Mooney Aerodynamics

[takair]

16 minutes ago, Cargil48 said:

That is exactly the aim of the entire subject I am posting here. Why? Because of the effecti I am constantly pointing at, of the air inside the engine department having nowhere to go when the cowl flaps are closed and the associated drag that implies. I must again point to Skip's example of the tin can with a small opening at the bottom and what I have been saying about that...For me that represents a lot of drag, but... 

I think the reason cowl flaps are used over inlet air is due mainly to simplicity.  It is just more difficult to elegantly make the inlet system work well.  Keep in mind that the engine is on rubber engine mounts, so one has to decide if the variable inlet is mounted to the moving (not obvious when viewed with the cowl on) engine or the fixed cowl.  Both have very complex geometry and tight fit.  This includes complex curves, cylinders that are staggered, and the need for it to work after wear and tear and sagging engine mounts.  I suspect that, in a perfect world, the variable inlet can be more effective at drag reduction, the problem is the execution.  I’ve had a few concepts on the drawing board, but they fall apart when it comes to simple, reliable execution that the FAA would demand. (Well, the FAA doesn’t demand simple, but they do demand reliable).

[afward]

39 minutes ago, Cargil48 said:

That is exactly the aim of the entire subject I am posting here. Why? Because of the effecti I am constantly pointing at, of the air inside the engine department having nowhere to go when the cowl flaps are closed and the associated drag that implies. I must again point to Skip's example of the tin can with a small opening at the bottom and what I have been saying about that...For me that represents a lot of drag, but... 

Don't think of it from the perspective of flows.  Flows are a result of pressure gradients.  So it's better to think about it in terms of "where is the pressure" and "how will that pressure move" (yeah, I know, sounds like a flow...  stick with me).  For a cowl with a flap, you're not changing the airflow so much as changing the pressure gradient in the lower deck of the cowl (and specifically the outlet area; some cowl flaps even create a pocket of even lower pressure due to their effect on the exterior airflow).  By opening the flap, the lower deck pressure is reduced, allowing a lower upper deck pressure to sustain the same flow rate (vs. cruising with the flap closed).  It's really elegant in its simplicity.

All that said, I'd _much_ prefer a passive system...  ducted cooling plenum and a cowl with CFD-designed ramps will do everything we need without the extra drag of an oversized fixed system or an open cowl flap.

[PT20J]

Here's an interesting experiment. Take your favorite brand of canned soup and remove one end with a can opener. After a tasty lunch, take the can out to your car and drive down the freeway. Roll down the window and hold the can open end forward and horizontal out in the airflow and note the force (drag) you feel. Now, turn the can around so the closed end is forward. Is the force different? 

Skip

[PT20J]

11 minutes ago, afward said:

Don't think of it from the perspective of flows.  Flows are a result of pressure gradients.  So it's better to think about it in terms of "where is the pressure" and "how will that pressure move" (yeah, I know, sounds like a flow...  stick with me).  For a cowl with a flap, you're not changing the airflow so much as changing the pressure gradient in the lower deck of the cowl (and specifically the outlet area; some cowl flaps even create a pocket of even lower pressure due to their effect on the exterior airflow).  By opening the flap, the lower deck pressure is reduced, allowing a lower upper deck pressure to sustain the same flow rate (vs. cruising with the flap closed).  It's really elegant in its simplicity.

All that said, I'd _much_ prefer a passive system...  ducted cooling plenum and a cowl with CFD-designed ramps will do everything we need without the extra drag of an oversized fixed system or an open cowl flap.

Details at https://cafe.foundation/v2/pdf_cafe_reports/localflow2.pdf

[afward]

Just now, PT20J said:

Details at https://cafe.foundation/v2/pdf_cafe_reports/localflow2.pdf

Well, I got a permission error trying to load that.  CAFE's Mooney cowl is a thing of beauty, though.

