I'm in the middle of building a highly modified version of Paris White's 1/12 scale DC-C and am building it VERY light for electrics. The idea being that the lighter the plane the less the wind loading and the slower it will fly. Am planning on at least 3 degrees washout since the wings taper fast from almost 14 inches at the root to 5.5 at the tip.

Is this right thinking or nonsense?

By the way - since a 1/12 scale DC-3 should cruse at 12.5 mph and have a VNE (redline speed) of 20, is there any prize for scale flight over scale looks and has anyone been able to pull it off? And does anyone actually keep to a 1/12 scale DC-3 climb rate of 100 ft?min?

Steve

You should be fine with the 3 degrees. As far as flying at scale speed, that could be a challange. Flaps down will allow the model to fly slower but you can get into trouble using flaps.

My first "general aviation" flight instructor, a woman actually, Mary Lippit, taught me, "Fly high fly slow, fly low fly fast." This was 1966.

Charles

Thanks for the confirmation on my washout guess. I'm not building flaps to try and save weight and will hopefully be able to test fly it here next July over the many acre thick green airbags (other wise known as soybean fields).

While a 1/12 flight speed would be a perfect scale speed I think you'll need to temper that a little with reality. A 7 foot span model will need to be built to a crazy light sailplane like weight to keep the flying speed down to even near to the mid 20's mph region. This may just not be practical.

And I also concur with the washout figures. You can't go wrong with lots of washout on a highly tapered wing such as this.

Good luck with an interesting project.

Those who have clocked flying speeds of 20 mph will know that it is all but impossible to do with wingloadings much over 10 oz ft you can wash out wash in wash up -whatever makes no difference once you attempt speeds where required low wing loadings are impossible to produce
My Seniorita which weighs 4 lbs and has 800 plus squares -on a low aspect ratio wing -will barely stay controllable at 20 mph.
My ERATIX weighs 3.75 lbs and has well over 700 squares on a very low aspect very numb thichk wing and yet -it's min flying speed is about the same

My 450 sq in flat foam wing ONE LB 3D model will fly controllably UNDER 20 mph
note: it has a 1/4"thick flat wing an no coupled flaps
yet it easily flys the slowest-controllably
Why?
it ain't the airfoil -it's the weight that gets you at low speeds.
Scale speeds would work on th DC3 if it were a flat foam electric weighing a few pounds
you could even add a scale airfoil -if the weight stayed the same but in this case- weight is everything.

Don't get put off by the naysayers, it can be done, as Dick says, by really good weight control, but you really need to build bigger, 1/12 is likely to be too small to achieve the required loadings with the fixed weight of the propulsion and control systems needed. 1/6 scale will increase the areas and the speeds for scale and still let you fit the fixed equipment and maintain the proper loadings and strength requirements. It just gets easier as the size increases. Build as big and light as you can.
Evan, WB #12.

Wow Evan, though I'd really like to agree (what I do in general here), especially in this case there's a downer. I did the math (quad wing loading) with DC-3 spec's from Wikipedia: Because the DC-3 is such a big slow-flying featherweight full-size (I hear you laughing) you'd have to build it to 20 lbs weight in 1:6 scale (190" wingspan) and even only 1.2 lb in 1:12 scale (95" wingspan) to have her fly "scale" speed. (I hope I've got the initial question right and made no calc error.) Those would be real featherweights, even 1:6.

I wholeheartedly agree not to give up and build as big and light as you can. There are several successful examples for instance on the annual Aspach e-flight meetings, not only [link=http://www.rcgroups.com/forums/showthread.php?t=294293]2002[/link] (especially the BAe 146 and other models by Hans Bühr). The result may not fly "mathematically scale" (just faster) but it should still look very scale.

If you aim for the the original cubic wing loading, which is sometimes seen as a measure for correct scale behavior, the 1:12 version could weigh even 14.5 lbs what should be possible (lighter still being better). Wing loading would be 34 oz/sqft what is not really low for a 8 ft high-aspect-ratio wing but is not bad either. In this case, I would find flaps instrumental in getting scale landings, though. The washout is advisable at this wing loading, and flaps would additionally help cure the tip-stall problem.

