On Silent Wings (Airborne 224)
For this issue we have a special treat for you, a short article by John Skinner on how to achieve the best performance from your thermal sailplane in “dead air”. John is one of Australia’s leading soaring competitors. Amongst other accomplishments he represented Australia at the World F3B Championships in 2007, placing tenth. This is a considerable achievement considering how few opportunities there are in this country to hone your skills in top class F3B competition.
Much of the content of this article first appeared, as a series of short questions and answers, in a discussion on “dead air” flying on one of the Internet discussion groups. With John’s permission and assistance I have edited it to suit the format of this column. Over to John:
Dead Air Flying.
The basic theory for dead air flying is quite simple. For a given launch height, and sink rate, the model will arrive at ground level in a given time. For example:
Launch height = 200m
Sink Rate 0.3 m/s
Flight time = 200 / 0.3 = 667 seconds (11.1 minutes)
Obviously maximising launch height, and minimising sink rate will give the best dead air duration time. Also note that it appears possible to get the 10:00 minute task from quite modest starting conditions. As usual, the devil is in the details!
How to maximise launch Height
A good launch is vital for a good dead air time, so the first fundamental in dead air flying is a high launch. In dead air you must launch like there is no tomorrow.
For launching in dead air, using a FAI legal winch (1.1 kW), tension is hard to build in the line, so launching like there is no tomorrow means:
Hook the model up onto the line. Select launch flap. Power the winch up, and hold the model with winch power on until you get the maximum tension that you can reliably hold, and still be able to throw the model accurately (see Figure 1).
There are a variety of techniques to throwing the model, but the end result is to throw the model vertically, throw the model up. I emphasize this, because just letting go of the model causes it to leave the launcher’s hand horizontally, with the model moving forward five meters or so, before rotating into the climb. Tension is lost, and more line is wound onto the winch drum. You don’t want the model to rotate. It should leave the throwers hand already in the climb (see Figure 2 and Side Note 1).
Side Note 1: I should mention the importance of the team around the pilot. In Figure 2, you’ll notice from the left, Steve Keep driving the winch, Florian Lindner applying silicone spray to the line on the winch drum, myself throwing, Nick Chabrel timing and calling, and Tim Kullack, the pilot, obscured by Steve, standing back to get a good pilot’s view of the action. When Mike Reynell was launching he sat in a chair behind the pits!
If the model then climbs to 15m in height, tries to peel off the back of the line, then drops its nose, and then starts to climb again it is not set up correctly, or you didn’t push a little bit of down elevator at the right time. The bob of the nose just cost you another few meters of height, and remember each meter is worth 3 seconds of flight time!
Keep the power on all the way to the top, otherwise tension is lost, and cannot be re-built by the time the model reaches the top of the launch. All the way through the climb on the line, the model should be flown as gently as possible, with no aileron input. Any coarse control application simply creates inefficiency, drag, loss of lift and loss of tension. Fly gently. Drop a wing and you’ve just lost another 15 meters of height (50 seconds)
At about a ten-degree angle before being directly above the turnaround pulley, neutralise the launch flap by selecting speed mode. Immediately push over, aiming the model at the turnaround. Diving this way avoids sweeping the line through the air, which creates drag, and therefore energy loss. You want to convert the tension built on the efficient climb, efficiently into airspeed. Just before line tension is lost, stop the winch power, pull up (with drag reducing elevator to camber mix), and climb out nearly vertically at high speed (the ping). A near vertical climb converts the high airspeed most efficiently into height. The mix minimises the loss of energy as you pull up. Done wrongly this can cost you big time in terms of launch height, say 30 meters, or about 1.5 minutes worth of height!
Just before the stall, click from speed into thermal mode and push over strongly to arrive at the top of the climb out at or near minimum sink speed. Don't stall; if you do you will lose five to ten metres of height! (15 – 30 seconds duration)
If you’re throwing the model flat, don’t dive at the right time into the ping, don’t have the right setup and don’t select the right flight mode, then embracing the technique I have just described will gain you at least another 15m in launch height and perhaps much more! Given that modern models sink at around 0.3 m/s, 15m is equivalent to another 50 seconds airtime at minimum sink!
I should stress that these techniques are for dead air. In any sort of a breeze, tension is attained far more easily. Less line is wound onto the winch drum to hold tension, and tension is easier to build. The model can be kited up to near the top of the launch, and then power applied. This has an advantage because line is not wound onto the drum, so directly relates to a higher height at release. As the breeze gets stronger, the dive to accelerate starts earlier, so in windy conditions the model might only climb to forty degrees from the vertical line above the turnaround before the dive starts. In these conditions a heavier line, and smaller winch drum diameter are used (see Side Note 2).
