Need help to understand airflow

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

REVAN

Well-Known Member
Joined
Dec 6, 2016
Messages
279
Location
Tucson, Arizona USA
I need you guys to explain something to me. On one hand for any object moving through the liquid ,shape is very important to reduce drag. we know that shape of the front part of this object is equally important as shape of the back side.

On other hand I am hearing this explanation that airflow on the back side of abruptly shaped object will form turbulent Triangle that will work like a bearing and this void will be naturally filled. So, we should stop worrying about building anything for this area?

I believe Peter Garrison at one point was telling me that all this fancy looking Spinners that you guys using on canards complete nonsense.

I also see how they shape fuselage on icon A5 right before propeller. They must have the same logic as far as airflow concern.

when we standing on the bridge and looking down we can see how water is moving around concrete columns. You can clearly see still water forming that kind of area that’s really not moving at all and acting like filler.

so, do I see it correctly? In certain area like prop spinner for pusher or backside of fuselage after body doesn’t need to be perfect?

View attachment 111518
The long pointed tail only matters if the flow remains attached to the contracting body. If you can maintain that flow attachment, the pointy tail makes a big difference. If the flow separates, there is a big drag increase that will be proportional to the size of the separated region. When this happens, it makes little difference if you cut off and remove any part of the body that is within this separated flow region.

Cars have a rear window and a surface bump at the roof to window transition that tends to separate the flow. Designers found that if the rear window transition did not drop too fast, the flow could remain attached over the transition bump and recover more of the car body's form drag than a traditional rear window that drops down to trunk level. This is where the Honda CRX got it's look. Even today, cars that focus on aerodynamics tend to look similar to compact station wagons with a mildly sloping rear roofline that cuts off abruptly. The goal is to contract the flow as much as possible before it separates from the body. At some point size constraints of the car dictate that the car can't be any longer than 'X', so you terminate it with a flat end and let the flow separate (i.e. - cut your losses, literally).

I hope this explanation helps.
 

Jay Kempf

Curmudgeon in Training (CIT)
Lifetime Supporter
Joined
Apr 13, 2009
Messages
4,314
Location
Warren, VT USA
Cutting off the back end if you are going to use a car analogy means you have to look at the details. NASCAR has done tons of research and so have IMSA and GTP ALM participants. Truth is the chopped back end has to be done right. NASCAR figured out early that you need to add a proper trip strip or spoiler at the right angle with the edge in just the right place to get that nice tapered turbulent -stagnant- zone to form. If you don't you just create a lot of suction as the flow tries to slam shut behind the tail light plane. For it to work it assumes that one side (the road) isn't moving which has a very stabilizing effect on the overall flow. If you don't form the following eddy current correctly it will be overly turbulent and not stagnant and so will create huge drag and turbulence that can affect handling negatively.

So if you are trying to analyze a blunt rear end but you mean leave off a spinner you need to model the flow properly to understand what to do. Many pushers with bad engine installations don't allow the flow to stay attached after the fattest part of the fuselage. So they install VG's to trip the flow to reduce the thickness of the turbulent boundary layer. If there is not prop flow then there is no more energy introduced to pull on the turbulent parts ahead of the prop so simpler but probably even more turbulent after the blunt separation. Props are turbulent but also motive so it is a different analysis based on how efficiently you accelerate the tube of air.

So I guess you need to clarify first what you are doing then apply the best overall guess. On a chubby pusher there isn't any laminar flow past the fattest part of the fuselage or wing normally so that drives you into VG's to get the best you can out of a bad situation. Fuselage shape and camber line are pretty important as well.
 

WonderousMountain

Well-Known Member
Joined
Apr 10, 2010
Messages
2,294
Location
Clatsop, Or
Birds have porosity, slot gaps & Aero-Elasticity in addition to morphing form.

An airfoil is just not the same. Laminar flow is excellent for keeping drag down, but most wing surfaces have both transition flow and a very draggy portion near the trailing edge. A peculiarity of wind tunnel testing is even though most real craft have flaps, unflapped section description dominates.

Results as high as 1% Chord blunt trailing edge has improved drag numbers and increased lift performance in the stall-stag region AoA. Interprating lab results can be difficult, flight trial experiments sometimes see improvement that is really quite small, but thought provoking. Testimony seemed to be in favor of 1/8th inch and under blunting, for GA craft. Measure is preferred to fractional percentage in nearly all experimentors. 1/4 inch seems to be the limit of positive GA testing, and IIrc that has been on longer chord wings.

Pintle type spinners show lowest drag (pusher);
These are the ones with concave revolved body. You will need to direct axial flow tangent to the
surface, not linearly in the direction of flight. Not every propellor manufacturer will be so attentive.

In my humble opinion, a 1/12th body blunting, To keep from poking your eye out is reasonable. The Base area is the figure of Merit, Diameter
fraction ^2, so that 1/10 becomes 1/100 known as common 1%. Or you could not bother with it.

