Designing for a High Performance 3M Racer
Do you feel the need for speed?
In this instalment of my four part series, I’m going to go through the basic thinking and outline design process used for my high-performance Redshift 3M model, and this is why there are so many references and pictures of it only as I don’t really know how others do their designs. I’m not going to call my design an F3F Racer per se, as I think it has a far wider flight envelope than just hard, repetitive competition use so, let’s just call it a 3M Racer. Though some if it may be a little controversial, these are my ideas and given honestly. Now actively developing the Redshift Mk II Spada version, I hope that this article will help to give you an insight into the thought processes behind where I was going, and incidentally still am, on the model’s design evolution. I may reiterate or repeat parts of previous subjects here and I hope the reader will forgive me but the recapped points are important and relevant to this article, and in any case, I can never tell who has read my previous work. Hopefully it won’t be too boring. — JH
First Thoughts…What Do I Want from a 3M Racer?
Development. I have a few 3M Racer designs already under my belt and there is nothing like learning from a previous design, so with that in mind; in order of appearance — I’d like to see: outright straight-line speed, great acceleration, fast controllable turns, overall agility, light or heavy air capability and convenient ballast adjustment all wrapped up in a good-looking package, and let’s not forget strength and toughness. Actually, most of those requirements are complementary, and with the knowledge we already have under our belt, none of them cancels out any other, which is useful.
Let’s Get Down to It — What Do We Already Know?
Wing Span
Well it’s called a 3M racer, so it may as well be a 3M (120”) racer — but why? Over the years this kind of model has gone from slightly longer flying surfaces through phases leaning towards slightly smaller wing spans. The thinking has been that a slightly larger span could fly in lighter air or maybe carry more weight and still be FAI legal, while a smaller model can be more agile and turn faster, and each is undoubtedly true — though which is the more valuable, performance-wise could be something of a dispute. But the mean tends to remain close to 3M, or 120-inch span and that’s what I have found to be the best size. So, lets kick off with that.
Takeaway: There are an awful lot of Racing models out there these days, all close to the 3M benchmark. Ergo: Unless many of us are wrong, 3M is the best size for the job.
Lift
With a 3M racer its absolutely essential to get the lift pattern distributed along the wing in the places where you need it, and ONLY where you need it. There really is no point at all in having large chords out near the wingtips. It’s not needed there and it can actually be harmful to the overall performance at both high and low speeds. Just like an Allrounder slope model, the 3M racer needs to have a planform that optimized for its job — only even more so, as some of the ‘wannahaves’ on an Allrounder become ‘gottahaves’ for the 3M Racer. The most important Gottahaves are straight line speed, fast, crisp control response in three axes, energy retention in turns, and out of turn recovery acceleration.
To recap a little from the last article in this series, we need lift to make our model fly, and we know that lift can easily be swapped for speed, but there is a balance needed here. We know that we need to optimize the planform shape in order to have an elliptical lift pattern span-wise across the entire wing with the most lift close to the fuselage, and the least amount at the wingtips.
We have also learned from the Allrounder design that while a true ellipse might be great for the lift — at least in theory — it’s actually not so good for model flying qualities. What tends to happen is that the Mean Aerodynamic Chord (MAC) and the Centre of Gravity (CG) can find themselves too close together, which can lead to instability and a tendency for the wing to stall if even mildly provoked. Conversely, separation of the CG and the MAC by too much, can lead to over stability and sluggish control responses. We know that a true ellipse that extends to the tips will bring problems, plus we know that the pesky chord distribution might also have an effect too. As on any self-built model aileron chord and span problems are OK because we can deal with them after the model is made.
We need to control the wing chord size to limit it to what is needed to put the lift in the right places. I also need to control what happens at the tips. So, putting those together, just like the Allrounder, I come out with an elliptical shape but with the rear (Trailing edges) pulled back to make the rear curve flatter, and the front part (Leading edges) more bowed — so that should sort out the MAC Vs CG problem very nicely, but I’ll still keep my elliptical lift pattern. For the tips we’ll just cut them off and give a more focused and controlled point for the isobars to depart in a more organized manner. A bit like sweeping the wings back on the straight-edged model, I know I am going to give up a bit of pitch and roll maneuverability, but I’ll gain stability and control, and best of all limit the tip stall possibilities.