Side Note 2:
The match of line diameter and winch drum diameter is extremely important. In dead air, less tension can be built; so smaller diameters are chosen to allow the line to stretch to closer to its maximum length. A large stretch in the line allows a better conversion of the line tension into speed for the ping. The lower diameter provides less drag during the tow.
In low winds less tension can be built. Smaller line diameters and larger drum sizes are used to limit the tension that can be applied by the winch to avoid line breakage. The larger drum also provides line speed so that tension can be built faster when the power is on.
The reverse is also true. In high winds, the higher tension automatically built allows thicker lines to be stretched to near their maximum too. Drum sizes are smaller to allow the winch to provide maximum tension. Line speed is less important.
Quality of winch line is also important because of different diameter to breaking strength (which relates to line drag during the climb), and elastic properties so that all the tension built returns like a rubber band rather than like a piece of chewing gum (which relates to the height of the ping). In competition the line is chosen for these characteristics, in sport flying durability is more of a concern.
In wind speeds above around 3 to 5m/s, it is possible to build huge tension just by weaving and circling the model on the line, and I think it illustrates how, with wind, tension can be built by kiting the model. Contrast this to dead air, where tension is built using the power of the winch alone.
I have described here the two extremes. Most conditions will be somewhere in between, and therefore the technique will also be somewhere between full winch power all the way and power at the top of the launch only. The release point also varies, but in all cases the model should be pointed at the turnaround during the dive to avoid sweeping the line through the air (see Side Note 3 for the effect of wind direction and thermals on launching technique).
Side Note 3:
Wind direction and thermals also change things during the launch. Flying into a thermal on launch will build tension very quickly, and sink will kill your launch! In wind, weaving has the advantage of positioning the model for the most efficient into wind climb.
Weaving also has an advantage in dead air because tension can be built and maintained before climbing to the top of the launch. Thermal duration competition tasks are nearly always flown man on man, so weaving or circling on launch is likely to result in mid-air collisions during the launch unless you’re on the end of the row of winches!
Launching 102: There could also be another chapter written on setup of the model for launching, covering topics such as flap settings, elevator presets, tow hook position, mixes and multiple launch modes and speed mode for the ping. In short, the model setup is extremely important to attain a high launch, and the setup also interacts with the launch technique. Models that are set up on the edge need the perfect launch technique!
How to minimise sink speed
So, now we’re at the top of the launch, and hopefully at a high launch altitude because we have a good technique, and a good setup. To minimise sink speed remember that every control input increases drag and reduces flight duration, so the second fundamental is that in dead air, minimum sink speed must be maintained, and all drag production must be at a minimum. In these conditions your glider is a radio GUIDED free flight model. Don't touch the sticks unless ABSOLUTEY necessary. Never bank more than a few degrees, fly long straight lines or a VERY large orbit, judged just right so that the model is on the spot when it runs out of height at 10:00 minutes. If you must turn, try using small inputs of rudder alone, and be patient. With small control inputs the model will take some time to respond.
The other advantage of flying hands off to the limits of vision with a stable trimmed model is that you can see what the air is doing; you have the horizon in your field of view. Even the slightest lift can be seen. I have NEVER seen truly dead air. At large distances, subtle lift can easily be seen in reference to the horizon. You can choose to slow down even further to prolong the time spent flying through it, or turn back to fly another straight line through it. How strong is that buoyancy? It might not be strong enough to circle in, or it might be; it’s your decision! If you choose to circle, you should re-assess the air and your height on every part of the circle. Be prepared to stop turning and fly straight again. Be aware also that the circle might maintain height. That is good. Lift is any air that stops you coming down as fast as you might have done by flying straight!
To achieve an ideal hands-off flight pattern, trim and balance are extremely important! With a forward CG your model is incredibly pitch stable, and so does not need constant control by elevator; it flies itself, without the inefficiencies caused by frequent use of the controls to maintain a constant airspeed and heading. If you do encounter subtle lift, the soft nose heavy trim allows you to breathe on up elevator to sample the air.
To follow this through, a more forward CG means a little more up elevator trim for slow speed flight. So if the model is upset nose down, and wants to fly faster, the up trim slows it down. If it is upset nose up and wants to fly slower, the more forward CG, now overpowering the up trim, drops the nose and makes the model fly slightly faster. The more forward the CG, the more one speeded the model becomes, and therefore the easier it becomes to fly at the trimmed (minimum sink) speed. Easier means fewer control inputs, smoother flight, and I believe a better performance at minimum sink on those "dead air flights" because it is flown hands off in pitch. This type of trim holds the model in a very narrow range of airspeeds (see Figure 3) with a minimum of input from the pilot. It is stable.
