Considerations in Engine Cowl Design

[skippydiesel]

I am a few day's, to a week away, from starting the design of my Sonex /912ULs engine cowl. Like any designer I need to set the parameters/goal of the final product.

 

Note: This Sonex has its;

 

• Rotax coolant radiator mounted, by the original builder (see photos), in such a way as to be wholly dependent on cowling exit air for it to be effective. I am committed to following through on his ideas in this area and very interested to see how it will work in practice.
• Muffler and tail pipe shrouded (see photo), with the idea that the exiting exhaust gas will act as a venturi, drawing cooling air past the muffler and out of the cowling (reducing under cowl heat loads).

 

So I would suggest that the cowl for my aircraft should;

 

• Have a minimal frontal area, consistent with internal clearance for engine movement.
• Have minimal sized air entry vents, sufficient to meet the cooling requirements of the engine, coolant radiator, oil cooler and exhaust system.
• Excellent air seal (fit), so as to force most entering air, to be vented through the coolant radiator & the exhaust shroud.
• Be sufficiently rigid so as to resist the potential for air loads to distort the structure.
• As light weight as possible consistent with other requirements listed.
• Have such access "doors" sized & located, as is necessary for pre-flight checks/inspection.
• Readily removed for engine inspection & servicing  (split roughly upper 1/3 & lower cowling 2/3)
• Protect the engine from weather both in flight & on the ground.
• Be aesthetically pleasing.
• Other  -  at your suggestion

 

At this conceptual stage in the planning/design, I think I can pretty much take care of all but the second point - Minimal sized air entry vents/inlets ;

 

How big/small should I make the two (only at this stage) air inlets, either side of the spinner/front of cowling ????.

Note: My exit air vent is limited to the size of a standard Rotax 912 coolant radiator (plus a little out of the exhaust shroud).

 

I have no prior experience or any real theoretical knowledge of what goes into sizing of entry air vents, other than an understanding that the air volume & flow must meet the cooling needs of the engine both on the ground & in the air and that the exit venting (aperture) is as important as the entry.

 

I am open to all suggestions/factual comments, constructive help of any kind.

 

 

[Bruce Tuncks]

I like and agree with your wish-list skippy. With water-cooled  heads, you shouldn't have too much of a problem keeping things cool enough. My suggestion is to shamelessly copy sizes etc from a known good-performing plane with that engine. The Dimona comes to mind for me.

If I were starting from scratch, I would look carefully to see if simple cowl-flaps could be put in too, but this is just a thought bubble.

In a recent eaa article, they paid attention to streamlining the air exit from the lower cowl. Wow, drag can occur anywhere.

 

[cscotthendry]

The cowling on my Legend matches all the specifications you laid out. It works well on all but the hottest days after a heat soak. There has only been one time when my engine looked like overheating on the ground. The situation was this; My wife had gone for a TIF with and instructor in our plane and had landed. The plane sat for 1/2 hour on the ground afterwards and then we started up and taxied out to the runway. It was 35C on the ground and we were on bitumen taxiways. There were 5 aircraft in the circuit so it took quite a while for us to catch a break where someone wasn't on short final. The temp gauge went up into the red, but as soon as we started rolling at full throttle I could see the gauge dropping like a dead fly. By the time we lifted off the ground, the gauge was back in the yellow and headed south. AFAIK the engine suffered no lasting ill effects from this.

However, I have a normally aspirated, carburetted Rotax 100HP. The same style cowling as mine was used on two subsequent Legends with injected engines and they had a lot of overheating issues and had to have the cowlings modified.

So the moral of the story is that the cowling design isn't just a straight numbers exercise. The environment and the engine characteristics probably will require some trial and error no matter how much you plan and calculate.

Edit:Something to consider. On my cowling and on other cowlings I've seen with the radiator at the exit of the cowling, there is always a flare around the exit hole which I think creates a low pressure area around the hole and increases the flow a little. The flare is always small and not a pronounced angle. I'm guessing that is because at 100Kts, it doesn't take much to be effective. PS, I have never had any cooling issues even in the longest continuous climbs on the hottest days.

