Mike James

Ok, now you've all done it...

All the recent discussion of multi-segment flaps and leading edge devices has got me interested in doing some tests. I'll construct a .60-size high wing platform that can use several different test wings. Each constant-chord wing set will be approximately 68 inches in span and 12-inch chord. (just to keep the attachment to the fuselage simple) For structural reasons, I'll probably use the F-16 style droopable leading edge flaps, rather than (more fragile, at this scale) slats with a slot. Maybe one of you will build something with deployable slats? My model will have a T-tail and tricycle landing gear. On one or more of these, I'll also employ slot-lip spoilers for roll control.

This winter, I'll build three sets of wings for these tests, with the expectation of testing them all in early spring of 2003:

1. Wing with single-slotted Fowler flaps and fixed leading edge, with (fixed) NACA droop in front of the ailerons.

2. Wing with single-slotted Fowler flaps and leading edge flaps

3. Wing with double-slotted Fowler flaps and leading edge flaps

4. If the double-slotted Fowler flaps work, I might have time to tinker with the triple-slotted type.

Those of you with advice on airfoils and other details about these items are welcomed to post here or email me. I expect to use an Eppler 197, but am open to the idea of trying some others. Let's collect some real data!

Ollie

It sounds like you have the test model hardware well covered. However, you have said nothing about exactly what you hope to measure, the instrumentation to measure it or the test procedures. For example, to get air speed, are you going to measure the ground speed and direction, the wind speed and direction and then compute the vector sum? Take off length? Load carrying ability? Stalling speed?

Will you be able to extract the maximum coefficient of lift of the various devices from the measured data? How?

Mike James

Hi Ollie,

Unless someone wanted to contribute (or loan) me some instrumentation to prove my opinions, this will be a "seat of the pants" test. I simply intend to construct several dimensionally-similar wing sets, each with a different possible configuration, and report on "what works, based on eyeball observations". I'd simply like to see the effect of all these devices on a medium-size sport model, focusing on the ability to takeoff and land in a smaller area, at a lower airspeed, and also to investigate extremely slow speed maneuvering.

I'd be interesting in analyzing some of the other aspects of flight that you mention, Ollie, so maybe this can become a team effort...

I'd need some technical assistance, as well as some hardware, to produce truly scientific results. If readers here would like to contribute expertise, equipment, or suggestions to achieve more finite results, I'm willing to be the builder and flier part of a team effort, and document the results both here and on my web site. Then maybe we can all benefit from the results.

I'll go ahead with the plan on my own, regardless. It should be fun!

BMatthews

Sounds like fun Mike. Some observations if you don't mind.

Our models enjoy an unprecedented power to weight ratio thanks to our totally free design practices and a wide range of very powerful output to weight ratio engines. For takeoffs we don't need flaps or other high lift devices for many sport models, just point them up and let go. To a certain extent it's the same with landings. Funfly models sometime use negatively reflexed ailerons to kill lift rather than augment it in order to get their flyweight frames onto the ground on demand without arriving as a controlled crash. I have an 84 inch 5 1/2 lb Old Timer model that has a 10 foot takeoff roll in dead air and settles into a 25 degree climb on an old OS loop scavenged 35. By all accounts this should qualify as a STOL model. In my old timer's case the high lift device is the highly cambered Gottingen airfoil.

The full size world is a much different story. Here they don't have the surplus of power. Or if they do, as in fighters, they are using the high lift to compensate for painfully small wings and for reducing the take off and landing rolls by lowering the stall speed to allow for slower speeds in the pattern while still having a small enough wing for the high speed mode.

So for these high lift devices to provide an advantage we need to be talking about heavier and lower powered models to set the stage. I don't know what parameters you're thinking around but obviously there's no point in adding high lift devices to, say, a Goldberg Eagle with a modern 40 on the front. There's plenty of lift for it's weight already. Drag flaps to kill the float on landings is another story mind you. Gliders have used this for quite a while. Also they've used these flaps for camber changing over a few degrees to add or subtract a small amount of camber to great effect. With the gliders once you get beyond a few degrees then all they are interested in is drag and not lift. In fact they've got too much lift on landing approaches as well.