Wing Aspect Ratio
This is an important if not critical part of the wing planform design that at least on my designs plays a large part in the flying quality and speed potential. The Redshift is designed at over 19:1 aspect ratio, where many of the older designs are around 16:1. On my model this is no accident. I have said that putting the correct amount of lift in the correct place is possibly THE most important part of a racing model design. Having broad chord wings will not help with anything except the ability to carry weight. The model needs to be able to carry weight up to the FAI limit of 5Kg in total weight with a loading of less than 75 g/dm² and no further than that, Ergo: having really large wing areas will not help anything except the ability to fly really slowly in very light winds. My problem is — and maybe I’m wrong — I don’t design racing models to fly, slowly…in really light winds… To be honest, if I encounter that kind of conditions, I know there will be no race, and the model will stay in the car anyway. Large wing chords and carrier deck wing areas are simply not needed on a racing glider and at 3M span will only slow it down.
Ailerons
From the Allrounder design we know that the tip shape and the position of the ailerons is also critically important. An elliptical or rounded tip shape is likely to cause a lot of trouble as will ailerons that end too close to the tips. On a racing model expected to perform fast, tight turns any disturbance at the extreme ends of the wing is a recipe for disaster so it’s good policy to keep the wingtips clean.
Takeaway: The importance of a well designed and developed wing planform cannot be over-emphasized on a 3M racer.
Takeaway: Put the lift in the right places.
Takeaway: We don’t want the problems with the wingtips so keep them clean.
Aerofoil
I don’t know if all would agree with this, and I suspect that many will not, but as I have said, this is how I do it: There are many aerofoils out there. Unfortunately, there are also many that are “proprietary” or “secret” — there are even those that are available for a price (!) I hope you will forgive me, I but strongly oppose this doctrine for two reasons:
- I’m a commercial model aircraft manufacturer, but I maintain that it’s not in the true spirit of aeromodelling or progress in model design to restrict the information because of commercial reasons. It’s not only about the money, guys.
- Over the years much progress and a hell of a lot of fun has been had due by sharing information. I’ll say it again: I am a commercial model sailplane designer and producer and yet I still admire and always do my best to help those who want to design for themselves, and that includes open-source availability of my aerofoils.
The Best One?
I’m sure that this is going to be controversial but anyway. This is a myth. There is not and never can be a “best” aerofoil as the way that the aerofoil is used and is positioned on the wing has such a huge effect. Good ones? — Yes, Great ones? — Probably. Best one? — Nope.There really is no one killer aerofoil that will blow all the others away, nor will there ever be one.
Why?
Problem is we demand a lot of different things from our aerofoils and some of those demands influence others, so at best we end up with a bit of a “Jack of all trades” — even if we are lucky enough or smart enough to come up with a good one. Some work better in outright speed, while some work better in the turns, just as some carry ballast better — which is better?
Is the Aerofoil the Most Important Component of a Fast Model?
This is where I will get into trouble, but the answer here is no. It’s an important contributor for sure, but not the most important part of the whole. Having a good, fast, responsive, low drag aerofoil working for us is very important, but not the key.
Fact: Wing planform, or how the (good) aerofoil is positioned and thus the lift distributed over the entire wing is far more important and influential. Yet, surprisingly many people think that the aerofoil is the single biggest deciding factor in deciding what’s a killer plane versus an “also ran” Logically it’s really no good having a super aerofoil if it’s distributed in the wrong positions across the wing, because however good it is, it just can’t do its job properly.
So, Let’s Look at a Couple More ‘Digital Myths’
Fact: Yes, some sections may be better than others when compared on a computer simulator, but the actual flying difference is slight and not enough to give any kind of clear advantage. There is in fact little practical difference in performance between most of the more often used aerofoils — and I have lost count of how many I have wind-tunnel tested over the years, so I can promise you that this is true.
Fact: Computer simulations may give some idea, and can help to compare one aerofoil against another to some degree, but believe me folks, it ain’t necessarily so. The results that you get on a computer simulator, and it doesn’t matter which one, can be quite different to what the aerofoil shows in wind tunnel or even more accurately, flight testing. Just think about it, if computer simulations were perfect, or even in the ballpark, then why would organisations like NASA spend so many millions and millions of dollars constructing huge wind tunnels? Has anyone ever seen the test unit at NASA Ames? You can lost in there.
A good example of this digital world vs real world phenomena is a series of profiles that were specifically designed for use with flaps some years ago by a well-known and highly respected designer. When tested on a computer simulator they did not show up too well, in fact the results were possibly below average. But — put them in a wind tunnel, and more importantly on an actual flying model and they were really good.
Why the Difference?