Edited January 5, 2022 by cscotthendry

[skippydiesel]

1 hour ago, cscotthendry said:

The cowling on my Legend matches all the specifications you laid out. It works well on all but the hottest days after a heat soak........................................................................................................................................................................................................................................................................

 

Edit:Something to consider. On my cowling and on other cowlings I've seen with the radiator at the exit of the cowling, there is always a flare around the exit hole which I think creates a low pressure area around the hole and increases the flow a little. The flare is always small and not a pronounced angle. I'm guessing that is because at 100Kts, it doesn't take much to be effective. PS, I have never had any cooling issues even in the longest continuous climbs on the hottest days.

I have looked at photos of your aircraft - The two upper inlet vents look to have a similar cross-sectional area as that of the Rotax radiator. If this be correct (?)  it is confirming my own thoughts on the matter  (that is entry air openings, same - slightly larger than radiator exit air opening). However your aircraft also has a lower vent  (for oil cooler?) which changes things considerably - by the look at least + 35%.

 

As to the exit air "lip" yes this is a common feature and as you suggest  is designed to create a venturi (low pressure zone) like effect to improve air extraction from the cowling.

 

My previous aircraft, the ATEC  Zephyr (same engine), has a similar cowling to yours, but with with a larger fixed exit air opening  -  I never had any cooling (usually the opposite) issues but then  I rarely took off in temperatures higher than 30C (my personal limit). On a summer trip, where forecasts were for high temps, I  " managed"  the hot weather by taking off at first light, landing after 2-3 hrs, still in the cool of the day for next TO. Second leg would usually see a stop for extended lunch break  Further" opps" late arvo.

[skippydiesel]

2 hours ago, Bruce Tuncks said:

I like and agree with your wish-list skippy. With water-cooled  heads, you shouldn't have too much of a problem keeping things cool enough. My suggestion is to shamelessly copy sizes etc from a known good-performing plane with that engine. The Dimona comes to mind for me.

If I were starting from scratch, I would look carefully to see if simple cowl-flaps could be put in too, but this is just a thought bubble.

In a recent eaa article, they paid attention to streamlining the air exit from the lower cowl. Wow, drag can occur anywhere.

 

Thanks Bruce,

 

I agree using existing successful designs is the way to go (no need to reinvent the wheel) however I dont know of any Rotax set up that uses a rear mounted coolant radiator,solely  dependent on cowling air for its operation. 

 

All the Rotax  cooling set ups, that I know of,use front mounted radiators or have a large duct from the front directing dedicated cooling air to the radiator. The exception might be the Pipstel Virus SW which has a rear top mounted radiator but still with a dedicated air inlet.

 

The set up I have inherited "looks" to handle exit air very effectively - time will tell.

 

Your cowl flap idea is certainly feasible. My son & I have already speculated on a simple pilot operated shutter/flap to control air entry to the radiator (assuming over cooling at cruise)

[rodgerc]

Perhaps consider locating a NACA vent (or blast tube) strategically where it will provide cooling to your electronics....Specifically, ignition modules, regulator and condenser. 

[Old Koreelah]

Skip has anyone mentioned heat tolerance of resin type? Most common epoxies go soft above 60C and after shutdown the cowl above a hot engine cops lots of heat.

After a dozen years of regular “cooking”, my cowl still seems very strong. Maybe the epoxy has become “heat treated”. Maybe it survived because Inadding a couple of layers with Vinyl Ester resin, which can tolerate much higher temperatures that other resins. 

[skippydiesel]

2 hours ago, rodgerc said:

Perhaps consider locating a NACA vent (or blast tube) strategically where it will provide cooling to your electronics....Specifically, ignition modules, regulator and condenser. 

Good point - perhaps having the front two vents very near top of cowling would create a relativly cooler top environment (where he modules are).

 

I have located/mounted regulator & condenser on cockpit side of firewall.

 

[skippydiesel]

1 hour ago, Old Koreelah said:

Skip has anyone mentioned heat tolerance of resin type? Most common epoxies go soft above 60C and after shutdown the cowl above a hot engine cops lots of heat.

