Airfoil selection for an aircraft out there?

[WARPilot]

I think 1200 is where they usually end up. Lots end up going overboard with finishing and glassing.

I know of the P-51 in Asheville. It weighed 1200# at T.O. The owner said it was fast. The weight was way out of line for the design.

 

[Vigilant1]

It's a long way to 1.8 with any kind of simple flapped wing.

But, if he builds it, he will find out.

Yes, I assumed flaps. Many planes achieve an overall flapped wing CL of something close to 1.8 (or more). For example, with plain flaps deployed, per the CAFE Foundation tests (calibrated stall speed, actual weight)
Thorp T-18: CL 2.0
RV-6 : CL 2.1
Now, it is also easy to find planes with lower CL at stall in landing configuration:
Bushby Mustang: CL 1.4

Reynolds number: The Thorp's wing chord is within 3" (7%) of the proposed WAR-FW wing, it seems unlikely that would cause a major difference in expected CL at VSO. The aspect ratio of the Thorp is 5.1, the AR of the proposed WAR FW wing is 5.2.

There's certainly nothing fancy or sophisticated about the T-18 wing or flaps, but they do perform well. I don't know why a well designed and well built wing on a new WAR plane couldn't also give a CL of about 1.8 with plain flaps. But, it is nothing to take for granted, either. As you say we won't know until it is built. The T-18 wing got considerably better behaved after some tweaks.

 

[rv7charlie]

A/c designers often say that a good starting point is to 'sanity check' against an existing successful design. Instead of the RV6, how about the RV3? 1100 lbs gross weight, 51 mph stall speed. Wing area is 90 sq ft & span is just under 20 feet.

BTW, the reason the Mustang looks worse than the others is likely because the flap is really more of a belly board; it goes all the way across the underside of the fuselage.

Thorps are great flying a/c, but they might not be the best reference point if you're looking for decently docile performance (and a wider audience).

Charlie

 

[Vigilant1]

A/c designers often say that a good starting point is to 'sanity check' against an existing successful design. Instead of the RV6, how about the RV3? 1100 lbs gross weight, 51 mph stall speed. Wing area is 90 sq ft & span is just under 20 feet.

RV-3 by those numbers: A stall CL of 1.87.
I avoided manufacturer's numbers because, well, not everybody trusts them and there's usually not much background on how they were collected. But, the point remains the same: For a wing of about this size and with plain flaps, a VSO coefficient of lift of 1.8 or better is achievable in the real world.

Charlie, did your Thorp have the original airfoil or did it have the later Lu Sunderland airfoil?

Mark

 

[rv7charlie]

As far as I know, mine was bone-stock (purchased from the widow of the builder). I thought it was a great flying a/c, but I was likely well prepared for it, having learned to fly in a Luscombe 8A. Low time pilots who learn to fly in the later Cessna/Piper/etc trainers that are typical today will not be prepared to handle a Thorp. Stories I heard & read back in the 1990s indicated similar, but worse, issues with the WAR replicas. I have no experience with them, but 'word on the street' back then was that scaling down as much as they did caused significant issues with stall speed and control sensitivity.

EDIT: What I should have written was 'significant issues with low speed handling qualities and' etc. Stall speed often doesn't tell the whole story about low speed handling, especially with 'short wing' designs like the Thorp (and even RVs in some situations).

 

[Pops]

FLying my F-12 you had to use 80 mph at light weight with just me for the approach and anything under that it went into a high sink rate. Heavy with 2 people, had to use 85 mph to stay out of the high sink rate. Not mph less.

 

[Vigilant1]

EDIT: What I should have written was 'significant issues with low speed handling qualities and' etc. Stall speed often doesn't tell the whole story about low speed handling, especially with 'short wing' designs like the Thorp (and even RVs in some situations).

With a rectangular wing (T-18, RV, Sonex, etc) there's a very strong tendency for the wing will stall from the root outward, which helps a lot as far as low speed handling. The ailerons on my Sonex work fine well into the "mush" (though, obviously, abusing that is a good way to get some spin practice).
A tapered wing (like the FW will have) requires more work (usually incorporation of washout) to achieve the same thing. It's obviously doable.
The revised LE on the Sunderland wing was reportedly a significant improvement WRT the reduced suddenness of the stall break. This goes to my previous point: Airfoil attributes can be subtle, and using the exact airfoil that has worked in other similar airplanes can save a lot of trouble.

 

[3V152]

So, who is flying the S8036 and how does it perform in the real world?

Not sure of any fullscale aircraft that would use S8036. It was designed for powered model aircraft.
bigfoil.com (shameful plug) does have S8036 wind tunnel data, but it is all at low Reynolds numbers, appropriate to model aircraft. The wind tunnel data is from UIUC Low-Speed Airfoil Data, Vol 3.

 

[BigL]

Going back to a post on page 1.