First, computer testing is done in a digital, number-based bits and bytes environment and not in actual gaseous air. Yes, all the numbers can be manipulated to simulate different linear conditions, but the big problem is, we don’t fly in different linear conditions — we fly in constantly varying conditions.
In a computer simulation the digital air flows smoothly over the digital section and the results are displayed digitally according to the parameters we input.
But, on the slope the nonlinear air flows erratically over our physical section.
On just one pass, observing his model’s behavior, what passes through the controlling pilot’s mind could be:
- more lift on the right side?
- more compression?
- thermal coming though?
- too close to the edge?
- climbing slightly?
- diving slightly?
- Whoops…too close (MOM).
All of these variables in most cases would result in small correcting control inputs…therefore the aerofoils spatial position changes constantly. This is reality.
What’s the Solution?
In fact, the aerofoil needed will be very similar indeed to that required for a slope Allrounder, so to recap: What’s required is a semi symmetrical section (not flat bottomed) with a thickness of between 7.5% and 8.5% — and a camber of around 1.8% to 2.5%. This is the sweet spot. Why? because at this thickness the camber line of the section will have a good curve, and will create enough lift to carry ballast if needed, and it should still be quite responsive.
At this thickness range the section can deal with a large variation in model weight, yet its thin enough to be low drag, while still being thick enough to be structurally viable and capable of withstanding high aerodynamic loads. There is no point going below a thickness 7.5% because there will be little or no advantage on a slope soarer, and even possibly a loss of performance due to the wings having to be strengthened and made heavier to compensate for the lack of structure. By the same token there is no point in going over 8.5% as the extra lift is simply not needed, while the drag escalates pretty fast with thicker sections.
Last but not least: any modern aerofoil with a decent alpha performance does not need any rigging angle.
Takeaway: Don’t think that the aerofoil selection alone is the ultimate path to a fast model. Wing planform is far more important.
Takeaway: Don’t believe that a computer simulation is the ultimate selection process. A computer selection is only a guideline — the actual performance of an aerofoil in flight could be quite different so talk to people who know.
Takeaway: Choose a nicely proportioned aerofoil between 7.5% and 8.5% to get the best overall performance from your model — look at what others are using.
Takeaway: Sections like the RG15, MH32, My JH25, and JH35, some of the HN sections, the HQW sect ions all fit the bill
Here are some Lift Vs Drag curves for the JH35, an aerofoil that I designed for the Alpenbrise alpine soarer to give low drag with high response to control inputs. For this section, the flaps and ailerons are designed for a 25% chord position. Flight tests will tell if it works as well as I hope it will.
The Back End — To Vee or Not to Vee, That Is the Question
The other obvious variable throughout the years of racing sailplane design has been the backend configuration: That old cookie, X-Tail or V-Tail? Both have advantages and disadvantages so, in case you did not read it, let me hark back to my last article in RCSD:
V-Tail
Good: Less pieces/joints so theoretically less drag, can be helpful in stabilizing the model in strong winds, and less chance of landing damage. Also: Fashionable — I kid you not, this is also a big reason for the V-Tail
Bad: Loss of much of the rudder control, slightly less stabilization surfaces control effectiveness overall, little or no drag advantage in practice as the inputs need to be greater for the same model responses, not so good for nice aerobatics as the control forces can be in the wrong directions.
X-Tail
Good: Better overall control, little or no actual difference in drag, decidedly better for nice aerobatics.
Bad: Not fashionable, more pieces so theoretically more drag, more risk of landing damage.
X-Tail, V-Tail, the choice is up to you. I have done both types through the years though many of them were back in the days where personal “one-offs” or friends group builds were the only option. But if I was asked which type is better for a racer, I’d go for the X-Tail every time due to the more open and detailed flight envelope that this type allows.
Fact: The world FAI F3F record has been repeatedly set and broken here in Taiwan on many occasions now, by the famous and talented ‘Mr. O’, flying a special version of the Needle — an X-Tail racer. Alas, as a commercial model aircraft designer I have to bow to the biggest influencing factor here and that is fashion. V-tails are fashionable at this moment, and while many things can be displaced by logic, fashion is not one of them.
Stabilizer Shape
Follow the wing shape that you have used as much as possible — this is not only for looks, but also effectiveness as the things that we have discussed for the wing shape are all valid for the Stab too.
Takeaway: Both X-Tail and V-Tail work and both have tradeoffs — which one is better for you?
Takeaway: Secondary considerations are the Stabilizer aerofoil to be used, and its span, plus V-angle and area.