For a scale take-off and climb, plan for a thrust-weight ratio of less than 0.45 (even 0.35 should be enough) and use low-pitch props, just sufficient for target speed. That will help you keeping the weight low (small motors and batteries), maybe even more than a c/f skeleton framing construction with thin shell Ã* la Hans Bühr.

You will definitely need that 3 degrees of washout to compensate for those terrible reynolds numbers you will be operating in. Remember that cruise speed is typically double that of the stall speed and is about as slow as you want to go in normal manuvers. Even at 16 oz wing loading which is extremely low for what your doing, that would put you at a 40 mph cruise speed. It is more likely you would be in the 32 oz wing loading so something more 50-60 mph range would be more likely.

Ustik, it's not fair to compare the cubic wing loading from the full size to the model. That's too big a change and doesn't take into account the whole reynolds number shift for that big a change. And I guess it depends on your tolerance for flying bricks but when you're talking 34 oz/sq foot on anything that is sub 100 inch span I would suggest that this is well and solidly in the flying brick category. A much nicer weight to achieve a scale'ish look to the flying would be more in the 8 to 10 lb range. But to make it look like it's really hanging in the sky like the real ones do I suspect the weight will need to be more in the 5 to 7 lb range.

Let's see..... The 1/12 model would have 6.8 sq feet of area. At a sailplane like 10 oz/sq ft that would imply a 68 oz, or just a little over 5.5 lbs all up weight. Sailplanes with that wing loading typically can fly at around the 20 mph mark with ease. And I'd say that the illusion of scale speed could be well, if not accurately, maintained up to likely 16 oz/sq ft which would give a weight target of 6.8 lbs. I think that the 6.8 is quite do-able as long as extreme measures are taken to making the airframe weight come in low enough. Even 5 to 6 lbs is do-able if this design does not require solid skin covering but instead uses open structure that is covered with material of some form. I'm not familiar with the plans mentioned so I can't comment further on them.

In the 20 to 26 oz/sq ft range you'd have a model that many would consider a complete success but the flying speed would definetly not look scale like. But as the loading hits 30 and up in this wing area range bad things will become the norm.


Another test I like to do in these discussions is hit up NASA's Foilsim II to get an idea of how slow something can fly or how fast it needs to fly to avoid any harsh issues.

http://www.grc.nasa.gov/WWW/K-12/airplane/foil2.html

I set up the simple span and chord to provide 6.8'ish sq feet and then setting the camber to 3% and the angle of attack to achieve a lift coefficient of 0.75. At 0.75 you're up in the high lift region but should be comfortably away from the stall of most higher lift style airfoils. With those set I altered the flying speed until the lift amount hit 5.5 lbs. The flying speed to achieve this was 20.25 mph. A very nice sailplane like speed even if it is well above the 12.5 mph scale cruise speed.

Foilsim says it would need to be down in the 2.0 lb all up weight range to fly at a true to scale speed of 12.5 mph. If you're willing to cheat on the airfoil and use a higher camber with 6% camber and run it up near a Cl of .9 and where it's getting a little close to the stall speed for most folks comfort then you could get away with a 2.5 lb model. You can see why I was saying that you'll need to compromise for the sake of reality. Not to say that a 2.5 lb model is not do-able but it would need to be constructed like a rubber scale model with lots of minimalist strips and covered in only the very lightest of films or jap tissue and dope. Handling and flying of such a model would require the calmest of conditions and care. Oh, and the wing loading on a 2.5 lb model would be only 6 oz/sq ft. Right smack dab on the money for wing loadings enjoyed by the lightest "gasbag" sailplanes from the early days of gliding like the Olympic 99 and others of that era.

Incidentally at 14 lbs and a Cl = .7 for a safe'ish landing speed or minimal cruise Foilsim says that the flying speed would be 34 mph. It doesn't sound like much but it would make a 6 foot DC3 model look more like a WW2 fighter doing a quick pass.

How are people figuring scale speed? I might think of it in terms of airplane lengths traveled per second. Is that reasonable?

Wow! Lots of posts to read and think about!

Thank you all very much!

For what it is worth I have a target weight of 4.5 lbs. and so far I'm right on.

The span is 93inches and the wing area is conservatively 850 square inches.

From the start I planned to stringer rather than full sheet and finish the entire thing with tissue and dope.

