitted with the correct size jets etc and that there is nothing more to do with the carby. Carby issues and workings are another complete subject on its own. It is well worth getting on the Bing USA site (www.bingcarburetor.com) and buying their tuning and parts manual for the carby. Its a bit of an eye opener. Sadly they don't have it for download (at time I wrote this) you have to purchase a hard copy but it's well worth the $10.
7.I have read, but not tried myself that the metal aerofoil in the chamber after the carby can be removed and replaced with a cylindrical (12.5 mm diameter is quoted) metal rod. I can't vouch for this but it has been mentioned on the Jabiru forum on the recreational aviation website. Jabiru are working on a new plenum chamber (photos also to be found on the Recreational aviation website in the Jabiru section.)