Hot rear cylinder example Jabiru 3300A-2575 and plenums

[Old Koreelah]

15 hours ago, RFguy said:

You'd be suprised of the head material  temperature gradient if being forced air cooled.
 

the top screw mounted CHT thermocouples cop a real cooling blast...

I installed the neat little stainless steel insulating covers for CHT probes sold by Ian Bent of CAMit.

Years later, inspired by discussion on another forum, I added shrink wrap to the sensor cable to shield them from the air blast.

 

The temperature readings shot up 10-15C.

After lots of retesting with other thermometers, I’ve come to the sobering realisation that the higher readings are closer to the real temperatures inside the head than I’m happy with.

[RFguy]

Alright everybody - here are my head suggestions.... I 'm posting in anpother specific Jab engine group

- you do not need to be logged into view them

- you might as well look at the photos there instead of double posting

-based on the profile photos, they show where the air must go- and how to do it...

this will only work if there are no leaks where the air can take an easy path.

 

https://jabcamit.groups.io/g/main/message/13777

https://jabcamit.groups.io/g/main/message/13778

https://jabcamit.groups.io/g/main/message/13779

 

 

 

[facthunter]

Without a fan assisting, I can't see the volume of air you will get being enough to do the job. Nev

[RFguy]

the plenum is the pressure box .... plenty of head pressure there.

Now, this will only work IF there are no big leaks, as this needs a fair pressure drop to get the airflow. 

 

BUT- the utilization of the fin surface will  be far higher.

The question is will getting the utilization up from the essentially doing-nothing fins be better than lots of air over a small number of the fins ?

SUggestions for a high temperature bead  or rectangular section gasket I can use to push into the tops of the deep fins ?

There is no/little room between cylinders, needs the bead in the fins or 0.1mm thick ally sheet 

 

[facthunter]

Yes  it is very restricted.  Can you get blast tubes over the top of the rocker boxes or from forward?  Nev

[RFguy]

could use blast tubes- but the bulk of the air needs to go down the middle -  deep- yeah I could blast directed air down there, but in a pressure box, the drop drop will do that for me. 

 

-and the only way this will happen is with that method... no other way.....so  restricted is important- otherwise all those  fins do bugger all. 

I can measure the pressure drop to meet airflow  required for this on my bench... probably need to. 

[facthunter]

You are very limited on how much you can  force through there. Squared off fin edges impede flow a lot The screws in the rocker covers can be used to locate   substantial  ducts. Air from between the cylinders  directed to the exhaust port area? That's your hot spot. Near the exhaust valve.  Cylinders show BLUE  there. Nev

[Old Koreelah]

1 minute ago, facthunter said:

...Air from between the cylinders  directed to the exhaust port area? That's your hot spot. Near the exhaust valve.  Cylinders show BLUE  there. Nev

Updraught cooling may be ideal for the Jab engine, maximizing inflow on steep climbs when it’s most needed. The hottest bits get the cooling airflow first; what is the effect on the rest of the heads?

[RFguy]

the fins are clearly designed for TOP to bottom cooling, it's obvious when looking at the finning and shape  deep in....

[RFguy]

I've run the numbers, needs 300 Pa pressure drop to work (1.2" of water) so with 3" available, even with gross errors  , there is enough pressure drop to work .

that assumes the flow in the fins is laminar, which it may well be not at a fin airspeed of 25 m/s......

- non laminar flow will cost pressure drop margin. 

 

 

 

[Kenlsa]

Glen,

this is what I did after a heap of testing.

I ensured a gentle slope for the air transition instead of a “step”

Not shown are the gull wings underneath on the cylinders.

Note the rubber grommet— this was required to reduce the standard factory hole size (19mm-3/4in) to slightly increase the amount of air staying in the plenum 

Overall no lowering of the temps on the front heads but from memory a drop of 15 deg on the rear heads. 
better temps on climb as well, 85kts@3000rpm to 1500ft, then cruise climb at 250 to 300 fpm @ 2800

[RFguy]

Ken

I like the 3rd pic , showing the metal sheet over the head caps . My thoughts exactly . that is low hanging fruit. 

I like the meringue work. yes the step would generate all sorts of nasty airflow behaviour + drag..

 

 

[Kenlsa]

Further to the above, I did try increasing the length of the plenum at the rear.......VERY bad. It increased the rear head temps by 15deg,

Spoke to Ian Bent about it and as it followed the 90degree bend at the rear it increased the drag slowing the air too much, making a stagnant air pocket.

Ken

[RFguy]

that's right- the simple CFD model  I did showed the generation of a low pressure ' well' at some air speeds depending on back wall distance 

 

need to break up that eddy  with some choice vanes.  a set of steps in the rear will do that... make sense, sort of ?

 

 

 

 

 

 

 

 

[Yenn]

Two things. Updraught cooling of any aero engine looks attractive as hot air rises, but you have to be aware that the heated air has to exit somewhere. If it is at the top there is a good chance that any oil blowby or worse leaking oil, will end up on the screen.

The problem you are having seems to be that the air goes down through the fins until it gets to the centre of the cylinder and then it can get away from the fins beneath the cylinder. As someone earlier suggested the answer to that is gullwing ducts beneath the cylinder to keep the air flowing between the fins.

[kgwilson]

Personally I think you are overthinking the whole process. My Gen 3 3300A has always run cool even on the hottest days and sustained climbs at full power. I use exclusively Mogas.

