Considerations in Engine Cowl Design

[facthunter]

The velocity of the air through the radiator matrix is not going to be as high as the propwash is so the intake can be smaller to get minimum drag. More benefit on fast planes and you will see it o them whether an aircooled motor or a liquid cooled one is installed. This obviously requires sealing so  you have a clue to what's really happening. The outlet can/ has to be a part of it.  Why not really?

  A 9 rotax will overheat if it runs just right in a colder environment and goes to a hotter place.. This is just physics. An adjustable flap at the air exit  will work well to minimise drag, but is an extra control to forget. .Nev

[skippydiesel]

2 hours ago, RFguy said:

Hi Geoff. I agree !

Hence the suggestion of just putting the whole radiator  in the wind...and make it as simple as possible so deviation from the ideal and number of assumptions are both minimal.

Sonex are designed with speed in mind  - "whole radiator  in the wind" !!! might as well throw an anchor out.🤣

[Thruster88]

The Shark.aero as flown by Zara Rutherford is one of the fast ones. It has quite large cowl intakes and a pilot controlled "scoop" on the main naca inlet duct. In google images of this aircraft the scoop is open a fair bit.

 

Putting pre heated air through your radiator will surely increase the total cooling air flow required,  that has to be more drag and less speed. 

Edited January 7, 2022 by Thruster88

[RFguy]

Skippy , Just means that some experimentation and iterations will be required.

At the end of the day, a certain number of kilograms of air is required. 

 

Where that shark.aero has the inlet scoop, that is a fair way from the prop, a bit cleaner airflow -  and might give you some food for thought ---- that is in I would call the transition region , compared to the cooling openings adjacent to the spinner where the airflow is highly disturbed by the prop.

 

SKip, I had thought about having my radiator there approx where the cowling lip is and be able to adjust the amount of tilt of the radiator into the wind. IE from full blow to tilted 'up'  into the cowling gap (relying on the cowling thru air)  

It's not going to change drag much because I am no longer using the cowling suction lip  . But it's likely I will put it just below the prop. 

 

[Bruce Tuncks]

I think it was Old K, but somebody tufted a plenum and found that the air flow directions were sometimes not as our intuition would expect. For example, airflow near the crankcase center was forward !

So I learned from this that you needed to test out theories first, and tufting and photographing was a good idea.

Once, this was done on a glider wing and it was found that the flow on a glider aileron was ALONG the wing towards the fuse on the top surface. A separation lay ahead of this flow.

[skippydiesel]

28 minutes ago, Bruce Tuncks said:

I think it was Old K, but somebody tufted a plenum and found that the air flow directions were sometimes not as our intuition would expect. For example, airflow near the crankcase center was forward !

So I learned from this that you needed to test out theories first, and tufting and photographing was a good idea.

Once, this was done on a glider wing and it was found that the flow on a glider aileron was ALONG the wing towards the fuse on the top surface. A separation lay ahead of this flow.

Yes I have heard of this type of unexpected airflow. Unfortunately I lack both the equipment and knowledge to do anything other than go with possible flawed logic of observation.

[RFguy]

Skip, what size is the radiator that you have ?

 

[skippydiesel]

1 hour ago, RFguy said:

Skippy , Just means that some experimentation and iterations will be required.

At the end of the day, a certain number of kilograms of air is required. 

 

Where that shark.aero has the inlet scoop, that is a fair way from the prop, a bit cleaner airflow -  and might give you some food for thought ---- that is in I would call the transition region , compared to the cooling openings adjacent to the spinner where the airflow is highly disturbed by the prop.

 

SKip, I had thought about having my radiator there approx where the cowling lip is and be able to adjust the amount of tilt of the radiator into the wind. IE from full blow to tilted 'up'  into the cowling gap (relying on the cowling thru air)  

It's not going to change drag much because I am no longer using the cowling suction lip  . But it's likely I will put it just below the prop. 