[afward]

3 minutes ago, PT20J said:

Here's an interesting experiment. Take your favorite brand of canned soup and remove one end with a can opener. After a tasty lunch, take the can out to your car and drive down the freeway. Roll down the window and hold the can open end forward and horizontal out in the airflow and note the force (drag) you feel. Now, turn the can around so the closed end is forward. Is the force different? 

Skip

If anything, I'd say the open end will experience slightly higher force than the closed end, primarily because it will experience more turbulence and a region of reversed flow that the closed end won't have.

[carusoam]

But will you feel it?

Part two of the experiment...

Remember the lid you cut out off the can...?

Repeat the experiment using just the flat lid...

 

I think you may need a strain gauge to help out with collecting data...

 

Flow and pressure differentials go together....

If a lot of pressure was needed to get air to flow through the cooling fins....

it wouldn’t work so well if you used an Iris to adjust the nostrils of the cowl based on pressure...

 

keep in mind... the pressure drop through the cooling fins isn’t highly restrictive...

Either way... one has to consider all the pressure drops from entry into the cowl, to exit, and all the turns, and changes of shape...

And know the source of pressure is dependent on air density and speed of the plane... and in some cases the attitude of the plane....

And don’t forget the cooling capacity of air is also dependent on the air density and its temperature... and moisture content...

 

Geeesh getting a good cowl to work must have taken decades of development...  

Best regards,

-a-

[PT20J]

9 minutes ago, afward said:

If anything, I'd say the open end will experience slightly higher force than the closed end, primarily because it will experience more turbulence and a region of reversed flow that the closed end won't have.

Maybe. But, since air is incompressible at freeway speeds, after the  pressure stabilizes wouldn't the air at the opening be primarily stagnant to first order approximation?

[PT20J]

3 minutes ago, carusoam said:

But will you feel it?

Part two of the experiment...

Remember the lid you cut out off the can...?

Repeat the experiment using just the flat lid...

 

I think you may need a strain gauge to help out with collecting data...

 

Flow and pressure differentials go together....

If a lot of pressure was needed to get air to flow through the cooling fins....

it wouldn’t work so well if you used an Iris to adjust the nostrils of the cowl based on pressure...

 

keep in mind... the pressure drop through the cooling fins isn’t highly restrictive...

Either way... one has to consider all the pressure drops from entry into the cowl, to exit, and all the turns, and changes of shape...

And know the source of pressure is dependent on air density and speed of the plane... and in some cases the attitude of the plane....

And don’t forget the cooling capacity of air is also dependent on the air density and its temperature... and moisture content...

 

Geeesh getting a good cowl to work must have taken decades of development...  

Best regards,

-a-

This is why physics has two disciplines: theoretical and experimental. The theorists have to wait for the experiments that prove (or disprove) their theories. The old saw that something works in theory but not in practice is not correct. If it doesn't work in practice, the theory is wrong 

[afward]

Just now, PT20J said:

Maybe. But, since air is incompressible at freeway speeds, after the  pressure stabilizes wouldn't the air at the opening be primarily stagnant to first order approximation?

That might be beyond my skill to answer.  Yes, but real-world is messy so it's never truly stable?  Honestly I'm not sure.

[afward]

13 minutes ago, carusoam said:

Geeesh getting a good cowl to work must have taken decades of development...  

^^^ Awesome ^^^

14 minutes ago, carusoam said:

Flow and pressure differentials go together....

If a lot of pressure was needed to get air to flow through the cooling fins....

it wouldn’t work so well if you used an Iris to adjust the nostrils of the cowl based on pressure...

 

keep in mind... the pressure drop through the cooling fins isn’t highly restrictive...

As I recall, the pressure drop target across the cylinders is usually about 1/2 psi.  That's not much, but is sufficient to ensure good cooling at speed.  This is, of course, why leaks in your baffles wreck so much havoc on CHTs.

I still think a cone in a round inlet is the best design (some clever people in Burbank used that one for a, uh, "high speed" application back in the late 50's).  It just won't work as well on our planes due to prop clearance concerns.

