Thanks Steve - but the time horizon here is two years, so I don't expect to make it for the Scale Build Off 5. I couldn't even manage to complete my little Chipmunk last year which is why I haven't made the entry myself. But, never mind, now its there and we'll see what happen :-)

I have to start with "I am no airfoil expert either" but.. You can't scale air molecules and on a highly tapered wing like the Mossy or the DH 88 wash out might just save your model from a death spin. You are simulating scale and not going for top speed I assume. Also the model wing span is on the smaller size for wing loading. I am curious about the comments that will follow.

Very interesting project Jorn !
I will be watching!

Ed in Indiana, USA

Well, regarding the taper - you are probably right about tapered wings having higher risk of tip stalling - and washout helping with that. I don't have either experience or theoretical knowledge to answer that. It seems to be the consensus - but apparently it also depends on other things (like people building slightly twisted wings or placing the center of gravity wrong).

My personal take on this is 1) it is not a beginners model - the Mossie was a fast flyer, not a soaring bird or glider, and as such it has to maintain good speed at all times, and 2) it has flaps to help landing at low speed and thus reducing the risk of stalling at that moment.

Speed has always been a problem for me - I learned RC flying on a Multiplex TwinStar which is a great model and very suitable for the Danish weather. But it is also a very forgiving slow flyer (almost feeling like a glider) which meant that I never learned to land properly - and believed other aircraft would idle around on 25% without problems. Duh!

Low speed cost me a full model and lots of gluing when I started flying fighter models Especially my little Hobbyking FunFighter - its a fun plane, but do *not* fly it too slow! Landings is still an issue for me too - somehow I always get in too slow with the plane wobbling and almost stalling. But I have never lost a plane that way though - rough landings are annoying but usually also repairable ...

So, no, I do not fear a death spin due to tapered wings and missing washout - but I will probably break the landing gear once or twice before mastering the model

Nice to meet you Ed

Hi Jorn,

I love the Mosquito so I'll follow along, nice CAD work too.

Looks like you have a thing for DeHavilland's....

Denny

Originally Posted by Dennis Sumner

Looks like you have a thing for DeHavilland's....

Yep. For no really good reasons.

My local club is close to "Flyvestation Værløse (EKVL)" (http://www.jetphotos.net/showphotos....ase%20-%20EKVL) which is where I saw the Chipmunk first time at an airshow. It looked like a nice (Danish) subject which is why I chose that one.

Short after I heard about the Mosquito and thought it could be a suitable big brother to the Chippie. And now, having read a lot about de Havilland, I might just be interested in more historic builds from that company later on.

The biggest problem might just end up being storage space

Hello Jorn, I am glad you don’t mind comments on your wing design. I am finding this pretty interesting.

I hadn’t realised that you had used RAF 38 section as the basis for the wing geometry. It is a good choice as a starting point. I presume that you used the basic section all the way down the wing and at the inboard stations you extended the section forward to form the radiator geometry You still show the basic section that you used in the construction so I used that for the comparison with RAF 34 – it was bigger and a little more accurate using that than the smaller outboard sections, but I realise that the sections at those stations will incorporate the radiators.

I am going to have to stick with my assessment of the zero lift axis of RAF 38. Basically, the flow leaves the aerofoil section at the mean angle of the trailing edge and that mean angle is roughly the zero-lift axis of the section. If that axis is aligned with the onset free stream flow direction, the flow leaves the trailing edge at the same angle as the free stream. There is no flow deflection and therefore there is no lift force generated. There are graphical methods to determine the zero-lift axes that were derived by Max Munk at NACA in the 1920s. For the first order method, you determine the mean line at 50% chord and joint that point to the trailing edge and that is your zero-lift axis. For sections with complicated mean lines there is a second procedure that is a bit more complicated but the first method will work for RAF 38. For the figure I posted, I just sketched the axis in by eye for RAF 38. The figures for RAF 34 were taken from another figure I had already constructed based on a quoted zero lift angle – so the comparison of the two sections come from two different sources. That is why I didn’t recognise that the noses of the aerofoil sections should have been identical. They were pretty close, as I noted.

So that leaves you with a 3 degree difference in the zero lift axis of your wing with the full scale wing (supposing that it used RAF 34). It means that the aircraft will sit more nose-down in flight at a given speed in order that the lift balances the weight. That is a matter of aesthetics and not everybody is as sensitive as I am to the attitude of an aircraft in flight, but there are a couple of cases that come to mind that may be more important. In a 3-point take-off with the tail down, the RAF 38 wing will be at a much higher incidence than the full scale situation and may actually be close to stall. There is a video of a DH88 Comet coming to grief in just this way doing a take off from a 3-point attitude (like you might use to get off from a grass strip):

RC de Havilland DH.88 Comet, Flight, Crash and Burn. (2 min 22 sec)

The answer is to get the tail up right away and accelerate to take off speed on the main wheels but in that case with the wing producing significantly more lift and induced drag the take off run will be longer – not a problem for a model but a real issue for the full scale due to limited runway length.

