A couple of years ago Steve and I were flying different models of P-51 Mustangs; both were very adequate in light lift but were rather semi-scale in outline. Both were eventually sold on, and within a very short period of time it became obvious that disposing of ones’ favourite light-lift model was probably a mistake; anyone attending PSS events must have at least one model for light lift conditions, there’s no point in making the journey and then not being able to fly because the wind isn’t blowing hard enough. Accordingly, thoughts turned to co-designing a suitable but more accurate replacement that would satisfy three key requirements:

a.)  Aerodynamics to be optimised for light lift conditions.
b.)  Easy to build.
c.)  Outline to be as accurate as possible, with some minor changes allowed in order to improve the flight performance.

It was reasonably straightforward to make design decisions that satisfied requirements a.) and b.), a few calculations suggested that if the wing was enlarged very slightly then a wing loading of about 11oz / sq. ft. could be achieved, well within the range for a correctly-balanced PSS model to float in minimum lift conditions. The choice of aerofoil section was slightly more difficult, partly because the wing shape dictated a built-up construction, so a flat-bottom section was preferred. Although Clark ‘Y’ would have generated more lift, it was felt that S3021 would fly slightly faster, thus providing a more aerobatic performance in very light conditions.

Satisfying the outline requirement was more problematic; most drawings that were consulted disagreed with photographs in one or more areas, even allowing for known differences between airframes that were built at different factories (the canopy shape, for example). In the end, most discrepancies were resolved by the purchase of a 1/48 scale Tamiya plastic kit. A few minor changes were made to improve the flight performance and simplify construction, i.e. the oil cooler is slightly reduced in depth and the fuselage section is simplified near the fin post.

Experience suggests that assembly will be quicker if a kit of parts is produced first, and in any case it’s always best to get the less interesting stuff out of the way when enthusiasm for a project is at its peak! Reasonably light wood should be used throughout, the exception being areas like the wing l.e., which should be quite hard to protect the aerofoil section from the inevitable knocks and scrapes. The model has been designed so that cyano’ can be used almost everywhere, the exceptions being areas that will expose a glue joint (e.g. nose laminations, fuselage top sheeting etc.). Using cyano’ for these joints will produce a ridge under the covering because it’s so much harder than the surrounding balsa; the best stuff to use here is balsa cement or a sandable PVA.

Work begins by preparing the main longerons, made from cypress. Mark out the former positions and glue on small pieces of balsa (e.g. 1/8” sq.) to act as registers for the formers. The easiest way to do this is to tape the longerons together with some 1/16” packing pieces between and glue the register pieces across both of them to ensure that they’re identical, then separate them again. If you can’t get cyparis then use basswood, obechi or spruce, laminated with balsa.

To minimise the nose weight required for balancing, it’s very important to keep the tail surfaces as light as possible. The tailplane and elevators are from light 1/4” balsa, the tips from two layers of 1/8” balsa laminated on 1/64” or 1/32” ply cores and a hardwood spar to protect against breakage. The t.e. reinforcement can be left off but, of course, the t.e. will be more fragile as a result.

The fin is built around a 1/16” balsa frame to which the fin ribs are added followed by the fin post, 1/16” sheeting and the l.e. (after the balsa frame is built over the plan, the rest of it can be built in the hand). The fin post is from 1/4” x 1/8” cyparis (or similar hardwood) with 1/8” sheet balsa strips on each side, shaped to the profile shown. Do use cyparis if you can get it because it tends to spring and bend rather than break, and produces a very tough airframe. The rudder is separate but non-functional and is covered in film rather than balsa. It’s built over the plan using a 1/16” sheet core with half-ribs each side, cut as 1/16” rectangular blanks and then sanded to an aerofoil section after assembly. If you wish you can omit the hardwood (e.g. lime) t.e., but if you do, be sure to soak some thin cyano’ into the edge instead. Note that the top and bottom tip blocks are a balsa / ply / balsa laminate. Finally, and before we move on to the wing, it’s worth cutting the dorsal strake, though it won’t be finally shaped until the fuselage is built and ready for covering.

The wing is easy to build, and goes together quite quickly. It’s based around a one-piece slotted spar on which the entire panel can be built, one half at a time, rather like the late-lamented Balsacraft model range. After building one half, the bottom skin is simply creased in the middle before building the second. If you can’t get any sufficiently long balsa sheet to construct the skins then splice extensions onto 36” sheet to make the required length. Alternatively, each wing half can be built on separate bottom skins, accepting that there will, of course, be some loss of strength across the centre joint unless reinforced with a small amount of glass tape / bandage and epoxy / PVA.
Join the sheet for making the wing skins by truing up the edges, taping them together on the outside faces, gluing together (don’t use cyano’ for this!) and pinning flat. When the skins are dry, unpin them and sand the outside surfaces smooth on a flat surface with a big sanding block - trying to sand them later when they’re attached to the wing won’t produce a flat surface because the skin will bend away from the sanding block between the ribs.

