Aerodynamics and streamlining computations

vlittle · #1 01-21-2016, 09:45 AM

So, if I have a cylinder of length L and diameter D sticking out into an airstream at, say 210 Knots, what is the most effective way to streamline it?

A streamlined bubble (half teardrop) is convenient, but it also has a lot of skin drag. What is the break-even point? At what size does the blister have the same drag as the cylinder?

A streamlined mast (airfoil cross section) of length L (think pitot mast) also has a drag coefficient. Same question as above.

We all hate the little things that stick out.... Fuel drains, vents and so on. Is it better to leave them alone or to fabricate streamlined fairings for them? I know what the gut feeling is, but is it supported by science?

skylor · #2 01-21-2016, 11:46 AM

Quote:

Originally Posted by vlittle

So, if I have a cylinder of length L and diameter D sticking out into an airstream at, say 210 Knots, what is the most effective way to streamline it?

A streamlined bubble (half teardrop) is convenient, but it also has a lot of skin drag. What is the break-even point? At what size does the blister have the same drag as the cylinder?

A streamlined mast (airfoil cross section) of length L (think pitot mast) also has a drag coefficient. Same question as above.

We all hate the little things that stick out.... Fuel drains, vents and so on. Is it better to leave them alone or to fabricate streamlined fairings for them? I know what the gut feeling is, but is it supported by science?

You may find this video on the science of drag reduction (and other related videos) very informative:

https://www.youtube.com/watch?v=ftq8jTQ8ANE

Skylor

Flying again! · #3 01-21-2016, 01:01 PM

I suggest going to this link and download ?Fluid Dynamic Drag? by Hoerner, it is a great book and a classic for aero engineers like me.

https://docs.google.com/file/d/0Bx0M...t?pref=2&pli=1

For a cylinder perpendicular to the local flow the CD is 1.2 (assuming Reynolds number of 1x105 (210 mph at 15C, 3500ft and pipe diameter of .5?). If you streamline it in a teardrop without changing the diameter, the CD drops to .06. The base drag is the big contributor due to turbulent airflow and separation, skin friction is almost insignificant.

A properly shaped blister has a CD of .036 but the reference area is higher.

So here are some examples (all at 210 mph, 3500ft, std atmosphere):
1. A .5? diameter cylinder 2 inches long has an un-streamlined drag of .77 lbf, streamlined it is .038 lbf. A blister would be about .092 lbf.
2. A 2? dia cylinder 48? long (gear leg). Un-streamlined drag is 73.6lbf. Streamlined is 3.68 lbf, 20 times more w/o streamlining. Explains why you pick up 10 knots or so with the gear legs streamlined.

There obviously are a lot of other factors such as surface roughness, critical Reynolds number, intersection drag etc. that can be accounted for on a case by case basis.

The book is heavy in theory but has some clear examples. Since it was written in the 1951 (1st edition), there are some great examples that take you back in history, such as adding a blister fuel tank to an Me-109 drops the speed from 320 to 318 mph.

It covers, scoops, radiator drag, struts, vents, sea anchor drag, drag of cars, nacelles, aircraft floats, landing gear, tires, external fuel tanks, bombs, parachutes, airships, antennas, Mercury capsule - virtually everything you can think of!

Take a look and see if it helps you understand how to reduce drag on your plane.

TJ

tjo · #4 01-21-2016, 01:18 PM

Those are a couple of great answers. In summary, the easy answer is, it will always be better with a fairing or aero shape because the aero shape will have roughly 5% of the form drag of the cylinder and skin drag will never make up for this difference.

That said, on a practical side, if you had five 1/2 inch diameter 1 inch rods sticking out of your plane, at cruise you would lose about 1mph. If you just used more power and kept the speed the same, you would burn about 0.1gph more, or roughly 45 cents per hour cost. Difficult to justify making little fairings for fuel drains and such, but if you are after every last bit of speed or efficiency, definitely the way to go.

Tim

scsmith · #5 01-21-2016, 01:26 PM

the old conventional wisdom from airship design days was that a fineness ratio of 3:1 was the best tradeoff between skin friction and pressure drag.
But that was for body of revolution, not a cylinder. Also, it was for constant internal volume, which is different from constant diameter.

But it gives you some idea. For a cylinder, the optimum would be more fineness than that (more chord for a given thickness).

My dear old friend and mentor, R T Jones, used to have a plaque on the wall in his office. It had a 1" stub of 1/8" rod sticking out, and it had a equal-length stump of an NACA 0010 airfoil, with a 1" thickness and 10" chord. The claim was that the two had the same drag.

DanH · #6 01-21-2016, 01:55 PM

....as does Vern, I'm sure.

Raymo · #7 01-21-2016, 02:13 PM

Quote:

Originally Posted by vlittle

We all hate the little things that stick out.... Fuel drains, vents and so on. Is it better to leave them alone or to fabricate streamlined fairings for them? I

Or just buy them over here at JDair.com

dmn056 · #8 01-21-2016, 05:31 PM

The numbers shown are based on 2-D sections. These are a very good approximation for things like landing gear legs, but not so good for short protrusions like temperature probes or tank vents.

See Hoerner again for examples.

Also, streamlining of low aspect ratio protrusions is much less effective when not closely aligned with the local flow. This means very careful measurement of the local flow direction at cruise (or whatever condition you want to optimise for min drag) before fairing installation. Since the absolute drag effects are mostly small, it may not be worth the effort.

However, design to eliminate the protrusion will always give 100% improvement at any flight condition ...

vlittle · #9 01-21-2016, 06:52 PM

The video was quite good. An airfoil shape is about 1/10 the drag of a cylindrical shape. Further reading says that a 1/2 airfoil (blister) is about 1/5 the drag (not exact, just rules of thumb).

This means that if the x-sectional area of the blister is 5x the cylinder, it's the same drag.

Therefore an airfoil shape (like a pitot mast) over a cylinder is better than a blister over a cylinder because the area is smaller and the Cd is lower.

So the JDAIR design is much better than a blister over a protuberance. Better yet would be a small afterbody fairing on the cylinder(aka fuel vent or temp probe).

QED.

scsmith · #10 01-21-2016, 08:00 PM

Quote:

Originally Posted by vlittle

The video was quite good. An airfoil shape is about 1/10 the drag of a cylindrical shape. Further reading says that a 1/2 airfoil (blister) is about 1/5 the drag (not exact, just rules of thumb).
QED.

It depends on the airfoil. For a good laminar flow section, its more like 1/30.