If the max speed of the original WAR FW-180 was 165 mph then it is reasonable to expect 180 mph from an engine with 30% more power.

If you follow Bilski's advise you will move towards the Riblett airfoils.

Everyone please note that for a comparison between the Eppler program Riblett was using and the NACA wind tunnel, Riblett provided PAGE 37 in his book.

Bearing the above in mind. And using an aircraft design program by Donald R Crawford with a few additions to 'up' the aircraft weight to compensate for Horizontal Tail down-force, I present this to you.

At 180 mph your aircraft would be flying at a Reynolds number of about 6.5-Million, the Coefficient of Lift will be 0.17. That points to a 2-series in the Riblett or NACA. Let's go with the Riblett GA 37A-215. First, it is designed for the Coefficient of 0.17, Second, it is a Laminar Flow Airfoil which has a lower Drag Coefficient than the 230XX. Third, its constant 15% thickness to the tips is better strength-wise and aerodynamically, and Fourth, Riblett seemed to have a sweet spot for this airfoil (see Pg 106 where he gives Flap and Aileron details).

Stall with the above at a Reynolds number of 2-million and a Coefficient of Lift of 1.38 will be 58 mph. Flaps will help.

Comparing the NACA 230XX to the Riblett GA 37A-215 the following will be of interest to you:-
NACA 230XX _ Min Drag = 0.0062 _ Max Lift (Reynolds Number 2-Million) = 1.3
Riblett ________ Min Drag = 0.0048 _ Max Lift (Reynolds Number 2-Million) = 1.41

The Riblett has notably less drag at zero-degrees AOA. That should make 180 mph easier to achieve, assuming you build for Laminar flow. The reduced drag will more than offset the extra Wing Moment Arm of the Riblett (which is very reasonable compared to some other popular sections).

Harry Riblett points out that the 230XX has a sharper and immediate stall, while his improved airfoils are 'soft stall', and trying to go beyond the Angle of Attack that produces Cl of 1.41 is not a problem. The 230XX is good for the Vans aerobatic aircraft but you may wish to land slower.

Hope that answers your search for a better airfoil.

Oh, there are other airfoils that may promise much - but computer generated data is not real flight Data. Riblett's airfoils, and his solutions for 'problem-airfoils' have been proven in real life.

Disclaimer: I am not being paid by any company or persons designing airfoils or aircraft performance programs.

 

[speedracer]

A/c designers often say that a good starting point is to 'sanity check' against an existing successful design. Instead of the RV6, how about the RV3? 1100 lbs gross weight, 51 mph stall speed. Wing area is 90 sq ft & span is just under 20 feet.

BTW, the reason the Mustang looks worse than the others is likely because the flap is really more of a belly board; it goes all the way across the underside of the fuselage.

Thorps are great flying a/c, but they might not be the best reference point if you're looking for decently docile performance (and a wider audience).

Charlie

A/c designers often say that a good starting point is to 'sanity check' against an existing successful design. Instead of the RV6, how about the RV3? 1100 lbs gross weight, 51 mph stall speed. Wing area is 90 sq ft & span is just under 20 feet.

BTW, the reason the Mustang looks worse than the others is likely because the flap is really more of a belly board; it goes all the way across the underside of the fuselage.

Thorps are great flying a/c, but they might not be the best reference point if you're looking for decently docile performance (and a wider audience).

Charlie

A/c designers often say that a good starting point is to 'sanity check' against an existing successful design. Instead of the RV6, how about the RV3? 1100 lbs gross weight, 51 mph stall speed. Wing area is 90 sq ft & span is just under 20 feet.

BTW, the reason the Mustang looks worse than the others is likely because the flap is really more of a belly board; it goes all the way across the underside of the fuselage.

Thorps are great flying a/c, but they might not be the best reference point if you're looking for decently docile performance (and a wider audience).

Charlie

RV 3's are FAST. 21 years ago an RV 3 beat me out of first place in the Copperstate Dash air race flying my 290 angle valve powered Long EZ. IIRC my speed was 220 MPH. I'm still kinda pissed about that.

 

[Marc W]

I think that's the answer. Make it the size of an RV-3 and you have a winner!

 

[Lendo]

Lets see the right Riblet Airfoil about Cl -1.6 at Landing speed, Plus good split flaps 1.2 =2.8 *.73 (0.73 is 0.93*Pi/4) =CL of 2 , but to get there one would need
Wing Area [S= Weight/ CL*0.5*P*V^2]. I'm sure you can work it out from here.

It's best to do a Excel Spread sheet and play with the numbers.

Make sure the flaps are Big 35% of chord and 70% of exposed Wing Semi Span, with 30% of Flaps. The exposed semi span is length less the tip and any concave cuff (fillet) between Wing and Fuselage.
Have fun, George

 

[Lendo]

Oops, don't forget to add the HT download to the weight (1+.67*St/Sw).
George

 

[PiperCruisin]

Make sure the flaps are Big 35% of chord and 70% of exposed Wing Semi Span, with 30% of Flaps.