With formed paper wing and tail fillets and engine cowls I am doing my best to build as light as I can.

Might add an inch to the tip cord and take it from 5.5 to 6.5 on the off chance that even this little might help.

Steve

Yep, you're going in the right direction. A long time ago a group of us were building big and light, and we settled on 16 oz/sqft as giving a reasonable loading for scaleish flight. Was easy to work out too, wing area in sq ft equals airframe weight in pounds. There was some modification, WW1 and earlier were better around 10~12 oz/sqft, and WW2 and later looked better around 20~24 oz/sqft. We didn't expand on the envelope as you had to build lots of models to test the theory, and we just didn't build that many...
we got lots of comments on our 'helium filled' models, but since most of them are still flying, and 15 or so years later, you needn't be concerned about the strength or longevity of the things either...
Evan, WB #12.

Oh, I didn't know that it's a matter of fair or unfair. I just tried to keep things simple and set an upper and lower limit for the design. The Aspach meeting reports should show several real, working examples in between those limits. There are even more, e.g. [link=http://www.modelflight.regheath.com/mf145/gallery.htm]here[/link] even a DC-3 (and there are more when you search that site). It's easy to do the math for their wing loading. Just build how you think it's best and see where you come out between those limits.

I never claimed that the original's cubic wing loading would give scale-like speed and behavior (though some people claim so) because I know that quad wing loading would. But the latter would indeed require a helium-filled gas bag and would be flyable only indoors, and the former would be quite fast - but it would not behave as twitchy as a fighter plane because it's a calm and stable configuration. The only relevant Re number problem would be less maximum lift coefficient (and aoa), but you wouldn't notice it at the high wing loading (and with flaps).

BTW, I found 95" wing span and 6.85 sqft (987 sqin) wing area for 1:12 scale, just to compare the figures. What I meant as an upper limit of weight and wing loading (34 oz/sqft) would make for the scale speed of a jet transport and not a DC-3, that's clear. But flight behavior would still be fine at this model size, and the model would land like a pattern model. One of the Aspach attendants (Bill Kleinbrahm) built airbus models with that wing loading and reports them flying much like trainer models (even landing if the slotted flaps work, being a Re number problem). The people building jet transport models handle even more than 40 oz/sqft, that's where "nasty" things begin. But they aren't really nasty if you don't think of a DC-3 but a DC-9 which just needs speed. You just have to know and account for that.

And may I mention again that small motors and batteries will save you more weight than a stick-and-tissue construction. BTW, 12.5 mph is slow flyer speed, 20 mph is park flyer speed, 30 mph is cruise speed of a thermal glider. You know how that handles. They all can take some wind if they are set up not too stable.

The really light stuff fits into my early rule of aerodynamics which is:
If th model is too heavy -the airfoil really does not matter
and
if it is light enough -it still does not matter.
When I got into the very small indoor flat foam aerobatic type models,I found that from a practical standpoint -you simply could not make the model too light - too flimsy but not too light.
these models fly in a no breeze enviroment
The avid contestants can do aerobatics which leave most wondering what happened to the laws of flight.
In scale - -the best still sems to be stick and really good Jap Tissue
most new flyers have no idea how this stuff even works
the one shown here, weighs a litte over 1/2 ounce

UStik, looks like we are not that far apart on our thinking after all. And yeah, on a near 100 inch model 34'ish is considered fairly normal. I did the span math and then somehow got the 6 foot thing locked into my head. Upside down 9?

Still, as you say the upper side of 30 sure doesn't provide for a nice scale look. Even on a jet liner (watch them going overhead sometimes. They aren't THAT fast). But with care and staying within the envelope they can fly well.

Spalex, you have GOT to post up a build thread in the Scratch Building forum on this one. I want to see all the chins drop when they see free flight scale model techniques applied to this size of project...

Of course at this size and weight it'll be a sub 5 to 7 mph wind model. That'll limit your flying opportunities but the alternative is a nasty tip over during ground handling or on landing. After all, if you're scaling the wind speed that's 60 to 85 mph! ! ! Yeah, yeah, I know the air mass doesn't work the same at these speeds but this will obviously be a calm day sort of model.