 

What I did

1. Installed a large 7 row Positech oil cooler on the Firewall with a sealed NACA ducted airflow direct to it with its own exhaust.
2. Installed the standard old style Jabiru Plenums with moulded intakes glassed to the front to fit my Intake nacelles. The Jabiru idea of having a dam wall at the front to direct air upwards causes more problems than it solves with eddies restricting the airflow.
3. The intake nacelles and moulded plenum intakes are sealed and smoothly shaped to maintain good laminar airflow into the Plenums so air cannot go anywhere but into the plenums. There are no spark plug cables to disrupt the airflow and each plug cap has a tight fitting O-ring so no air can escape there.
4. The moulded plenum intakes curve around to be directly in front of the cylinders and a 25mm hole also allows airflow across the block.
5. Each plenum has glassed in vanes to deflect air downwards between the cylinders. 16mm shaped pool fence aluminium tube is glassed in to direct cool air on to the coils.
6. A 1.2 metre wide air exhaust with a bottom cowl lip for 1 metre to create suction.
7. A 50mm diameter hole in the lower cowl directly in front of the sloping sump ribs.
8. The nose leg and exhaust pipe have a close fitting cover screwed onto the bottom cowl to reduce turbulence.

Photo 1 shows the air exhaust & lip with the cover around the exhaust & noseleg

Photo 2 shows the oil cooler NACA duct and suction lip

Photo 3 shows the moulded intake nacelles during the build

 

[RFguy]

excellent post KG..appreciate your efforts to shopw and explain .

 

Hey what did you do with the airflow over the sump fins ??

 

In my 230D, there is a deliberate inch gap above the oil cooler to put air over the sump fins. 

 

Compared to the oil cooler I imagine they dont do much (surface area comparison). It is also not constrained over the sump fins at all so abotu 10% of that bypass air does any work on the sump fins.... . That inch gap is a considerable airflow into the lower cowling that will screw up the attempts to neg pressure it - 

?

.agreed on the front dams.  bad for airflow. drag etc

 

 

[kgwilson]

There is no constraint as my oil cooler is on the firewall with it's own duct so the 50mm hole is only 20mm from the sump as in this construction photo.

[RFguy]

geez that's done nicely.

[kgwilson]

On 26/01/2021 at 4:32 PM, kgwilson said:

It is the standard Jabiru setup with a shroud around the exhaust before it enters the muffler & scat hose to the airbox so the air hot or cold is always filtered unlike Lycomings where hot air is unfiltered.

I told a lie there. It is the standard Jabiru setup but all 6 exhausts go into the muffler and there is a hat shaped piece welded on to the back of the muffler with air entry at each end and an outlet in the middle where the scat hose sucks hot air into the airbox when carb heat is on.

[Old Koreelah]

2 hours ago, Yenn said:

Two things. Updraught cooling of any aero engine looks attractive as hot air rises, but you have to be aware that the heated air has to exit somewhere. If it is at the top there is a good chance that any oil blowby or worse leaking oil, will end up on the screen.

Yen I presume this response is in reply to my post this morning. I agree that exiting the air above the engine could because problems for pilot visibility; that seems to be why it’s only used on pushers or twins.

 

The “hot air rises” concept is only a minor reason for my interest in updraught cooling. A major attraction is that my cooling air exits into the low pressure zone above the wing. Another is that the hottest bits will feel the cooling air first. The main perceived advantage is that updraught cooling should work best during steep climbs, when the standard set-up is most stressed.

2 hours ago, Yenn said:

The problem you are having seems to be that the air goes down through the fins until it gets to the centre of the cylinder and then it can get away from the fins beneath the cylinder. As someone earlier suggested the answer to that is gullwing ducts beneath the cylinder to keep the air flowing between the fins.

I have gull wings both above and below, as well as close-fitting ducts around all four cylinders. The air can only escape via the fins.

 

I suspect the main problem is the big air leak into what should be a slight vacuum below the engine- caused by the oil cooler. 
As outlined in an earlier post, I am in the process of completely separating this airflow.

[facthunter]

Not neglecting the excellent work but it's clear the faster aircraft types have less issues with cooling on a Jab. engine installation and I've always found increasing climb speed to have a significant effect. (Done BEFORE the temps get too high).. Hot air rising. Yes it does but the  effect in a confined area of minimal height is pretty miniscule. A tiny bit of pressure difference will do much more. Instead of "oily belly" you'll have oily windscreen. if you vent upwards. Nev

Edited January 27, 2021 by facthunter
more content.

[tillmanr]

Should we differentiate between climb speed and speed when climbing. Climbing at 65 knots while giving good height gain also sees the engine temps climb. To climb with the same revs but at 75knots will see lower engine temps. There is more air flowing over the fins and less engine load,

[kgwilson]

1 hour ago, facthunter said:

Not neglecting the excellent work but it's clear the faster aircraft types have less issues with cooling on a Jab. engine installation and I've always found increasing climb speed to have a significant effect. (Done BEFORE the temps get too high).. Hot air rising. Yes it does but the  effect in a confined area of minimal height is pretty miniscule. A tiny bit of pressure difference will do much more. Instead of "oily belly" you'll have oily windscreen. if you vent upwards. Nev

I agree with that. The slowest I ever climb out at is 70 knots. Mostly I climb at 80 knots as this gives the best rate and angle of climb combination with distance travelled.

[facthunter]

In the right airframe your rate of climb is hardly diminished. Your angle is though. If you are headed places a cruise climb is used often.( Need a clearance for that if in CTA.) Nev