Was never sure how to do this, but here it goes. Assuming GA37A215 0.35c flaps at 35 degrees and 70% of span. The flaps have a Clmax of ~2.5 an the outboard wing would be at a minimal angle of attack so a Cl of about 0.1. So 70%*2.5+30%*.1 =~1.78. Maybe wack off 5-10% for inefficiencies.

 

[WonderousMountain]

Just do Cri-Cri wing.

 

[BigL]

Realistic Maximum Coefficient of Lift (Cl) for a General Aviation aircraft occurs at a reduced airspeed and therefore a lower Reynolds Number (RN) like 2-Million. An airfoil with a Cl of 1.6 operating at an RN of 6-Million changes its behavior when slowed to a stall speed where the RN might be 2-Million, its Cl also reduces and becomes 1.45 according to the Cl-optimistic airfoil program. Correction is made for the real world and becomes 1.41. Another behavior change is that when approaching the stall, the Center of Pressure (CP) Moment about the Center of Gravity (CG) moves towards the quarter chord point so Tail Down-Force becomes a minor issue - if at all, and if the Center of Gravity (CG) were very aft, it could become an 'up-force'. So no simple formula can be applied. Horizontal Tail (HT) Down-Force is a moving target. It has to be worked out for each CG location and airspeed Cl.

At the stall, the wing maintains some Angle of Attack (AoA) even if its flaps were full span and set to a standard NACA setting of 60-degrees. See Theory of wing sections where the unflapped NACA 63-215 stalls at 14 degrees (RN = 3-million), but when deploying 60-degrees of flap the AoA reduces to an estimated 10 degrees (NACA Flap data doesn't cover RN = 3-Million). So the combined Maximum Coefficient of lift for a conventional Flap and Aileron layout also becomes a moving target. In this case, were we to use a 50/50 split, the AoA has to be 10 degrees - as dictated by the stall angle of the deployed Flap. Working it out - the flapped portion operates at approximately 2.4 and the outboard is at 1.2. The math indicates the Cl of that wing would be 1.8 at an AoA of 10 degrees. Please note that by being able to reduce the stall speed using flaps, the RN has reduced so all the figures for Cl have reduced. Eventually the calculation for Cl lies somewhere between 1.975 and 1.35. An educated guess would arrive at about 1.7 at an AoA of 9 degrees for that particular NACA data set. Nowhere near as good as we would want, but realistic.

Plugging that into the original figures using the Riblett GA37A-215 the Stall occurs at 53 mph for all that complexity (down from 58 mph). If flaps are not added but the Gross Weight were reduced to 800 lbs (from 924 lbs), the Stall speed becomes 54 mph. Put the two together and we arrive at somewhere near 49 mph.

Hope that clarifies some of the workings of a wing.

 

[Vigilant1]

Plugging that into the original figures using the Riblett GA37A-215 the Stall occurs at 53 mph for all that complexity (down from 58 mph). If flaps are not added but the Gross Weight were reduced to 800 lbs (from 924 lbs), the Stall speed becomes 54 mph. Put the two together and we arrive at somewhere near 49 mph.

Thanks, that's a helpful way of understanding the situation.
If it is possible to lose some weight and achieve an acceptable stall speed in that way and without using flaps, it offers a lot of advantages (less complexity = faster build, better ROC on the same power, a smoother, cleaner wing, etc). OTOH depending on the mission and design, flaps can be very useful as drag devices for glidepath control/energy management. In that case, flaps of fairly short span and occupying more of the chord become an option (allowing a shorter wing center section and easier trailerability.).
In the case of the OP, I suspect the target gross weight is already a bit optimistic, so setting it significantly lower may not be realistic. If this weren't a pseudo-replica, using a larger wing would be another option for achieving the desired stall speed without flaps.

 

[PiperCruisin]

 

[BJC]

That is a high Reynolds number for most E-AB airplanes with flaps deployed.


BJC

 

[WARPilot]

Using the construction method, infused-molded wing and fuselage, the surfaces will be glass smooth. Paint will be spayed into mold and bonded to the parts.
Flaps may be worthwhile but add more steps in the construction and some redesign to get flaps of meaningful span. They will no doubt add several lbs to the airframe.
This has been a wonderful thread…..very helpful and eye opening.
We are designing another airplane with my friend and had an Aero Engineer with his notebook CAD program running some preliminary lofts. We talked about the various wing sections, Riblett, NACA. He suggested the 63215 but said the GA37A-215 would be good also.
2 degrees dihedral, 2 degrees twist. There was discussion of using change in camber to mimic twist or building in the twist to the wing.