After a dozen years of regular “cooking”, my cowl still seems very strong. Maybe the epoxy has become “heat treated”. Maybe it survived because Inadding a couple of layers with Vinyl Ester resin, which can tolerate much higher temperatures that other resins. 

 I invite you to "wax lyrical" on the subject -  What are the bests composites for a cowling ??

[Geoff_H]

Many years ago the epoxy landing gear on the Cost IV often sagged with time.  The cure was to 'cure' the epoxy by constraining the legs and heating them for several hours without any load on them.  The suggestion was to stop them in black plastic and put them out in the summer sun.  It was claimed that there was no sag on undercarriage legs so treated. I don't have details not numbers.  It does make a little sense when you look at prepeg fibreglass with uncured fibreglass is placed in a mould then heated to cure it.  

[skippydiesel]

So Vinyl Ester is  the right composite system for the job??

[Geoff_H]

Personally I think it is but I am not writing from any point of knowledge.  It definitely has more strength at temperature.  95% of what I use is epoxy.

[RFguy]

Sean,

if you can effect the radiator facing the airflow, IMO it cannot be beaten .

There is no substitute for frontal area and inches of water slamming into the radiator.

My guess is your setup will need about 1.3kg of air per second, 

 

and at a density altitude of 5000 feet, that's 34kg per square meter at 80 kts  (full power Vy climb, and hot say) 

so for 1.3kg/34kg = 382 cm2 of radiator area facing the airflow to get the cooling you need.

Most methods of like what you have, relying on the neg pressure and exit air,  IMO are ineffectual.

There is a reason that spitfires etc hang their coolers in the wind. cant be beaten.

 

[Old Koreelah]

5 hours ago, skippydiesel said:

So Vinyl Ester is  the right composite system for the job??

Skip I used Vinyl Ester (Derakane) for several heat-affected jobs after consulting a chemical engineer retained by my material supplier, FibreGlass International. Can’t remember the figures, but he assured me it was very tolerant of heat, as well as fuels and ethanol. Might be worth asking your supplier.

[skippydiesel]

5 hours ago, RFguy said:

Sean,

if you can effect the radiator facing the airflow, IMO it cannot be beaten .

There is no substitute for frontal area and inches of water slamming into the radiator.

My guess is your setup will need about 1.3kg of air per second, 

 

and at a density altitude of 5000 feet, that's 34kg per square meter at 80 kts  (full power Vy climb, and hot say) 

so for 1.3kg/34kg = 382 cm2 of radiator area facing the airflow to get the cooling you need.

Most methods of like what you have, relying on the neg pressure and exit air,  IMO are ineffectual.

There is a reason that spitfires etc hang their coolers in the wind. cant be beaten.

 

No doubt all correct, however as with any  "performance' machine there will always be compromise or if you will the best result from trying to balance many diffrenet competing factors.

 

I have "inherited" my aircrafts cooling system as is,  I am determined to give it and its designer my best shot at getting it to work.

 

So what size air inlets to your computations suggest ?

 

The following quote , from the Sonex Forum, on this very topic may assist (take special note of his comments for 5200RPM):

 

“......................................................... Rotax states the radiator needs 0.75 cu.m/sec. of airflow (28.58 cu.ft/sec.)
If I make an opening of 5" x 9" that should give the required airflow at 54kts or 4" x 8" would require 76 kts to give enough airflow.
The prop blast is morre interesting, at 5200 rpm the airflow is almost double what is required for a 4" x 8" opening. I wonder if this would work out in the real world though, ........................................................”

 

[Thruster88]

I have one question, what are the thermodynamic qualifications of the previous builder. I feel that feeding hot air from the engine cylinders, oil cooler and exhaust system through the exit radiator has little chance of success. Van's aircraft could have done the cooling on the 12 any way they liked, their system is just like everyone else's. 

 Our club 912 aircraft is only just on a hot day with front mounted radiator and oil cooler.  

[RFguy]

I agree with Thruster .

Well, I still recommend 382 cm2 as before, is the opening I would suggest to cope with sustained 80 kts full power climb at 35C, DA=5000.