One of the guys in my club built a 1/12 DC-3 last year.
He is builds some pretty light airplanes, his with a pair 0f .32's for power flew very nicely at around 6 lbs...
It did however have the classic tip stall/loss of aileron authority when he got too slow. I would say the slowest he could get it for landing was closer to 30 mph. any slower than that was an issue.
Given that you are building it for electric power your framework would not have to be any where as robust as his so if you could get your target weight closer to 4 lbs you may be able to get the performance you want.
As everyone has pointed out your most critical issue will be keeping the wing loading as low as possible.
If possible use laminated balsa instead of ply if you need a bit of strength as far as motor mounts and alike, you would be amazed just how much strength you can add to balsa with some 1/2 oz fiberglass applied with epoxy with all of the excess epoxy blotted off with a paper towel. Carbon fiber veil applied with dope also will add strength without adding a lot of weight for stress points such as spars and shear webs ribs and alike.
Using a very lightweight receiver as well as servos, and opting for pull pull control linkages, rather than rods is also good for shedding an ounce or two.
Using the absolute smallest lightest li-pols for your power requirements is also a must, granted this may limit your flight time but given the high C rating batteries available today this is not as much of an issue as it was even two years ago.
Bottom line get as much air into your airplane as possible!

Good luck!

Good point in the issues related to the taper of the wing.

Given that we know the lift coefficient for the root area will be up around .7 to .8 you may want to look at what angle of attack it takes to achieve that. Then lower the angle at the tip via washout to get the Cl down to around .2 or .3. And THEN use some fairly extreme differential on the controls as well as plan on using some rudder along with aileron at lower speeds.

A buddy of mine built a super light 36'ish inch span SPAD 13 for electric. In his quest to fly it at a scale speed he spends all his time about one notch away from a stall. Then he wonders why it produces a lot of adverse yaw even when he used extreme differential. I kept telling him he needed to use a little rudder with the aileron but he tried to avoid it. Finally he set up a mix and found that it worked much better.

This issue with the ailerons is a biggie with sailplanes as well. I've seen some nasty tip stall and snaps because someone got greedy with the aileron at low altitude and while trying to "stretch" for home. Low and slow calls for miniscule control inputs on the ailerons in any design and far more so with sailplanes or extreme tapered wings.

Guess you couldn't find it so here it is: the [link=http://www.rcgroups.com/forums/showthread.php?t=240358]Aspach 2002 report part 1[/link] with Charlie Binder's 1:10 DC-3. Wing loading is 33 oz/sqft! View the video to see how that looks (not "scale", of course).

[link=http://richard.ferriere.free.fr/3vues/douglas_c47_3v.jpg]Here[/link]'s a 3-view to have an impression of the wing. The [link=http://www.ae.uiuc.edu/m-selig/ads/aircraft.html]Incomplete Guide to Airfoil Usage[/link] tells the DC-3 had NACA 2215 airfoil at root and 2206 at tip. A quick check with [link=http://www.mh-aerotools.de/airfoils/javafoil.htm]JavaFoil[/link] shows the former being very docile (stall-wise) while the latter is nasty, at least at model Re numbers.

That might have been a problem even in full scale, considering the wing sweep giving additional pitch up when stalling the tips first. (And the normal down pitching moment of the airfoil is reduced at stall.) Had the original DC-3 wing washout or did they simply fly at high aoa only with flaps down? Both would work but it's still speculation. Someone here who knows the facts?

Good question, Jim. This discussion has passed over some of the basics.

To make a model visually appear to be travel at a realistic speed, you have to scale the relationship of size and speed. As you suggested, this means that the model would travel the same number of fuselage lengths per second as does the real plane. Then it looks right to the eye as it flies by. To accomplish this requires that the model weight be scaled by a definite ratio to the real plane. Without going into the math, it works out that "fourth power scaling" will get you the appearance of scale speed. As an example:

If we want to make a DC-3 model that is 1/12 of full size, the weight needs to be 1/12 X 1/12 X 1/12 X 1/12, or 1/20,736 of the full size airplane's weight. Since the weight of a full size DC-3 is around 35,000 lbs, the model would need to weigh 35,000 / 20,736 = 1.7 lbs to fly at scale speed with a similar stall margin.

That's why we don't see many models flying at scale speeds.

