Radio Control: Scale
Bob & Dolly Wischer
Finish for all-metal prototypes
The silk-and-dope finish method (described last month) may produce the longest-lasting and toughest surface for models expected to have a long life. In our collection is a Fleet Trainer that is 37 years old, finished with nitrate dope. Our Emeraude, at nine years with butyrate dope, shows few blemishes. Some of the flaws that appeared were due to the use of balsa substrate that was too soft for weight-saving reasons; the tiny warps really stand out in the glossy surface. Careful selection of balsa would avoid these problems. We have heard it said that butyrate dope continues to shrink throughout its life, but we have not found this to be true.
To simulate the surface of an all-metal prototype there is probably nothing more realistic, durable, and easy to apply than fiberglass cloth and polyester resin. Lightweight fiberglass cloth weighs only one-half ounce per square yard. Finishing resin, applied properly, adds minimal weight while affording maximum strength and toughness.
Preparation and application
Preparation of the structure—sanding smooth with progressively finer grades of sandpaper—is the same as for silk covering. Dope is not used as a base for covering because polyester resin will not cure if applied over dope.
Fiberglass cloth is laid dry over the surface to be covered, and finishing resin (such as that supplied by Sig or K&B) is applied through the cloth using a short-haired, rather stiff-bristled brush. Resin intended for boats or for patching rust spots on cars is not suited for our purpose: it is very difficult to sand and builds weight rapidly, being too thick to apply lightly. Finishing resin can be brushed out to a minimum layer. If the quantity applied appears too wet, it can be removed before setting up with absorbent tissue. However, very little resin can be removed this way, and rubbing down does more good as a means of bringing the cloth into close contact with the balsa surface. An alternate method is to sand afterward, which requires additional time and effort.
Fiberglass cloth is very easy to apply and forms willingly around even complex compound curves. After curing, the cloth and resin are sanded lightly. Special sandpapers minimize effort—for example, open-coat silicon carbide in 180 and 320 grit, intended for wet-or-dry sanding of painted surfaces.
Filling and priming
At this point the cloth weave remains evident, and the weave must be filled to achieve a metal-like surface. Adding more resin will fill the weave but adds weight and requires more sanding. Filling is done by brushing or spraying a lacquer primer such as Ditzler Primer 32 (light gray), available from an auto paint supply store. After sanding, if weave is still evident, another primer coat is required. Examine the surface closely with a magnifying glass to determine whether the weave is truly filled.
Primer is heavy and will affect final weight. It can be sanded all the way down to the cloth for minimum weight, or given only a light sanding (first with 180 grit and then 320) for a polished, scratch-free surface. Ditzler Primer is extremely thick and needs thinning (we use dope thinner) at about two parts primer to three parts thinner for spraying.
Simulating panel lines
Simulation of panel lines is easily accomplished using masking tape and unthinned primer. Lay tape along the desired line and brush a generous coating of primer along one edge. After it dries enough to sand, sand down to the tape line using the same two grades of paper, feathering into the surrounding area. Remove the tape and a realistic panel line appears. Use thicker tape or two layers for deeper lines. Intersecting lines require careful sanding to avoid removing previously built-up lines.
An ideal tape for sharp lines is Karoden Flex-Mask (available from hobby shops). It will make the curved lines that represent fillet outlines more easily than standard masking tapes.
Topcoats and finishes
Any popular model finishing material can be applied over Ditzler Primer, including dope. We have used Floquil model railroad paint for a flat finish to simulate camouflage. Its advantages are light weight coverage and good hiding power—one coat covers large areas with minimal material, colors do not bleed, and adhesion is good. Our test panel, treated with raw fuel, seemed unaffected except for a slight change in gloss due to oil residue. Five or six fluid ounces will cover a six-foot model.