V-Angle
Normally anything between 100 to 106 degrees seems to be the norm. I use 104 degrees because I’d prefer to err more on the side of elevator effectiveness rather than rudder.
Stabilizer Aerofoil
A low drag symmetrical aerofoil of between 7 to 10% is required. For all my recent models I have used my JHSYM-10 aerofoil, and recently the JHSYM-9 at a controversial 10% and 9% thickness respectively — more thickness than most people would go for, but there is method in my madness. Through testing the aerofoils WITH elevator movements, I quickly found that the thicker aerofoils actually have less drag and more control response than the thinner ones.
Stabilizer Area
Remembering that a V-Tail — if used — will need to do the job of horizontal and vertical stabilizers, if you make something about 17% to 20% of the wing area, you’ll be on safe ground. In this range, the Stabilizer will be big enough to be effective, but not needlessly over large. Too small and you will need to make a lot of pitch adjustments and too large actually has the same effect as the model will be over-stable and then needs to be forced in pitch. On the MKI Redshift I erred on the side of smaller tail area — which eventually turned out to be a mistake as under some conditions the model was marginal on pitch and yaw control.
Stabilizer Shape
It’s a good idea to follow the wing shape that you have used as much as possible — this is not only for looks, but also effectiveness as the things that we have discussed for the wing shape are all valid for the Stab too.
Takeaway: Thinner aerofoils do not necessarily have less drag, and may actually lessen control response.
Takeaway: A tail volume of between 17 to 20% of the wing area will work well.
Takeaway: Elevator setups work better than AMT type.
Takeaway: V-Tail or X-Tail, make the Stabilizer shape similar to the wing shape — the same rules apply.
The Fuselage
How Long?
The first thing to be considered for the fuselage is the moment arms. By this I mean the distances between MAC positions on the wing and tail, and from there, the wing MAC to the end of the nose. Think of these measurements in the same way as levers, but remember the weight considerations too. The longer the lever, which is the distance behind the CG to the Stabilizer MAC, the easier it is to move the load which in our case is the area in front of the CG.
Tail Moment
Tail moment distance is pretty critical as if it’s too short, you will need big control movements to change the pitch and yaw, but too long and you will add unnecessary weight behind the wing and all of the weight behind the CG needs to be counterbalanced by adding weight in the nose.
For the tail moment, a good ruler is somewhere between 3, to 3.3 x the wing chord. Shorter and you will get into dead pitch response problems, and longer, maybe have to add a lot of weight in the nose.
Nose Moment
Here again we have a lever — this time one that acts in the opposite way to the tail moment and counterbalances the tail end. There is no harm in having a nose that’s a bit longer than “normal” as the extra leverage length allows you to use less added nose weight, and the drag penalty is almost unmeasurable. In a perfect situation, the battery is all the nose weight you need. Practically a nose length of between 1.6 to 1.8 x wing chord works well for all considerations.
Cross Sections
Most racing model fuselages tend to break in front of the wing, behind the wing, or just in front of the stabilizers. This is always due to a rough landing involving sudden whip of the rear parts. I have tended to make my racing fuselages a tad wider than is the norm to counter this and to give a wider cross section in these sensitive areas, but in fact as we are making our own here, sensible use of materials can eliminate most of the danger. Look at golf shafts, and I bet you have not seen too many broken in play. Sensible tapering and careful arrangement to avoid stress raising of the cross sections will pay off. No sudden changes in diameter and try as much as possible to stick to around or triangular cross sections that aligned to resist the horizontal whip risks
Cool Factor
One last parameter for the fuselage: Make it your signature! Fool around with the lines until you have something that not only looks good, but also has large enough — but not too large — cross sections that will handle landing whiplash etc. — especially before and after the wing positions. The strongest cross unidirectional section is the round shape.
And In Conclusion…
Overall folks, there is a lot of value to the saying that “if it looks good, it flies good” — especially for really quick slope soaring models.
©2021 Dr. James Hammond PhD, DBA
This is the third part of a four part series. Read the previous instalments Designing for Slope Aerobatics and Designing for a Slope Allrounder in the March and April 2021 issues of RCSD respectively. In the fourth instalment, coming up in the June issue, author James Hammond turns his attention to really big alpine soarers. Don’t want to miss it? Best subscribe to our mailing list! All figures and photos are by the author unless otherwise indicated. Read the next article in this issue, return to the previous article or go to the table of contents. Downloadable PDFS: just this article or this entire issue.