[PT20J]

9 minutes ago, afward said:

That might be beyond my skill to answer.  Yes, but real-world is messy so it's never truly stable?  Honestly I'm not sure.

Well, I'm not 100% sure either. I think I know, but I've been wrong before. Where's a CFD guy when you need one?

[Blue on Top]

8 hours ago, PT20J said:

Ron, I don't have a good feeling for the effect of flap chord. A lot of the old NACA wind tunnel tests just compare different types of flaps, all with the same (frequently 25%) chord. Any insights on the effects of chord? Clearly Cessna favors fat ones and Mooney skinny ones.

Skip

@PT20J It's mainly area and Fowler motion.  Similar to wing aspect ratio, aileron and flap aspect ratios have some effect on efficiency (not a lot), but there are typically other factors that play a bigger role, such as: rear spar location, thickness/shape of the flap, desired stall speed, complexity of the flap mechanism/drive, etc.

There are sooooooo many factors that change aileron hinge moments and pilot roll control forces.  To start the list: aerodynamic forces, TE thickness and shape (which include centering and flutter considerations), airfoil loading (fore or aft … nothing to do with A/C CG location), system friction, aero balance (where the balance weight is located), aileron gap sealing effectiveness, aileron leading edge shape, control system gearing ratio, control system differential, control system throw (linear or exponential), yoke/stick size, a spade (aerobatic airplanes), control system springs, control system stiffness (Mooney ailerons are poor in this category), control system play, rudder/aileron interconnect, effects of wing bending, etc.

If there are enough Mooniacs that want lower roll control forces that can be looked into.  Some will result in a HUGE certification issue; some many not have any certification issues.  As I wrote that last sentence, what do Mooney pilots think of the rudder/aileron interconnect? (cost, maintenance, flying qualities, etc.).  I'm guess that it can be removed … which would make landing in crosswinds easier (lower control forces).

Here's an oxymoron for you, I'm an engineer that needs to quit talking so much.    All y'all are awesome; only 20 more posts to look at.  -Ron

[carusoam]

The fun part of the engines used in the drawings above...

1) Our Cylinders are nearly touching each other... fin tip to fin tip

2) All the blue arrows moving front to back... is air entering the cowl... sheet metal/baffling keeps air from flowing around the valve covers on the ‘top’ of each cylinder...

3) The air that is (mostly) cooling the fins uniformly for all six cylinders... is moving from the top of the cowl, through the fins, to the bottom of the cowl...  all six cylinders see the same amount of cooling air this way...

4) The NA Long Body has a couple of oddities where air enters the cowl... things like the alternator alters the air flow for the front cylinder #5... so much, so hangar fairies have design a special fix for the blockage...

5) So... it is a great challenge to get evenly distributed air spread across all six cylinders....

6) the vacuum effect of air moving across the bottom of the cowl... sucking the warm air out... I’m sure it is happening...

7) There is no special aerodynamic shaped lip to enhance the aerodynamics... just some simple geometry  to guide the flow, and leave excess space around the exhaust pipes...

8) The cowl has so many NACA ducts on it, the warm air exhaust didn’t seem to get any special duct...

9) Also consider as the volume of air turns downwards through the front cylinders then the next cylinders, and finally through the last set of cylinders....   the flow Channel at the top of the engine would need to be getting smaller and smaller to maintain the same pressure from from to back...

10) Overall the O cowling works incredibly well at cooling evenly...

11) It helps to have GAMI injectors so the heating of each cylinder is as uniform as possible...

PP observations of my plane...

Best regards,

-a-

[EricJ]

28 minutes ago, afward said:

I still think a cone in a round inlet is the best design (some clever people in Burbank used that one for a, uh, "high speed" application back in the late 50's).  It just won't work as well on our planes due to prop clearance concerns.