By contrast, a glider with its CG further back and corresponding less up elevator trim can still fly at the same airspeed. It is closer to the classic neutral stability. Models trimmed like this will hold the attitude they are pitched to, so if pitched to a thirty-degree dive, the model will hold that angle, and speed up. A ten-degree climb will also be held, until the model slows down and stalls. Such a model is “all speeded”; the elevator trim does not change for any speed at which you want to fly the model. The problem with this trim is that any disturbance, for example turbulence, will pitch the model one way or the other, resulting in a departure from the minimum sink speed and additionally the need for the pilot to apply a drag producing elevator control input to bring the model back to the correct airspeed.
This trim is closer to unstable, and is high work rate for the pilot to fly at constant slow airspeed. This almost ensures that the model is not flying at constant minimum sink speed all the time and certainly needs more control input to keep it there (see Side Note 4). End result, you come down faster!
Side Note 4:
The trim with rearward CG is actually closer to the best elevator trim and CG for speed flight. Moving the CG forward softens the elevator feel. To obtain the very tight turns required in F3b speed, the CG is moved back to make the elevator more powerful. The elevator throws are also reduced, and exponential used to cut the sensitivity due to the more rearward CG.
Notice also that I said “elevator trim and CG”. The two are interlinked only at minimum sink speed. I like to think of it this way: At slow speed the aerodynamic forces are small, so weight distribution (i.e. CG) takes on a more important role in defining the attitude of the model. Slight up elevator trim at slow speed requires more lead in the nose to counteract the up trim.
If you find the correct elevator trim for the “neutral dive test”, that will be very close to the correct speed mode setting, regardless of CG changes (i.e. the all speeded model mentioned earlier) I say this in a dead air article, because good performance in speed mode is important for the good launch. The correct elevator trim for this is close to the required trim got from a neutral dive test.
Don’t move the CG back without reduction of elevator throw, or you’ll have a model that tip stalls easily, and you simply can’t have that when turning in F3b speed, or for that matter, flying at a distance during a dead air flight. A forward CG will need more elevator throw.
I have never been convinced that a rearward CG signals lift better. All models signal lift, but they behave differently!
How do I know when I am flying at minimum sink?
I actually do this by feel. It is possible to have a trim where the model sits nose up, dragging its tail. It is on the verge of the stall and control becomes a little sluggish. This is too slow. The speed for minimum sink speed is slightly faster than this, perhaps by 1 to 3 km/h. I trim for this slightly higher speed by either putting a click or two of down trim in, or adding a slight bit (say 10g) of nose weight. So in summary, I find the stall speed, and work to a slightly faster speed from there.
In practice I tend to fly at a slightly higher speed than minimum sink. The response of the model to the slightest touch of up elevator tells me what type of air I'm in. In subtle sink a breath of up elevator sees the model sag. In subtle lift, the same input sees the model hold its energy without sagging. I think I said earlier, very rarely is true "dead air" experienced.
Minimum sink speed actually depends on a whole lot of factors relating to the type of wing profile, camber, aspect ratio of the wing, and wing loading. So, in general, for modern designs, minimum sink speed is a few km/hr faster than the stall. For the older types such as the Paragon and Gentle Lady minimum sink speed is much closer to the stall.
To summarise, when flying duration tasks in dead air:
- Launch Like there is no tomorrow.
- Set your model up with a forward CG to give adequate stability.
- Trim for minimum sink trim or very slightly faster.
- Fly long straight lines or very wide turns with the absolute minimum of control input.
- Be on the lookout for signs of subtle lift, the air is never really dead.
Captions for Figures.
Caption: Figure 1: Building the tension before launch. Launching Nick Chabrel’s Caracho 3000 at the F3b World Championships 2007. Note arm locked, feet wide apart, maximum effort to hold the high tension before throwing the model. Nose pointing up, green and gold parachute! (Photo courtesy of Steve Gleeson.)
Caption: Figure 2: It has to be right from the start! Launching Tim Kullack’s Radical at Jerilderie, 2007. Note that the nose is up, the model is already flying. Maximum effort into the throw. (Photo courtesy of Chris Adams. www.lsfaustralia.org.au)
Caption: Figure 3: The polar is for my own design F3b model with +2.5° flap. It shows how important it is to hold the model in a narrow range of air speeds. The optimum air speed (Vx) is at 8.9 m/s, with performance worse than the 0.3 m/s sink speed (Vz) benchmark when speeds are slower than 8.5 m/s (30.6 km/h) or faster than 9.2 m/s (33.1 km/h). The model has to be flown within a narrow speed range to obtain the best minimum sink.
(Note: This article is a copy of one published in Airborne #224. If there is any reason it should not be published here please notify the MRSSA administration via Contact to have it removed)