On J230, the heat, and suitability of the Rotax 912ULS

[Old Koreelah]

This morning the rain held off so I got to test my water injector. After taxiing up and down to warm up, I tied the baby to a post and did some full-throttle tests, with a fine spray of pure rainwater down the carby throat.

 

Results: intake air dropped 3 degrees C, but no discernable change in anything else. At least the water didn’t put out the fire.

 

Next test is to open up the nozzle to max flow. That would deliver about 120ml/minute, or about a third as much water as fuel.

Beyond that, the next step is to adopt Arron25’s windscreen washer pump, which should deliver much more water.

[cosmicray]

I recently fitted new pistons to a Jabiru 3300 and found the rear cylinder CHT's, although within Jabiru Australia specification, were getting hotter than US Jabiru recommendations on the ground. So I have been experimenting with water injection using a "Nice" (brand name) windscreen washer pump and a Snow number 4 nozzle. I am getting noticeably cooler CHT's injecting the water between the AeroInjector and input plenum. Have not flown with this arrangement yet but it is looking good so far.

[Old Koreelah]

Wow Cosmic, good to hear. Do you know how much water it’s getting?

That water/fuel ratio is one I’d like to get a pretty good idea of.

[cosmicray]

 

Wow Cosmic, good to hear. Do you know how much water it’s getting?

That water/fuel ratio is one I’d like to get a pretty good idea of.

 

Snow say the number 4 nozzle sprays at 225 ML/MIN but that figure is based on the feed from a high pressure pump. So have to assume something less than 225 ML/MIN with the windscreen washer pump. Tried a number 1 nozzle (60 ML/MIN) and there was no effect at all. Have no idea what the water/fuel ratio is but if you have too much water the engine runs rough and then stops. Have tried high pressure and while you could see a difference in the spray, the Snow 300PSI pump is too big and heavy for my little plane. 

[Old Koreelah]

I have a disused oil injector tube running from cockpit to the rubber collar between carby and plenum; that seems the ideal location for an injector, if I can get the spray properly installed. One advantage is access for adjustment is easier than the current location inside the air filter box.
One advantage I’d lose is the before-carby spray cools the air intake probe, giving me instant notice that the water injector is working.
That display also shows how much the carb heat is warming the intake air (not much at present).

[Old Koreelah]

Yesterday the rice padis were drying out and I got to test-fly my water injector. Full throttle climb past 5,000’ saw CHTs creep up to 156C, where during previous tests they spiked well over 160 in a fraction of that time.

That convinces me that even at only 80ml/minute, the water is helping with cooling.

One interesting change was EGTs; they normally stay fairly even around 650C but while the water was being injected, No. 4 was up around 720.

[facthunter]

Try some methanol in it as well as the water. That's a low rate of water injection. Nev

[Old Koreelah]

1 hour ago, facthunter said:

Try some methanol in it as well as the water. That's a low rate of water injection. Nev

Quite true, Nev. That’s the most my current little pump will deliver. The next step is to fit a windscreen washer pump, but I’d need a valve to control the extra flow- I don’t want to put out the fire!

 

What does the methanol do? I presume it helps avoid icing and adds power;

 I don’t need any more power and I’m wary of any alcohol getting near all those rubber seals.

[facthunter]

Mineral oil affects rubber. Makes it swell.. Castor oil and metho were used as a  temporary brake fluid  and system cleaner if water got into a brake system . .

  I've poured water into a  4 litre petrol engine at about 3 pints a minute at medium revs (3000) and it didn't falter  (as a test)  Check it on the ground (tethered)  Don't do it at the RATE I did and expect some contamination when/if the engine isn't really hot..Nev

Edited November 18, 2021 by facthunter

[Old Koreelah]

34 minutes ago, facthunter said:

Mineral oil affects rubber. Makes it swell.. Castor oil and metho were used as a  temporary brake fluid  and system cleaner if water got into a brake system . .

  I've poured water into a  4 litre petrol engine at about 3 pints a minute at medium revs (3000) and it didn't falter  (as a test)  Check it on the ground (tethered)  Don't do it at the RATE I did and expect some contamination when/if the engine isn't really hot..Nev

Yes Nev, I’ve been ground-testing each stage by tethering the tailwheel to a post.

Why add methanol?

[Bruce Tuncks]

Great stuff old K. I reckon you have solved the problem with spectacular success. Sorry I can't suggest a reason for the egt's being high. Maybe Nev's suggestion about adding some methanol will work. Darned if I can see why though.