The movable cone in the round inlet, deployed first by the Russians on the MiG-21, is primarily to manage shockwaves at supersonic speed and slow the inlet air down behind the spike.  If it were a good design for subsonic applications I think it'd have been deployed there.  It's seldom used even for supersonic applications.

[Blue on Top]

9 hours ago, EricJ said:

Everything has a tradeoff. 

@Cargil48  And they all said, "AMEN"

We all play by the same rules, but cars are a little different.  For example they have a "reflective plane" (the aero word for "the road" below them).  Aero also plays more with inlet and exit placements with respect to local surface pressures and NACA vs ram air scoops, etc.  Ironically down force makes a car go faster (sorry, yes, I have done a little Indy car aero … with a moving floor wind tunnel).  This is why they go so fast at Indianapolis … and fatally fast on a NASCAR track.

[afward]

Yep, hence the reference to the clever people.

I more thinking of the simplicity of the arrangement; Make the cone bigger than the inlet, and contour the inlet throat to fit.  One moving part.  In theory, having the cone stick out of the inlet can help to optimize the airflow around the lip (it would help move the pressure gradient), but I might be wrong about that.

The other idea I had was NACA inlets on either side of the cowl, which would be passive but as I said earlier probably has been considered before and rejected for good reasons.

[EricJ]

2 minutes ago, Blue on Top said:

Ironically down force makes a car go faster (sorry, yes, I have done a little Indy car aero … with a moving floor wind tunnel).

Mainly if it has to turn.   If it only goes in a straight line (like for LSR, Land Speed Record events), then drag dominates so you get rid of as much of it as you can.   The downforce isn't needed in those cases. 

 

 

[McMooney]

ahh finally an answer : http://cafe.foundation/v2/pdf_cafe_reports/Wax Job with PICT.pdf

 

[Blue on Top]

39 minutes ago, carusoam said:

6) the vacuum effect of air moving across the bottom of the cowl... sucking the warm air out... I’m sure it is happening...

7) There is no special aerodynamic shaped lip to enhance the aerodynamics... just some simple geometry  to guide the flow, and leave excess space around the exhaust pipes...

8) The cowl has so many NACA ducts on it, the warm air exhaust didn’t seem to get any special duct...

-a-

6) Yes, yes, yes!  Inlets in high pressure areas; exits at low pressure areas.

7) Ummmmmm. Yes, round is good with engine cooling inlets/exits (lowest drag shape due to boundary layer and less interference drag).  All the new cowlings have round holes up front

8) There need to be less NACA ducts (on the O cowl especially).  And a reverse one on the exit  

- MS needs to consider overly passionate engineers.  I'm out of emotion … again  

Bullet point thoughts for clarification (hopefully)

• Cowling inlets and exits are designed for worst case (already noted)
• Cars and airplanes differ greatly in the required aero to cool the engine.  Cars (water cooled) need A LOT of radiator surface area/volume and very slow airflow because the temperature delta is low ( ~100F worst case).  Airplanes (air cooled) have minimal surface area (cylinder surface areas), a faster air flow is okayish but the temperature delta is high (350F or more).
•  Mass flow into the cowling = mass flow out of the cowling (none gets trapped in there).  There is internal duct flow drag for every molecule that enters.
• Since EVERY condition (other than worst case) supplies too much cooling air through the inlets, the inlets should be designed to dump air efficiently.
• As I think someone mentioned a while back, closing cowl flaps all the way (and tight with the fuselage) will increase drag.  Opening them some will decrease drag slightly ... put a bigger hole in the bottom of your can. 
• There is a pressure box (some have dog houses) on the top of your engine for a good reason.  It increases pressure (up to ~ dynamic/airspeed pressure) and slows the air flow down for better cooling (better heat transfer from the cylinder fins to the cooling air).

Dang this is cool … ummmmmmm … awesome!  -Ron

[PT20J]

44 minutes ago, Blue on Top said:

Yes, round is good with engine cooling inlets/exits (lowest drag shape due to boundary layer and less interference drag).  All the new cowlings have round holes up front

The placement is important, also. Lopresti's designs placed the inlets as far outboard as possible to capture "ram" air from the prop. The blade shanks nearer the hub don't produce much thrust.