OK, so Rotax are suggesting 0.75cuM/sec (900g/sec at STP)   but at what DA, what temperature  etc ? sealevel, STP/std atmo etc ?

I beleive you need   1.4x factor on  STP to cope with hot and high DA.

I think the Rotax numbers are low for hot Aussie condix.  

 

Rotax DID increase the size of the recommended radiator in recent times to about what I am suggesting.  they suggest 485 cm2, but this is also for the 912is and 914 turb.

https://www.rotax-owner.com/en/rotax-blog/item/25-newrad

SO, it would seem my numbers on proportion (921 ULS 100hp vs  914 115hp) are very close  to the rotax numbers.  Incidently, I calculated for my setup 440 cm2, rotax recommend 447cm2 ! 

 

Sean, your 4x8" 80 kts you calced  is 200cm2 (sea level/ std atmos ?) which I think is probably fine for 55% cruise and moderate summer climate/DA. 

AREA required is proportional to power output, approx . I think you are about 50 % of what is really required.

It does take the system a little while for the temperatures to climb , and heat soak, so it might be OK to get off the ground, reduce power and go cruise climb at higher airspeed.

I've proportioned the numbers for heat soaked, full power no-holes barred ops. 

 

Fortunately as the cooling fluid system gets HOT, radiator performance increases due to the increasing difference temperature between the inlet air and the cooled fluid/device.

 

I assumed the prop blast didnt assist, since it can be rather turbulent  and non laminar around the cowling and prop root. 

Anyway, I've done the numbers.... and happy to bet my house on them AND my numbers for frontal area approx equal Rotax's own numbers.

 

 

Edited January 6, 2022 by RFguy

[skippydiesel]

1 hour ago, Thruster88 said:

I have one question, what are the thermodynamic qualifications of the previous builder. I feel that feeding hot air from the engine cylinders, oil cooler and exhaust system through the exit radiator has little chance of success. Van's aircraft could have done the cooling on the 12 any way they liked, their system is just like everyone else's. 

 Our club 912 aircraft is only just on a hot day with front mounted radiator and oil cooler.  

In my view the "previous builder" is an inspired amateur, who has consistently consulted, throughout the build, with the aeronautical experts available to him. The Sonex aircraft was designed to use the air cooled AeroVee engine, only very recently starting to facilitate the Rotax 9 range. This has presented those builders, who wish to use Rotax 9 engines, with considerable engine installation challenges. My "previous builder" has come up with a concept (not unique) that is a little outside the box - if it works it has the potential for low drag/increased efficiency. If it doesnt work? well I will have to spend time & money converting to something like your illustration.

 

Experimental aircraft; Since when did we have to adhere to convention. Sure it behoves us to take note and ponder the rationale of the tried and true but experimental is the opportunity to explore .

 

Your illustration: Aircraft manufacturers principal driver is cost. That is minimal to them maximum to you - if they can install a cheap system that works and sell you an expensive aircraft, the equation is in their favour. I would sagest that the cooling system you have illustrated is a simple. cheap workable system - not particularly efficient in terms of cooling or drag.

 

The coolant radiator installation in my Sonex would likely be a more expensive system for a manufacturer to install on a RAA class aircraft. 

 

Your comment  "feeding hot air from the engine cylinders, oil cooler and exhaust system through the exit radiator has little chance of success" would be reasonable, if the air is too hot to cool the liquid in the radiator. There is no question that the air will be heated above ambient however the "previous builder" has partially addressed this by having a shrouded exhaust muffler/tail pipe and ceramic coated headers, thus minimising (he & I hope) this source of air heating. Now we come to my contribution, to the cooling challenge- the cowling design. The system is dependent on having a sufficient (no more) air supply - hence my question regarding inlet air sizing.

 

[RFguy]

Hi SKip

That's all fine but the math doesnt lie.

 

AND the temperature difference between the inlet air to the radiator and the coolant temperature is a critical. If you heat the air hitting the radiator by 10 deg C it will be like flying at 40 deg C instead of 30 deg C all the time.... 

 

AND AND AND

 

There is a MINIMUM cowling input mass airflow that needs to be satisfied.  A small area air inlet and a huge radiator in the cowling wont work. Not enough mass air flow AND because the radiator is a pressure drop, the seal around the radiator you proposed on the cowling lower  must be very tightly sealed up.