Anyway, the whole thing was to use as little water as possible to achieve the bit of cooling needed, and you have done this.

[onetrack]

The beauty of some water injection is also the combustion chamber cleaning effect.

[Bruce Tuncks]

It should be possible to calculate the intake air cooling by working out the volume of air per sec and applying cooling due to water vapourization to this. I'll give it a try when I'm more composed. A smart guy could also adjust for the dilution of the inflow air with the water vapor. Don't think I'm that smart.

 

[Old Koreelah]

1 hour ago, onetrack said:

The beauty of some water injection is also the combustion chamber cleaning effect.

Yep. That’s what got me off the fence. My Jab runs a little rich and has quite a lot of carbon inside. Two doses of Mr. Subaru’s cleaner in a spraycan seem to have helped a little, but I’ve also had the engine run on after shutting off the spark, so the least carbon buildup the better.

[Old Koreelah]

1 hour ago, Bruce Tuncks said:

It should be possible to calculate the intake air cooling by working out the volume of air per sec and applying cooling due to water vapourization to this. I'll give it a try when I'm more composed. A smart guy could also adjust for the dilution of the inflow air with the water vapor. Don't think I'm that smart.

 

Bruce at the moment I’m giving it about 20% water and 80% Shell 98. The fire hasn’t shown any signs of going out and its running a little cooler than expected, but I’ve yet to do proper comparison tests in flight. Plenty of other projects to keep me busy, so I might keep using the current tiny little pump for a few more flights. When the proper hot weather hits I might install a bigger pump.

[onetrack]

Old K - I used a can of Penrite P26 foaming intake cleaner recently on a Ford Ranger, that seemed to be quite effective. I had an EGR fault code on the Ranger, which was related to improper EGR valve operation (sticking), and the P26 seemed to clean up the intake and exhaust nicely. I was able to clear the fault code after I used the can of P26, and the fault hasn't returned.

[facthunter]

Diesels with EGR are something else. Puts some of the exhaust back through the engine. This is to lower combustion temps and counter NOx. It reduces fuel mileage though and INCREASES CO2. as well as making some parts of the inlet tract full of BLACK GOOP.

. Do this before an oil change is advised as they  don't want the treatment traces in the oil.  Doubt I'd advise it for a Jabiru.

  Water will remove carbon on piston tops when injected into hot motors.. Subaru had a solvent for the higher comp motors that were detonating. Back when Subaru's road cars were competitive.  Nev

Edited November 18, 2021 by facthunter

[cosmicray]

The attached PDF provides info on what water and methanol do. methjuiceoptions.pdf

[Old Koreelah]

That article gives me more confidance that my current low rate of water injection will help keep the engine cool, clean and reliable.

The claim that water reduces the chances of detonation is great news for flimsy engines running uncertified fuels.

[Bruce Tuncks]

Well I tried to work out the theoretical temp drop but I must have got it wrong....  help needed!

Here's my attempt:

80 gm/min= 1.333 gm/sec and the latent heat of vaporization is 2260kJ/kgK so we have 2260*1.333*1/1000=3kJ/sec cooling .

Air/sec for a 2.2l engine at 3000 rpm = 5.5l/sec and air weighs 1.293 kg/mcubed therefore 3 kJ will cool the airflow by mass/sec times specific heat of air times temperature drop.

SO delta T= 3/(5.5*1/1000*1.293*1.005

=419 C !!!!

Well I don't believe this answer cos its too high intuitively.  But  I tried!

[Kenlsa]

You have given me a headache Bruce!

Ken

[RFguy]

Ken, sniffing too much Methanol  will certainly give you a headache. You are using it wrong !

[aro]

3 hours ago, Bruce Tuncks said:

Air/sec for a 2.2l engine at 3000 rpm = 5.5l/sec

I think that is too low - should it be 55l/sec?

 

~40C doesn't sound unreasonable, that would presumably be measured as a drop in EGT.

[Bruce Tuncks]

I agree about the 40 degrees

[Bruce Tuncks]

Here's how I got the 55 l/sec. A 2200cc4 stroke sucks 1100 cc/rev and 3000 rpm= 50 revs/sec. So the engine sucks in 1.1times 50 =55 litres/sec.

Please keep trying, there is something wrong somewhere.

It's the tiny weight of the air sucked in per second that throws the calculation off.

 

Edited November 20, 2021 by Bruce Tuncks