46 minutes ago, Blue on Top said:

As I think someone mentioned a while back, closing cowl flaps all the way (and tight with the fuselage) will increase drag.  Opening them some will decrease drag slightly ... put a bigger hole in the bottom of your can. 

On the early M20J, the left cowl flap was flat and closed off most of the airflow on that side when closed. Later M20Js have a rounded cowl flap similar to the right one that lets some air out even when closed. 

53 minutes ago, Blue on Top said:

Cars and airplanes differ greatly in the required aero to cool the engine.  Cars (water cooled) need A LOT of radiator surface area/volume and very slow airflow because the temperature delta is low ( ~100F worst case).  Airplanes (air cooled) have minimal surface area (cylinder surface areas), a faster air flow is okayish but the temperature delta is high (350F or more).

The Reno race P-51s put out more heat than the radiator can handle and they solve the problem by spraying water on the radiator to cool it. You can often see the vapor trail behind the airplanes. Another Kerchenfaut innovation was to realize that it wasn't the water that cooled but the evaporation of the water. He designed a system using fine spray nozzles in the inlet duct of Strega well ahead of the radiator so that the water would evaporate and cool the air before it got to the radiator. The result was more efficient cooling and a reduction in weight from carrying much less water. By the way, the Meridith effect is real, but it doesn't turn the cooling system into a little jet engine producing net thrust like some people think. But enough thrust is produced to reduce the cooling drag from 6%-10% of the total drag down to about 3% according according to David Lednicer's calculations. (https://arc.aiaa.org/doi/pdf/10.2514/6.1991-3288)

Skip

[carusoam]

For anyone reading this far, and not remember the Meredith effect...

https://en.wikipedia.org/wiki/Meredith_effect

-a-

[ArtVandelay]

I thought the change to the left cowl flap was because of going to dual tailpipes in other models and for manufacturing simplicity they made it for Js as well?

There is a lot of non aerodynamic shapes inside the cowling, I wonder how much of the drag could be improved?

I wonder which engine has more drag, L or C, seems having the large intake tubes on the bottom of the engine would make the Lycoming more streamlined.

 

 

Tom

 

[takair]

5 hours ago, PT20J said:

On the early M20J, the left cowl flap was flat and closed off most of the airflow on that side when closed. Later M20Js have a rounded cowl flap similar to the right one that lets some air out even when closed. 

Did the J not have the cowl flap cove on both sides like the E and F?  My E has crude sheet metal cowl flaps, but fairly well executed recessed coves such that when the cowl flaps are “closed” there is still a calibrated amount of air flowing through.  I think this alone is quite effective in reducing drag on a Mooney vs other designs when the cowl flaps are closed.

[Cargil48]

10 hours ago, PT20J said:

Here's an interesting experiment. Take your favorite brand of canned soup and remove one end with a can opener. After a tasty lunch, take the can out to your car and drive down the freeway. Roll down the window and hold the can open end forward and horizontal out in the airflow and note the force (drag) you feel. Now, turn the can around so the closed end is forward. Is the force different? 

Skip

I don't know because in the second example there is no build-up of air inside the can, it flows directly away, but you miss the point. Do I have to explain it again??? I don't want to suggest closing the inlets of air, I am all the time suggesting a device which directs the incoming flow to where it is needed, or down, directly to the air exit when not. 

The PT6 turbine has also an inverted flow system. It sucks the air near the firewall and expells it through the exhaust ducts placed right behind the proppeller. This makes it mandatory to install a proper tubing to direct the fresh air to where it is needed.  The Porsche equipped Mooney had the same, for cooling. Air got in through NACA ducts on the front side parts of the cowling and was directed to the ventilation turbine located aft. Just like in the PT6 but for a different purpose. 

 

Edited December 11, 2019 by Cargil48
editing typo...