So you have to ask yourself "what IS the niche where high lift devices would be a benefit a model". I believe the happy home and appropriate outcome for a project such as your's is a cargo carrying model. Not the SAE types where it's just a case of takeoff, fly for however long and land all on a HUGE paved field but rather the <ahem> grass roots field where you may want to tow a glider, haul a video or other heavier camera aloft, a bomb bay full of candies for kids at a show or whatever you can think off where there's a load to be taken up. In this case a good flap system added to an already relatively high lift airfoil could enhance the slow speed to some significant advantage. Takeoff rolls would be shorter allowing the use of more fields with higher payloads and landing approaches with higher wing loadings into tighter fields would benefit from the drag and lift of fully deployed flaps.

So I'd say your test bed should be a cargo carrying type with a large capacity fuselage or a removable lower mission pod type of design.

Mike James

BMatthews,

You're right, of course. My motivation is to accomplish 3 things:

Multi-mission plane -
I'd like to have a sport plane that could fly moderately fast and perform aerobatics, then switch to the opposite extreme "very slow STOL" mode, for takeoff and landing. (Based on my past experiences with concepts from Andy Lennon's articles, I know that the Eppler 197 is one of the appropriate airfoil choices.)

Test bed for personal project -
Paul Reese and I are working on a 1/6th scale King Air B200, which will include slotted Fowler flaps. This model can serve to prove some ideas for actuation, flap track construction, etc. ( http://www.nextcraft.com/b200_construction01.html )

Research for everybody here who is interested -
I'd like to get a feel for how much difference in lifting capability can be gained by going from single-slotted flaps to double-slotted flaps, and maybe even triple-slotted flaps. That's where your idea of a more highly-loaded model comes into play.

Last night I did some research on the internet on "high-lift devices", and found a ton of useful data, from drawings and CFD data (Computational Flow Dynamics) to NACA reports on various configurations. So, I've created a new link on my "Projects" page, called "A Study of High-Lift Devices", and will document the design, building, and flying of this model, in several configurations. It's at
http://www.nextcraft.com/j46_highlift01.html

All input, suggestions, and requests are welcome. And, as I mentioned before, if any of you can help produce a more scientific analysis of the data, by either doing some math yourself, or by perhaps providing some instrumentation, that's even better. This can be a team effort, but I will continue, regardless.

It should be fun!

banktoturn

Mike,

This is really cool, thanks for volunteering. One observation that I'll make is that the relative positioning of the elements of multi-element flap systems is very critical, and it is probably difficult to predict in advance of the tests what is the optimum positioning. I don't know whether this is something that is very dependent on reynolds number, but intuitively it seems that it could be.
One suggestion for you is to make you test wing so that the elements can be repositioned between flights. This probably means a somewhat clunkier implementation, but it is, after all, in the name of science. What pops into my head is a constant chord wing with flat mounting surfaces available on each end to allow the elements to be repositioned and remounted. The inboard surface could simply be the flat side of the fuselage, and the outboard surface could be a painfully ugly tip plate. How to securely mount the elements and also have the position be infinitely variable is a detail I haven't worked out, but I am assuming you might have some ideas about that. I will try to do some reading over the weekend to see if we can come of with a rational range of relative positions for testing.