1kJ of heat will raise 1kg of air at STP by 1deg C. So, there is a minimum amount of air you need.   An air dam/scoop has to intercept x kg of air per second.  (and either carry it through at same velocity or reduce the cross section and increase the velocity (and handling losses).

I am (likely) choosing to put almost my whole 490cm2 raidator into the airflow. I *could* use a smaller radiator and have a big air scoop  (min 450cm2 for me) and a reducing cross section (funnel)  and hit a smaller radiator. 

 

Also SKip, there are some interestign shapes you can get for motorcycles, that would lwork under the prop but still give you nice lines like this 

 

 

Edited January 6, 2022 by RFguy

[skippydiesel]

1 hour ago, RFguy said:

I agree with Thruster .

Well, I still recommend 382 cm2 as before, is the opening I would suggest to cope with sustained 80 kts full power climb at 35C, DA=5000.

OK, so Rotax are suggesting 0.75cuM/sec (900g/sec at STP)   but at what DA, what temperature  etc ? sealevel, STP/std atmo etc ?

I beleive you need   1.4x factor on  STP to cope with hot and high DA.

I think the Rotax numbers are low for hot Aussie condix.  

 

Rotax DID increase the size of the recommended radiator in recent times to about what I am suggesting.  they suggest 485 cm2, but this is also for the 912is and 914 turb.

https://www.rotax-owner.com/en/rotax-blog/item/25-newrad

SO, it would seem my numbers on proportion (921 ULS 100hp vs  914 115hp) are very close  to the rotax numbers.  Incidently, I calculated for my setup 440 cm2, rotax recommend 447cm2 ! 

 

Sean, your 4x8" 80 kts you calced  is 200cm2 (sea level/ std atmos ?) which I think is probably fine for 55% cruise and moderate summer climate/DA. 

AREA required is proportional to power output, approx . I think you are about 50 % of what is really required.

It does take the system a little while for the temperatures to climb , and heat soak, so it might be OK to get off the ground, reduce power and go cruise climb at higher airspeed.

I've proportioned the numbers for heat soaked, full power no-holes barred ops. 

 

Fortunately as the cooling fluid system gets HOT, radiator performance increases due to the increasing difference temperature between the inlet air and the cooled fluid/device.

 

I assumed the prop blast didnt assist, since it can be rather turbulent  and non laminar around the cowling and prop root. 

Anyway, I've done the numbers.... and happy to bet my house on them AND my numbers for frontal area approx equal Rotax's own numbers.

 

 

(Just read your last  X message) 

 

RFG - The standard Rotax radiator has a frontal area of about 49500 mm2 (330X150mm) . My initial thoughts were to just divide this by 2 for sizing the front cold air  inlet vents. I then remembered its not just radiator frontal area, its cooling surface area and air flow/throughput. Further confusing the geriatric brain and its 1.5 remaining cells is the photos of high performance Rotax 9 powered aircraft, almost all with tiny air inlets - there is much more to this.

[Geoff_H]

"That's all fine but the math doesnt lie."

 

That is if you have the right maths.  After years of calculations and observations of actual air flow in ducts believe me it is far more complex than simple maths.  Everything from zero wall velocity to air viscosity etcetera.  Best idea is to observe what works on a similar installation and copy it

[skippydiesel]

1 minute ago, Geoff_H said:

"..............................

 

.......................... Best idea is to observe what works on a similar installation and copy it

Not so many  "similar installation" like mine out there. I agree in principal.

 

My cautious intention is to try for a cowling design that can be relativly easily modified by enlarging the inlet and/or having a third (chin) inlet.

[RFguy]

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.

[RFguy]

Skip , I would avoiding inlets and  ducting unless you have a 2:1 room for assumptions and deviation from the ideal. Just cut the cowling and fibreglass reform it. 

 

 

[Geoff_H]

My cri cri similar project just puts the cylinder fins in the airflow, hopefully enough flow at idle.  Much simpler, more drag, as said earlier it's all a compromise.