There is a further complication in this weight scaling business. If a model is "fourth power scaled", it only looks right in straight flight. A model scaled to fly at that speed will make non-scale maneuvers. For instance, a turn at any given bank angle will have an unrealistically small radius. If you were to scale a fighter or aerobatic plane, the loops would look impossibly small compared to the size of the model.

In order to make the size of your maneuvers look right, it is necessary to do "third power scaling". That is, the model weight would be 1/12 X 1/12 X 1/12, or 1/1,728 of full scale. That would be 35,000 / 1,728 = 20 lbs. But then the speed looks too fast.

The bottom line is that there is only one scale ratio which gives a perfect match of both speed and maneuver size. Unfortunately, that's full size.

The best we can do with scale models is somewhat of a compromise between visual speed and maneuver size. That would be about "three point five power scaling". If you have a scientific calculator, you can do that. In the case of the 1/12 scale DC-3, the weight ratio would be 1/5985, or 5.8 lbs.

I'm sure there will be folks who will disagree with this analysis, but "the physics don't lie". There has been a lot of pretty good work done in this area over the years. An example of where serious scaling analysis is used would be flight testing of spin models by NACA and NASA.

Dick

Interesting analysis there Dick. Obviously there's some room for fudging how the pilot flies the model in the maneuvers but there's no cheating the stall speed. The trick is to fly a light model with minimal control inputs so that the model is smooth and slow to respond. This can give the impression of a big heavy model but while it's flying at the more scale like speed that can only come with super light wingloadings. Sailplane pilots are often good at this since they've learned that pushing the controls around just causes drag so they tend to just breath on the sticks and wait patiently for the model to SLOOOOOOOOWLY respond with course changes. Of course all that goes out the window when they find a thermal but the rest of the time it's all about smooth.

Ustik, thanks for the link. Yeah, the flying speed of the model definetly confirms all we've discussed here in this thread. While it appears to be a nice flying model it's far, far too fast from a scale looking flight perspective. And it seems like he's definetly landing it hot. Likely to avoid the issues you mentioned.

Dick, thank you very much for this spot-on explanation! I've never seen one that short and clear. I only used the terms cubic and quad wing loading without explaining them. May I add that "fourth power scaling" gives not only scale speed but also scale motion (acceleration, roll rate, etc.).

Sorry, again I have different figures, mainly only 25,000 lbs weight for the original, leading to lower weights and wing loadings of the model as well. But incidentally spalex chose a target weight that comes very close to your "three point five power scaling" even at the lower weight.

This similitude mechanics, as well as electric drive layout, are two of my special interests in this hobby. (The third being testing different airplane/drive layouts in a simulator, as Evan did in reality.) So even if nobody cares, only for my pleasure:

The 4.5 lbs weight for the 1:12 scale model, giving slightly more than 10 oz/sqft wing loading, will make it like a park flyer. The original determines it to be a calm and slow park flyer (as opposed to an aerobatic one). So expect about 20 mph cruise speed, only low wind capability, and easy landings provided the tip-stall problem is solved.

As to the drive I should have been more precise. On second thought I'd recommend a higher-pitch propeller after all. Reason is that the model will need quite little power in cruise flight, what should be most of the time. For take-off and climb much more power is needed, and the motors have to be chosen to deliver it. But they can still be quite small when overloaded during the short time of take-off and climb, say one minute maximum. After that, cruise power (and current) is typically only one third of climb power. You can afford to waste some energy during the short climb, but not during the long cruise. So the props should work with best efficiency at low power (1/3) and rpm (1/2), requiring not much diameter but some pitch. That's the basic idea.

Assuming you'll use common outrunner motors and standard LiPo batteries, I have some matching drive layouts with 22xx motors. For a very scale-like climb (and take-off only from paved runways), a 2208/34 motor, 2s2200 LiPo battery, and 9x6 APC SlowFly prop would suffice. One such drive gives 0.83 lbs static thrust (6 A current draw), 50 mph top speed (!), but will work with maximum efficiency (45% total) at 20 mph or less with 2.0 to 2.5 A draw. The static thrust/weight ratio 0.37 will just suffice, but one complete drive weighs only 6.5 oz. Two weigh 13 oz, of course, what is only 18% of the total weight of the model. (Who would have thought a few years ago.)