 

The Shark Aero would seem to have a plethora of inlets & outlets. Are they all functional?  Not sure its obvious what inlet/outlet serves what purpose. Nice looking aircraft somewhat spoilt by too many fish like accents. 

[skippydiesel]

2 minutes ago, RFguy said:

Skip, what size is the radiator that you have ?

 

I believe its a standard Rotax (could be wrong) Its approximate opening (finned area) measurements are 330 x 150 mm

[RFguy]

how deep is the core ?

[skippydiesel]

18 minutes ago, RFguy said:

how deep is the core ?

About 46 mm. Rechecked other dimensions L 330 W 140 (access not great so measurements have been rounded)

[RFguy]

46 on the core is thicker than most. Mine is 40 on the core itself. (CBR1300 motorcycle radiator) 

 

Looking at your original  photos.... OK so you'll need to have pretty good sealing to encourage air through that radiator...

I would suggest minimum inlet size 382cm2, however, since the insides of the plane will heat up the air a little, perhaps add another 10% to that, say 450cm2 inlet area. 

See how 450cm2 goes. You might need 10-20% more again to allow for leaks and gaps .  

Those numbers of course I came up with assumed sustained full power, 80 kts. If you have only plans for sustained full power 100 kts indicated, then divide the recommended area proportionally, IE (80/100)*450.. Note higher airflow/ air pressure inside will likely lead to more leaks so it may NOT go proportionally, precisely.

 

How much assistance you get from the prop on air input will be highly variable on inlet location, so I'd just assume no assistance and go with the above numbers... Geoff what do you think ?

-glen

 

 

Edited January 7, 2022 by RFguy

[Thruster88]

53 minutes ago, RFguy said:

46 on the core is thicker than most. Mine is 40 on the core itself. (CBR1300 motorcycle radiator) 

 

Looking at your original  photos.... OK so you'll need to have pretty good sealing to encourage air through that radiator...

I would suggest minimum inlet size 382cm2, however, since the insides of the plane will heat up the air a little, perhaps add another 10% to that, say 450cm2 inlet area. 

See how 450cm2 goes. You might need 10-20% more again to allow for leaks and gaps .  

Those numbers of course I came up with assumed sustained full power, 80 kts. If you have only plans for sustained full power 100 kts indicated, then divide the recommended area proportionally, IE (80/100)*450.. Note higher airflow/ air pressure inside will likely lead to more leaks so it may NOT go proportionally, precisely.

 

How much assistance you get from the prop on air input will be highly variable on inlet location, so I'd just assume no assistance and go with the above numbers... Geoff what do you think ?

-glen

 

 

I measured the inlets on the RV,  they are bigger than they look, total 640cm for 180hp. Your calculations are realistic. The RV cowl outlet is the same size so this may limit flow. Sustained climb last weekend in hot conditions resulted in 200°c CHT, ok for the lycoming.

 

 

[skippydiesel]

2 hours ago, RFguy said:

46 on the core is thicker than most. Mine is 40 on the core itself. (CBR1300 motorcycle radiator) 

 

Looking at your original  photos.... OK so you'll need to have pretty good sealing to encourage air through that radiator...

I would suggest minimum inlet size 382cm2, however, since the insides of the plane will heat up the air a little, perhaps add another 10% to that, say 450cm2 inlet area. 

See how 450cm2 goes. You might need 10-20% more again to allow for leaks and gaps .  

Those numbers of course I came up with assumed sustained full power, 80 kts. If you have only plans for sustained full power 100 kts indicated, then divide the recommended area proportionally, IE (80/100)*450.. Note higher airflow/ air pressure inside will likely lead to more leaks so it may NOT go proportionally, precisely.

 

How much assistance you get from the prop on air input will be highly variable on inlet location, so I'd just assume no assistance and go with the above numbers... Geoff what do you think ?

-glen

 

 

Radiator core thickness - as I said access is kind of tight - I measured the "tanks" thickness - I assume core will be a little thinner (but not much).