I think Ollie raises an excellent issue, and one that we hobbyists face when we try to generate objective data. In an ideal world, we would want to do something on the order of a wind tunnel test, to truly change one variable at a time, and know the values of the variables with some certainty. Having said that, serviceable wind tunnels are hard to build, and accurate wind tunnel testing is even harder that we hobbyists would believe. This leaves us with building a flight test vehicle and measuring some aspect of flight performance. The two obvious ones ( to me ) are stall speed and maximum lift. Measuring airspeed accurately has its own challenges, so I would be inclined to build a plane that can be loaded with ballast, so that maximum lift can be determined. Of course, since the airspeed and AoA will vary from flight to flight, this approach would not allow us to generate all the plots that a practicing engineer would want to have in order to evaluate different designs, but it would give you a pretty good idea which configuration would generate the most lift with the engine power held almost constant. This is a very meaningful measure, and would likely tell you a lot about which flap configuration is close to the best. Flight testing with a plane loaded to capacity would certainly be a challenge, and this might be the show stopper for this testing approach.

Comments welcome,

banktoturn

BMatthews

Banktoturn, that's a great idea for saving on building time. If you go with a stock Clark Y type airfoil and set up hardpoints at the trailing edge the hinges or tracks could then be added on as required for each experiment. Depending on the span I suspect there would need to be 3 or 4 hardpoints, root and tips obviously plus one or two in between to support the elements. To accomadate the differing TE shapes that would be required I think a sub trailing edge could be built in and then a final edge could be tack glued, screwed or just Monokoted into place to fair in whatever flap arrangement was being tried at the time.

Similarly the leading edge could be an add-on in the same manner to allow different slots or movable element shapes to be tried.

For the flap section you'd want to make the last 20% of the chord removable. For the front I'd guess at 10%.

One advantage is that you'd be working with the same core airfoil so the comparison between types of systems would be consistent.

And with a larger test bed the size of the tracks or hinges wouldn't be as significant. I'm thinking a 6 foot span model with a 40 and a cargo bay area here.

Fowler flaps at this size wouldn't be hard at all but Fowlers with even one little sub flap would be a toughy. The small element would need to be formed from something quite stiff. Carving from solid spruce at the very least. Actually I suspect it would need to be formed from carbon fiber to prevent it having all the stiffness of a wet noodle. With a 10 inch chord we'd only be talking about something that was about 3/8 to 1/2 inch in chord after all. This would call for two mid span supports so there would only be a 1 foot span between supports. That would make the spruce carved option a practical reality.

Ya know, this IS starting to sound like fun. It could sorta look like a big Storch or that PZL glider tug, service plane and wind up being a practical camera or glider tug in it's own right afterwards.......

Ollie

Consider this: As the coefficient of lift increase and, that is what you are after, the down wash angle increases and the tip vortexes get very large. This is acompanied by a very large increase in spanwise flow! Now the results are heavily dependent on the aspect ratio of the wing and its lift distribution. How are you to sort out the data for two dimensional flow so that the data can be applied to other cases? There is a lot of theory that has to be applied twice instead of once and this increases the uncertainty of the results because the real world never quite matches the assumptions inherent in the theory. It's not a case of the theory being wrong, it's a case of the real world not matching the assumptions twice in tandem. I'm not trying to be a wet blanket but I do want the large effort you are going to put into this pay off. Also, consider that part of the wing is going to be in the disturbed air of the prop wash and fuselage turbulence. Wind tunnel designers go to a lot of trouble to eliminate turbulence so that the data is something that can be used in a variety of situations. Twin engine planes have a larger percentage of the wing in propwash than single engine planes. If you get data from a single engine experiment and try to apply it to a twin, I think you may be surprised by the lack of similarity in results. If you can't define the flow conditions before the airfoil enters them, what will the data mean? Maybe people build wind tunnels because of these sorts of difficulties.