RC Scale — Flight Realism and Scale Speed
As one of the hardheaded Virginians that Dale was unable to convert to his test for scale speed, I feel a strong sense of déjà vu in offering a few thoughts on what's wrong with Dale's idea that a full-scale plane and a model—when spaced at distances that make them appear the same size—should appear to track together at the same speed. The problem stems from the increasing field of view of the human eye at increasing ranges. The viewer's perception of speed is largely a matter of how quickly an object traverses the range of his field of view, and perception of speed (specifically by a judge) is what we are dealing with. Since this field of view increases with range, far-away objects (the full-scale plane) always seem to be moving slower than nearby objects (the model) at any speed.
As you drive down a highway, for example, notice how quickly the telephone poles located near the road seem to move past you. Identically spaced poles farther away seem almost to stand still. Both sets of poles "move" at the same speed, but the near ones (the model in Dale's analogy) seem to be going faster. Stand close to an interstate and get a sense of how fast the cars are going by; move back a few hundred feet and they hardly seem to be moving at all. Both are going the same actual and scale speed. Dale's proposal is neither academically sound nor practically workable.
When attempting to achieve "realism" on a reduced scale, we are talking about dynamic modeling. Brad Powers' recent articles for MA are relatively complete—a tried-and-true tool of the engineer. Just as other parameters of a model are not simply the full-scale value divided by the scale factor (weight, area, and horsepower to name a few), neither is scale speed. One can consult the literature and apply the proper calculations, but since we aren't using radar or timing traps to accurately measure scale speed, such precision is unnecessary. What is required is a workable approach to determining flight realism—scale speed (however measured) being only one factor.
I have been the victim—in my scale modeling pursuits along the East Coast—of two misconceptions (largely from the ruminations of John Preston, for whom I have much respect): 1) flight realism consists only of "scale speed," and 2) this realistic speed is slow. At last year's Bealeton Scale Meet (which was well run and a complete pleasure, and in which I placed third in Sport Scale), a Byron Pitts was flown. The model fairly putt-putted around the sky and was on the marginal edge of underpowered: no vertical performance, small loops, and slow rolls. The judges, following the guidance that "slow is good," felt the plane was flying an excellent scale speed. However, during the full-scale airshow that followed the meet, a Pitts Special tore up the sky with blazing speed, seemingly unlimited vertical capability, and crisp maneuvers. Many agreed, to their surprise, that the model Pitts was too slow.
The answer is to judge flight realism on a range of factors, with overall speed being just one. Here are items we might include in a guide to flight realism judging:
- Does the model use throttle as a flight control?
- No aircraft that I'm aware of completes a loop at full throttle or runs wide open throughout an entire flight.
- Are the maneuvers chosen appropriate to the aircraft modeled?
- For example, you wouldn't expect a scale business aircraft to execute a loop or a full-scale P-51 to make an inverted pass on the deck. Maneuvers inappropriate to the subject should be downgraded.
- Are the maneuvers, even if properly chosen, executed in a realistic manner?
- An axial roll completed in little more than the fuselage length or retracts that extend and retract in a blink are not realistic.
- Is the entire flight realistic, not just the chosen maneuvers?
- A plane with the retract gear down for an entire flight (whether or not retracts were an option) cannot be judged highly realistic. Scale flight is sequence, rhythm, and tempo: how the model moves from one attitude to another, transitions from level flight to a climb, the size and shape of loops, roll coordination, and throttle use all contribute to overall realism.
- Are the control responses of proper magnitude and timing?
- Larger models have slower control responses; an exaggeratedly twitchy model betrays itself. Use of exponential or mixing can help achieve realistic control feel.
- Does the model present the correct sensory cues to the judge and audience?
- Scale proportions, paint, strobe and exhaust effects, and sound all play roles. A beautifully flown but unmistakably non-scale-sounding engine reduces realism.
- Are crowd perception and judge bias considered?
- Judges who equate slow with scale can be misled; training and clear definitions help.
Conclusion: It is incorrect to limit scale realism to a single dimension such as speed. A score should reflect multidimensional aspects of flight, including choice of maneuvers, execution, tempo, transitions, control harmony, and presentation.
Bob and Dolly Wischer Rt. 1, S-221 Lapham Peak Road, Delafield, WI 53018.
Transcribed from original scans by AI. Minor OCR errors may remain.