 

O ye of little faith - the cowling will seal effectively, with the only leak of a note being a small amount around the exhaust tail pipe. The whole point of this sort of cooling system is an air tight cowling.

 

"since the insides of the plane will heat up the air a little" - please expand.

 

"assumed sustained full power, 80 kts. If you have only plans for sustained full power 100 kts indicated" - Not quite sure what you are getting at here. The only time I have ever flown an aircraft at full power, is on take off. As soon as safe altitude reached (varies according to situation) I lower the nose & reduce power to cruise climb.

Possibly because this airframe can be fitted with a wide range of engines/ 80 - 130hp,  I am unable to find information for Max Climb, Best Rate of Climb, etc. An additional complication is the CS Airmaster. My expectations are for exceptional TO performance & climb out but that does not change my engine management habits.

I would be surprised if this little bird does not cruise climb at well above 100 knots (one up).

As for cruise, I am hoping for 120 knots at 16 L/hr or better.

In short I do not anticipate sustained high/max continuose power operations to occur very often and then only for a short adrenaline burst.

When specifying the cowl I will try for a cooling regime that ill work effectively in most situations. This is not to suggest that pilot management (reduced climb out angle, leveling out to allow for cooling befor resuming climb, etc) during hot weather departures will  not be required

 

 

[RFguy]

yeah, if you accept pilot management of cooling is required, certainly you have more wiggle room.  I design for continuous WOT at perhaps 10% above Vy

 

Sounds like we use power differently---- I usually climb at full power , I dont dawdle between flight levels.  that will be either a Vy climb, or a shallow angle higher airspeed-good cooling  climb ~ 95 kts  (observing continuous RPM limits) , but generally at WOT.  WOT usually isnt a heat issue, since the engine is at mega-rich mixture.

 

.....and, pilot managed / moderated  for cooling or noise abatement, as required....

 

If the (low) climb rate I want produces excessive RPM, then I'll either climb steeper or back off the throttle (EGTs usually rise from the full rich)

Thermal chasing in the summer can be most rewarding...

 

[Geoff_H]

Skippy

You are after a high  performance cooling system, one which cools the engine continually but presents the least drag on the aircraft.  In my 50 years as an engineer I have not seen a totally new design be got right on the first design.  What I think that you will need is to make a cowling then instrument it and data log the data probably including spaces inside the cowling when miscellaneous equipment such as fuel lines may lurk.  Amongst the temperatures that would need to be monitored is all cht, inlet air temperature exit, air temperatures air flow measurements etc.  I worked on the design of the fluidised boiler at Redbank Power Station, the first of its kind in the world,  we changed that design many times over a year or more before we got it as good as it could be without tearing it down.  But I do wish you the best of luck with what you choose.

Geoff

[skippydiesel]

12 hours ago, RFguy said:

yeah, if you accept pilot management of cooling is required, certainly you have more wiggle room.  I design for continuous WOT at perhaps 10% above Vy

 

Sounds like we use power differently---- I usually climb at full power , I dont dawdle between flight levels.  that will be either a Vy climb, or a shallow angle higher airspeed-good cooling  climb ~ 95 kts  (observing continuous RPM limits) , but generally at WOT.  WOT usually isnt a heat issue, since the engine is at mega-rich mixture.

 

.....and, pilot managed / moderated  for cooling or noise abatement, as required....

 

If the (low) climb rate I want produces excessive RPM, then I'll either climb steeper or back off the throttle (EGTs usually rise from the full rich)

Thermal chasing in the summer can be most rewarding...

 

 

Seems to me pilot management of engine parameters is always desirable.

I dont think we are that different -your speed references confuse me a bit. 

I suspect that's because I have spent much of the last 12 years flying one aircraft that has a very different climb performance to yours.

Full power climb in my Zephyr gets 60-70 knots and 1500ft/min climb,  so cruise climb at 90-100 knots/ 5200 rpm yields any easy 500 ft/min, with two up.

There rarely seemed any need to use continuous full power and anyhow she just seemed to fly so "sweetly" at 52000rpm.