banktoturn

Ollie,

You are definitely correct in that 2D results do not correspond exactly to 3D results. All wings have spanwise flow, even those built with the highest feasible aspect ratios. However, by doing the 3D experiment rather than the 2D experiment ( building a wing rather than testing a flapped airfoil section ), we would actually be getting away from doing the work twice, in the sense that we ( Mike ) would be measuring 3D wing performance rather than predicting it based on 2D airfoil results. As flawed as this process is, it is done this way all the time, and there is little choice, since buying the necessary equipment and getting the necessary people with the necessary training are not options for us. Modellers here are constantly getting advice on airfoil choices, often from you, only to go and implement wings of finite span. This experiment is every bit as valid as that practice. There will be error, but the data, if carefully collected, should be nicely applicable to models of various configurations. You are also correct that these problems are among the reasons for using wind tunnels, but a successful wind tunnel is probably out of reach for this endeavor. I did give some thought to a test rig in the bed of a pickup truck, or stuck out the sun roof of a car, but I think the disturbed flow and the difficulties of getting accurate force measurements seemed bigger than the problems with Mike's proposed test scheme. Maybe that's not true.

banktoturn

Ollie

Prof. Michael Selig has been running a program at UIUC for over a decade fueled by the dollar contributions of modelers, model clubs and model manufacturers and by the test models that modelers such as you and I build for the UIUC low speed wind tunnel. Why not contact Prof. Selig and see what you might do to enlist his help. You might also enlist the aid of scale modelers and their associations everywhere to contribute to the project in one way or another. Prof. Selig might give you some useful suggestions for the airfoil configurations that are of mutual interest and test them if people will build them. This might not be as much fun as building and flying a test model but it would be of far greater benefit to the modeling community and to the Aero. Engineering students who are participating in competitions. As I recall, the test models are 12 inches chord and 35 and a fraction inches in span with specified joiner tube locations. He will be looking for contour inaccuracies of less than 0.01 inches. The test model specifications and contact information are at :
http://www.aae.uiuc.edu/m-selig/

Mike James

Thank you all for your input! I appreciate your suggestions, and will do what I can. The issue with some of the suggestions, for me, is the money. I'm obligated financially to other projects already started. So, I can certainly build and fly this test model, but can't afford personally to build a wind tunnel, buy accelerometers, etc.. I'll do my best to come closer to "pure" measured results, but won't let these problems keep me from following through, regardless. Too much fun!

By the way, I corrected the specs in my earlier post, and will be using a 12 inch chord airfoil of 68 inch span. (816 sq. inches) and a .60 2-cycle for power. This gives a little more physical space both in the wing and in the fuselage for "stuff". The overall layout is "trainer-like", with the exception of the T-tail. I've already done the preliminary design in CAD, and posted an image below. (wheel pants mostly for show)

About airfoils:
I haven't contacted Professor Selig personally, but am certainly aware of the work going on by his team, and have downloaded as much of the airfoil database as I can find, to use in plotting airfoils. I'm using Selig airfoils on my King Air project, based on Ollie's advice, and have used several Eppler airfoils (based on Andy Lennon's work) on past projects. On several similar high-wing sport models I designed, I achieved a really nice balance of speed, aerobatic power, and slow speed flight (with single-slotted flaps) using the Eppler 197. One of my goals with this test is to fly the widest envelope possible, including relatively high speed aerobatics, rather than "just" slow flight. Until I discover a reason to change, I probably will begin this project with the E197 for the wing, and E168 for tail surfaces, since there's so much data already available for these airfoils. (Other suggestions are welcomed, especially if performance is improved.)

Ollie is right about single versus multi-engine performance. Not only can flaps be "blown", as on the YC-17, but I've also read a NASA report showing that the CL can be improved on many prop-driven multi-engine planes by placing the nacelles at a slightly downward angle, (downthrust) adding to the airflow over the upper surface of the wing.

Rigidity of small parts doesn't worry me too much, since small ply, metal, or plastic supports could be built into the primary flap and flap track to support the smaller flap segment, as is done on full-size flaps. I've spent the last few days reading NACA reports downloaded from NASA, collecting all the photos and drawings of the mechanisms I can find, and measuring parts in CAD. That, combined with materials I already have used, has put me pretty close to a final basic design for the test bed.

Again, thanks to everyone who is reading this and/or providing input. "Many brains are better than one". Happy flying!