Safer flying would be easier with the bigger 2212 motors. A 2212/34 with 3s2200 battery and 9x6 SF propeller gives 1.12 lbs static thrust, meaning 0.5 T/W ratio what is more than enough. Same speeds, same efficiency. Weight is 8.5 oz each, 17 oz both, 23.6% of AUW. Climb 7 A, cruise 1.5 A. With this drive, my parkflyer (same wing loading) cruise-flies up to one hour (demonstrated), but I use it for thermalling.

If you want to use scale-like props you'd need three-bladers with 11.75" diameter, which are far too big for this drive layout. But if you find a 12x7x3 with narrow blades you should be fine. With a 2s LiPo, the 2212/34 would give .38 T/W and a 2212/26 about 0.54 T/W. Weight would be 3.5 oz less (total). Current draw, rpm, and efficiency are different but still similar. Endurance is less but still more than you'll need.

Another idea is to build the model still with flaps, despite the low wing loading. The split flaps are small and simple, anyway. You'd not really need them but they might be handy. Forget the wee bit more lift, but the drag lets you control the landing approach, and the effectively increased decalage makes the model stable and easier to control (in calm weather, anyway) and should avoid the tip stall in landing configuration (take-off as well). The wing would be efficient in "clean" configuration, and "dirty" the washout would be achieved by increasing the inner wing aoa instead of reducing the outer wing aoa.

BMatthews,

Yes, I forgot to mention that aspect of scale flight. We can break "scale motions" into three general areas:

1. Linear motion. That is, travel along the longitudinal axis of the airplane. Speed, if you will. That's where fourth power scaling looks good.

2. Turning motion. That would be level turns and loops. Third power scaling works good at that weight.

3. Angular motions. Especially angular accelerations. These are motions around the axes of the airplane. Think rolling, wing rocking, and turbulence response. Full size airplanes have lots of inertia, so they respond slowly to control inputs and turbulence. Smooooth control inputs are definitely in order for scale-like flight. I once experimented with rate gyros in a park flyer (GWS Tiger Moth 400) and was surprised at the beneficial effect of the gyro in the roll axis. The roll gyro made the little Tiger Moth seem much larger, both visually and in pilot feel.

Dick

One other thought regarding scale motions that occurred to me after posting. Several years ago my friend Bob Hoey experimented with scale models of ravens (the birds). His models were full sized and had similar weights and inertias to the real birds. Even though they didn't flap, Bob's models looked eerily "alive" on a gusty, turbulent day. I'm sure it was the body motions that made them look so real. Bob even had occasion to soar with a gaggle of ravens, who accepted him without alarm. When Bob eventually ran out of lift and landed, one of the ravens landed along side and walked around the model, evidently wondering what was wrong with his new friend.

Dick

We used to have a utility C-47 at a place where I worked. I just now got off the phone with an old co-worker who was one of the company test pilots and flew our C-47 regularly. In answer to my question about stall characteristics, here is what he told me:

The airplane buffets very heavily as the stall is approached. In addition to the tail buffet, there is some pitch nodding. He said he did not recall pressing deeper into the stall because of the severe buffet. Also, his recollection was that he had satisfactory roll control at all times.

So there is our answer to why the full scale airplane has so many loyal admirers and such a good operating history. Apparently the buffet warns the pilot away from further increasing AOA and he never gets into trouble. As far as operating with the flaps down, the lift coefficient is increased with no further increase in AOA. I rode in our C-47 many, many times and don't the pilot ever doing a 3-point landing. Now I know why.

I agree with the other posters that any savvy modeler would look at the DC-3 wing and shudder. It has all the elements necessary to cause a nasty tip stall wing taper, thickness taper and leading edge sweep. Maybe if we could feel stall buffet in our transmitter sticks we could stay out of trouble too!

The only option that we have with our models is to carefully limit elevator travel such that we never quite get to the stall point. That's basically what the full scale plot is doing.

Dick

Dick, that story about the raven model is priceless! I laughed a good one at that.

The info on the DC3 and your experience with the gyro was also excellent. Far too often we, as modellers, want our models to ignore the charactarisitics of the full sized prototypes. We only do so to our own peril. We tend to get so used to not having the same limits as the full sized craft that we tend to get choked up when one of the full sized aircraft's foibles hits home on a scale model. Hopefully Spalex takes this to heart.