The only time I "step" climbed was on a return trip from Merimbula (sea level), on  hot day, with a large passenger and some equipment, he had not told me, he wanted to bring along. 

 

Edited January 7, 2022 by skippydiesel

[skippydiesel]

So it would seem that the initial design for cool air entry, will be to size the inlets at about the same/possibly slightly larger,  frontal area as the radiator. This will give the option to reduced radiator air throughput , if required, by blanking off (initially with tape) and if oil & coolant temperatures are just out of control (high), creating a third (strategically placed) cool air inlet  (NACA?) lower on the cowling front, directed at the oil cooler But also providing some more air for radiator.

 

Clearances around high points (front cylinder rocker box, exhaust system exit, fuel pump, etc) will be 10 mm, with 3-5 mm between spinner & cowling - see how we go.

[RFguy]

That is a good start- there are many variables, however for a specific amount of energy (heat) removed, this needs a specific weight of air required to heat up and take the energy away.

 

The radiator would have been sized for some airflow (airspeed)  for some power output, and likely have been sized for direct exposure to the air front at slow speeds. So its hard to go wrong if you go with the same area as the radiator. What you do have to ensure is all the air in  the front actually can go through the radiator to get out. Otherwise , as you'll well understand, air will want to escape around the radiator which is  a restriction.. and you'll need alot more air.

 

[skippydiesel]

7 minutes ago, RFguy said:

That is a good start- there are many variables, however for a specific amount of energy (heat) removed, this needs a specific weight of air required to heat up and take the energy away.

 

The radiator would have been sized for some airflow (airspeed)  for some power output, and likely have been sized for direct exposure to the air front at slow speeds. So its hard to go wrong if you go with the same area as the radiator. What you do have to ensure is all the air in  the front actually can go through the radiator to get out. Otherwise , as you'll well understand, air will want to escape around the radiator which is  a restriction.. and you'll need alot more air.

 

Sealing to prevent uncontrolled air exit will be  the top priority.

[facthunter]

Same area as the radiator will be too big. The air is impeded when it goes through the radiator matrix and slows down considerably. Even fly screens slow wind  a lot.. Thicker and finer are the worst offending radiators. At that point the pressure drop must be considerable to get flow. It takes quite a decent sized belt and fan to pull air through a radiator. Have a look at road graders and they are diesel and more efficient.. The prop would impart a swirl to the air in that region and the openings should take account of that. Nev

[RFguy]

It depends on the assumptiosn when the radiator was sized I guess.

 

An airplane flying at 80 kts (41 m/s) into a radiator facind the airflow  is 4" of water in pressure and will drive air through just about any radiator you could dream of. but equally will squeeze around the edges ! 

 

Yes it DOES take quite a big fan and belt but do the math and you realise there isnt any substitute for putting the thing, as I said , out in the airflow.

 

(skip of course the equal size opening and the only way out through the cowling exit through your radiator will also work, assuming the path from inlet to outlet is large and fairly unobstructed.

If obstructed, you'll lose velocity and thus inches of water.

[marshallarts]

I'm throwing this in in the hope/assumption that "there is no such thing a a stupid question..."

 

We all know that modern ICE-powered cars have thermostatic electric fans to push or pull air though their radiators.  Given that the Rotax is a liquid-cooled engine with a radiator, is this something that could be considered in this situation?  I guess it's one more thing that could fail, and maybe a fan (or fans) could never move a big enough volume of air for an aircraft engine.  But it seems to me that it/they could make some contribution.  Just thought I'd ask.

[RFguy]

fans in cars  are only in there to deal with very low power output at low airspeed. traffic. idle.

consider at the 500cm2 radiator at 41 m/s (80 kts) , is intercepting some 4300 cubic feet per minute.... now go and look  up some thermofan performance into say at least 1 inch of water....

 

Edited January 13, 2022 by RFguy

[marshallarts]

OK, I thought it might be something like that.  Those damn laws of physics - you just can't get around them!  😄