~CK LuPii
cd8d44cac535e9fde8be34174ea19d348b331658.jpg
 

Eugene

Well-Known Member
Joined
May 26, 2017
Messages
1,566
Location
Merrill, Wisconsin, USA
The long pointed tail only matters if the flow remains attached to the contracting body. If you can maintain that flow attachment, the pointy tail makes a big difference. If the flow separates, there is a big drag increase that will be proportional to the size of the separated region. When this happens, it makes little difference if you cut off and remove any part of the body that is within this separated flow region.

Cars have a rear window and a surface bump at the roof to window transition that tends to separate the flow. Designers found that if the rear window transition did not drop too fast, the flow could remain attached over the transition bump and recover more of the car body's form drag than a traditional rear window that drops down to trunk level. This is where the Honda CRX got it's look. Even today, cars that focus on aerodynamics tend to look similar to compact station wagons with a mildly sloping rear roofline that cuts off abruptly. The goal is to contract the flow as much as possible before it separates from the body. At some point size constraints of the car dictate that the car can't be any longer than 'X', so you terminate it with a flat end and let the flow separate (i.e. - cut your losses, literally).

I hope this explanation helps.
Yes, very helpful and in plain English! If you are already in separation zone, then doesn’t even pay to worry about to get to correct shape.

I was talking to someone from Russia who apparently have some time playing around with wind tunnel. He was telling me that they usually see separation of you increasing 8°. Wondering if anybody else has similar experience with flow separation.

Back part of my fuselage in front of propeller will shrink much more abruptly than 8°. I wonder what would be better in this case? Maybe I should stay with 8° and have 6 inch wide cut out in front of Propeller.

Small cross section below is line right below the wing.

534125E4-4757-4FF9-BFD1-7FDF1D446A10.jpeg

and this large cross section is at propeller tip line

BFD04E92-7DE7-4C6B-832F-079D9147A331.jpeg
 

Eugene

Well-Known Member
Joined
May 26, 2017
Messages
1,566
Location
Merrill, Wisconsin, USA
On a chubby pusher there isn't any laminar flow past the fattest part of the fuselage or wing normally so that drives you into VG's to get the best you can out of a bad situation. Fuselage shape and camber line are pretty important as well.
I divided my problem into two parts. One part is in front of the prop. This part I am very limited as far as what to do. Another part is below propeller and I can create pretty nice shape there. I think176D552D-BA6C-474F-ABCA-EB64DAC9B5A0.jpeg
 

REVAN

Well-Known Member
Joined
Dec 6, 2016
Messages
279
Location
Tucson, Arizona USA
Yes, very helpful and in plain English! If you are already in separation zone, then doesn’t even pay to worry about to get to correct shape.

I was talking to someone from Russia who apparently have some time playing around with wind tunnel. He was telling me that they usually see separation of you increasing 8°. Wondering if anybody else has similar experience with flow separation.

Back part of my fuselage in front of propeller will shrink much more abruptly than 8°. I wonder what would be better in this case? Maybe I should stay with 8° and have 6 inch wide cut out in front of Propeller.

Small cross section below is line right below the wing.

View attachment 111589

and this large cross section is at propeller tip line

View attachment 111590
The propeller will influence the flow and tend to keep it attached to the contraction in front of it when the prop is under power. Also, tunnel data may be at low Reynolds numbers compared to a full sized aircraft (something you need to look at when evaluating the applicability of data). When looking at flow separation, Reynolds number is a big influence. What doesn't work in a tunnel, may work in real life at a higher Reynolds number than what was created in the tunnel. In the end, you'll need to make your best choices based on limited information, then build and test it to find out if your hunches were about right, overly optimistic, or if you left potential capability untapped.
 

Eugene

Well-Known Member
Joined
May 26, 2017
Messages
1,566
Location
Merrill, Wisconsin, USA
The propeller will influence the flow and tend to keep it attached to the contraction in front of it when the prop is under power.
Well, in my situation surprisingly some tufts were pointing straight up right in front of propeller. Not sure how to explain it and what that means exactly. Almost like prop is starving for air and sucking it from down low?

Screen Shot 2021-06-11 at 21.57.59.png
Screen Shot 2021-06-11 at 22.10.28.png
 
Last edited:

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
15,074
Location
Port Townsend WA
The prop is rotating the air in a corkscrew. So it pulls air upward on the right side of the fuselage some distance ahead of the prop. It also swirls more to the center, the wake is smaller than the prop diameter.
 

Eugene

Well-Known Member
Joined
May 26, 2017
Messages
1,566
Location
Merrill, Wisconsin, USA
The prop is rotating the air in a corkscrew. So it pulls air upward on the right side of the fuselage some distance ahead of the prop. It also swirls more to the center, the wake is smaller than the prop diameter.
Of course I don't know what I am looking at, but tufts seems to behave the same way during takeoff, landing or at idling power. My point is, that prop RPM doesn't seem to fix airflow at all. Not on this machine

Screen Shot 2021-06-11 at 22.17.14.png
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
8,049
Location
Saline Michigan
The three most important things in powered airplane performance are:
  1. Keep weight near minimum for the mission;
  2. Let the wings work like they are the only thing in the air;
  3. Let the prop work like it is the only thing in the air.
The Skyboy is an LSA and does pretty good on 1.
The sharply tapered fuselage through the wing and uncowled engine violate the daylights out of 2 and 3.