There is also a landing attitude issue, especially with the flaps down. In this case the lift balances the weight until the point of touchdown and the nose is significantly more nose down. At touchdown there will be sudden horizontal loads imposed at the wheels due to the drag of the wheels (especially on grass) and spin-up loads as the wheels accelerate to their running speed. If the aircraft is nose-down already it may already be close to its nose over tipping point and not have enough margin to overcome the undercarriage nose-over loads. So there may be some real cases to consider, not just the aesthetics of the situation. The angle is about 3 degrees between the sections and if you decide to stick with the more cambered section, maybe you should think about changing the wing incidence setting to compensate.

Now the foregoing all depends on what aerofoil section was actually used on the full size aircraft. I had a look at the photos on the New Zealand restoration web site you posted and I think it shows that the section was based on RAF 34 for sure. If you study the photos – especially of the flaps – you see that the upper surface rear is absolutely flat, no dell and no convex curvature. The lower surface does have convex curvature on the other hand. Actually, the standard RAF 34 has very little concavity in the rear upper surface and any modification made by De Havilland must have been simply to ease manufacturing since straight members could be used to make the ribs of the flap. I attach a couple of pictures that I borrowed to illustrate the geometry.

By the way I think your attitude toward flying this aircraft is dead right. There are no experts on the stalling behaviour of a new aircraft design. Changes to lift curve slope and stalling angle vary gently with Reynold’s number for most aerofoil sections but there are exceptions and you may be unlucky enough to get something that behaves differently from the full scale. So you have to assume the worst and operate with the aircraft in trim, with a speed margin to the stall and in the approach use flaps and power to adjust decent angle and use power to cushion the touchdown – until you find out whether you can take liberties with flaring manoeuvres. I wouldn’t be too worried by the behaviour of the DH88 Comet models you see. Its wing has a much higher aspect ratio than the Mosquito and much higher taper. The Mosquito’s wing is also swept slightly forward, which also acts to unload the tips. Even if you incorporate some washout you won’t know until you test it how it is going to stall, so your approach to flying has to be the correct one.

I think that it is time for me to go back to watching mode. Regards John

Regarding that poor Comet - my Chipmunk has a really bad habit of doing the same kind of roll if the take off speed is too low. Take a look at the maiden video at 0:54 sec - https://youtu.be/S_7g_o3iQHg?t=54s. Right after start it rotates strongly to the right, but luckily I manage to get it somehow under control

It happens easily for it if there is some wind which will make it take off too early if I am not careful. The problem is also aggravated by the small wheels which requires a lot of up-elevator to avoid tipping over - and up-elevator means taking off to early if you are not careful. Add to the equation a less-than-perfect grass runway which will make the poor machine jump up and down while gaining speed - and, well, it takes some training

So hopefully I'll be prepared to handle the same kind of behavior on the Mosquito when time comes - giving it a lot better fate than the sorry Comet there.

Regarding the zero lift line: there is something here I do not understand. If you take the polar plots for C-lift versus alpha for RAF 38 then it shows Cl=0 for alpha = -2 deg which is a lot less than your suggested zero-lift line. See for instance http://airfoiltools.com/airfoil/deta...af38-il#polars. If you take the RAF 34 then the zero-lift line is almost the same. Where is my misunderstanding?

This is a fascinating discussion. Being a "that looks about right" modeller myself, I've never dug into the aerodynamic theory.

Jorn, I must apologise for adding you to SBO 5; I was in the process of adding several new entries and I guess yours just got caught up in the excitement. I can remove it if you'd like.

Steve

Steve, no need to remove me from SBO5. At least it wasn't myself who added me - which mean I won't feel bad by not making it on time

Originally Posted by Steve85

This is a fascinating discussion. Being a "that looks about right" modeller myself, I've never dug into the aerodynamic theory.
Steve

It's been a "light bulb" moment for me too. I think what John is telling us has great practical value and it would be good to tease out the general conclusions a little more. If it needs the space, maybe an aerodynamics boot camp thread?

Hello Jorn, I just did the Munk first approximation calculation accurately (based on your posted figure) and I get a prediction of -3.1 degrees for the zero lift angle of RAF 38. If you take the calculated curves for that aerofoil section you get an angle closer to -2 degrees for this section., as you say.

The zero lift angle for RAF 34 was given by an aerfoil source as -0.5 degrees. Munk's second method should be used for this reflexed section to derive a number that is consistent for comparison, but taking the quoted number for now, zero lift angle for RAF 34 seems to differ from the calculation by the same sort of amount but in the opposite direction.

I must admit that my first instinct is to base design on measured trends or consistent simple inviscid calculations rather that put a lot of faith in complex calculations. However, these calculations can expose physical problems to watch out for even if you do not put faith in their absolute accuracy.

I have backed out the points on which the calculated curves were based and I have added some general aerofoil section trend lines and shapes to put the calculated curves in perspective on the attached figures. There certainly seem to be features in the calculations that don't fit the general picture. I'll work out Munk's method for RAF 34 and give the attached figure some thought in the morning if you like. The moment curves have some very unexpected trends and excursions. I would be skeptical right now about the them if they don't ft the trend. I'll see if it looks the same in the morning, regards John

" If it needs the space, maybe an aerodynamics boot camp thread?" That would be great if John could come down to where us mortals can understand some of the things he is peddling.