A good cowling will help bunches with 2 and 3. Decent fillet radii at the wing root may help there too. VG's on the sharply tapered aft fuselage is the usual solution on improving the feed of air into the prop.

If all that you were looking at was drag from the fuselage, and you had bunch of separation coming off the tapered fuselage, it might indeed be lower drag to reduce taper and accept a somewhat wider cutoff. In the Skyboy, the fuselage drag and the prop efficiency are highly coupled... You can not handle them as separate issues because almost all of air coming off the fuselage ahead of the prop is going through the prop. You really need to minimize variation in air velocity and direction as it goes into the prop. You will get more thrust per hp that way.

If I were determined to improve the Skyboy, I would have a minimum width at the cutoff end, and play with VG's to attach and straighten flow into the prop as much as I can. I would also drop my cabin air and engine cooling air through outlets sized to be bigger than the inlets based upon changes in absolute temperature of the air going through - the percentage change will be much bigger for engine air than for cabin air. In this way, the air coming out is about at the same velocity as it was coming in, and thus close to airflow velocity around the fuselage and engine.

Billski
 

drgondog

Active Member
Joined
Sep 24, 2015
Messages
32
Location
Scurry, TX/USA
ALL of the airflow shown is turbulent. There is no laminar flow for RN> ~ 500K and most full scale a/c are in the 2x10^6 and greater in flight. What the tufts show is whether it is immersed in free stream or in an 'eddy' created during full blown boundary layer separation.

BL behavior may be somewhat addressed by Chaos Theory (for those that either took the math - or watched Jurassic Park) but there is zero constant (and predictable) separation 'line' in 3-D flow.
 

WonderousMountain

Well-Known Member
Joined
Apr 10, 2010
Messages
2,294
Location
Clatsop, Or
Those are beautiful lines, I am sorry you are running into obsticles.

You will see thumbed two sketches, of no artisnal virtuousity. First,
What I suspect you have is a paired vortex, powered by detached flow
Most probably starting at the lower wider region above where the gear
takes leave of the cabin body. It follows half rolling in a crescent shape
Meeting at the trailing edge heading notably upward. Since both sides
are of symmetric design, shear does not provide resistance to updraft.
The prop will aid this some, but even off, it is likely to be the same event.
IMG_20210613_125705.jpg
Second sketch offers a 'Quick' solution. A thin fin to straighten flow back
toward your rapidly converging fuse. There is no very good place for this,
But I believe the right wing will do what is asked of it. A mount could be
made to try different planeforms & airfoil sections. Good luck with it.
IMG_20210613_125740.jpg
~CK LuPii
 

Eugene

Well-Known Member
Joined
May 26, 2017
Messages
1,566
Location
Merrill, Wisconsin, USA
Those are beautiful lines, I am sorry you are running into obsticles.

You will see thumbed two sketches, of no artisnal virtuousity. First,
What I suspect you have is a paired vortex, powered by detached flow
Most probably starting at the lower wider region above where the gear
takes leave of the cabin body. It follows half rolling in a crescent shape
Meeting at the trailing edge heading notably upward. Since both sides
are of symmetric design, shear does not provide resistance to updraft.
The prop will aid this some, but even off, it is likely to be the same event.
View attachment 111838
Second sketch offers a 'Quick' solution. A thin fin to straighten flow back
toward your rapidly converging fuse. There is no very good place for this,
But I believe the right wing will do what is asked of it. A mount could be
made to try different planeforms & airfoil sections. Good luck with it.
View attachment 111839
~CK LuPii



I am convinced after my flying experiments with the form that airflow can be improved quite a bit with correct shape of fuselage below the prop. In front of the prop VGs making big difference as well.

Screen Shot 2021-06-13 at 15.33.23.pngScreen Shot 2021-06-13 at 15.29.56.pngScreen Shot 2021-06-13 at 15.28.28.png
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
6,505
Location
US
Eugene,
Have you applied tufts to the gear leg fairings and wheel pants? The way the aft surfaces are shaped might be less effective than a concave "pressure recovery" shape or even just a longer, more gradual taper.
Not that you need another project.
Apologies if this has already been addressed in another Flyboy thread.
 

WonderousMountain

Well-Known Member
Joined
Apr 10, 2010
Messages
2,294
Location
Clatsop, Or
Not sure if you are trying for very low drag, but those tufts are all over the place in video 2-3, you may need to generate a larger vortex, or add cosiderably to the body closeout. I did not see the issue coming, but it is in plain view now.
 
Top