The wing construction proper begins by constructing the main spar over the plan, marking its position on the bottom skin and then gluing it to one side only (left or right) of the bottom skin. The ribs are then added (gluing them to the skin only behind the main spar at this stage), the sub-spar fitted (but not yet glued), followed by the aileron spar. Now glue down the ribs forward of the spar and add the false l.e., which should be pre-chamfered on its lower surface (as should the sub-spar). Add the top surface spar and other details, then lift and support the half-finished wing so that the spar contacts the other half of the wing skin. The second wing half is then built in a similar manner to the first. When the joints are dry the sub-spar is glued in place and the top wing structure sanded smooth in preparation for the top skin.

Before the top skin is added, make a small washout wedge to fit under each aileron spar. The wedge should taper from 1/16” deep at the tip to zero at the root, and is quite easy to cut from 1/8” sheet with a straight edge, provided many small cuts are taken. If the wing is only supported at the tip the aileron spar will bow into a curve instead of being straight from root to tip - known from experience! Packing pieces should also be inserted every inch or so under the l.e. to prevent distortion when the top skin is added. Once each wing half is pinned down and packed appropriately, the top skin is applied using thick cyano’ or PVA; make sure the skin contacts the structure closely, and check that the t.e. tip and root don’t lift from the board when curving the skin down towards the leading edge. After the wing is skinned the (hard) 1/4” l.e. is added, followed by the balsa / ply / balsa tip block laminates. The model will fly better if you take some care over shaping the l.e. accurately for the S3021 aerofoil section.

The fuselage is a lot simpler than it looks and is built around the battery box, to which the front formers are added followed by the ‘V’-shape longeron assembly and the rest of the structure. The first job is to build the longeron assembly by gluing the longerons to F7, which will splay them at the correct angle. Whilst that’s drying the battery box is built using four rectangles of 1/8” sheet, to which formers F1 and F2 are fixed. N1, N2 and N5 are then added, followed by the longeron assembly. This might sound a bit hit-and-miss, but the whole assembly can be aligned easily by eye. The remaining formers (F3, F4, F5 and F6) are then added, followed by the top 1/8” square stringers, the tailplane support (take care to ensure that the tailplane support is parallel with the longerons), the wing seat doublers and the cockpit floor. Don’t forget the braces underneath the cockpit floor, otherwise it will probably bend when the fuselage sides are added.

The whole structure is then sanded smooth (paying particular attention to N1, N2, the wing seat doublers and edges of the cockpit floor) before the 3/32” fuselage sides are added; note that the sides should be cut oversize to allow for the fuselage curvature. They can be persuaded into the correct shape by gluing the top and bottom halves together at the longeron and then dampening the outside surface before taping in place; when dry it will be quite easy to attach them to the formers with cyano. A couple of words of caution here - don’t make the sides too wet, and don’t tape anywhere other than the former positions otherwise you’ll be faced with the classic ‘starved horse syndrome’, since balsa shrinks when it dries. If this happens, the situation can be rescued by adding hard 1/8” balsa strips to the inside edges of the skins between the formers. The elevator snake and receiver aerial tube are then added, followed by the wing bolt plate and the remaining stiffeners, triangular section etc. at the skin edges.

Sand the fuselage top flat from F1 - F2, F2 - F3, F4 - F5, and similarly the bottom from F1 - F2 and F5 - F7. Add tops and bottoms, engine fairing blocks, spinner and the tailplane-fin fillet. Drill the hole for the wing bolt and fix the wing in position, ensuring that it’s straight and square with the tail surfaces - if not, now’s the time to adjust it! When the wing is perfectly aligned, unbolt it, protect the wing with tape and bolt it back in position, trapping the 1/64” ply wing root fillet bases between wing and fuselage (some mild persuasion with hot water may be necessary to get the fillet bases around the wing l.e.). Tape the fillet bases to the wing, unbolt the wing, add glue in order to fix the fillet bases to the fuselage and bolt the wing in place again until dry. Add light balsa fillets to complete, then sand to shape when dry. Bolt the wing in position again, build the oil cooler and tack-glue it in place. Add the under-wing fuselage fairing (tack-glued at this stage) and ply facings at the wing trailing edge, then carve and sand everything to shape. Finally, remove the oil cooler, which will be re-attached permanently after covering and / or painting.

Anyone contemplating a highly polished all-over silver airframe might like to know that the wing of the full-size was not actually natural metal; in order to preserve laminar flow over the wing surface it was filled, primed and then sprayed with silver lacquer; P-51 enthusiasts might therefore care to consider using different makes of film for the fuselage and wing. The markings shown on the plan are for an early-production aircraft just out of the factory, without the dorsal fin fillet fitted - most P-51Ds had the fillet retro-fitted. A specially-moulded canopy is available from Vortex Plastics, and 1/12 scale pilot figures can be purchased from Pete’s Pilots. The exhausts are from scrap balsa and 3/16” diameter aluminium tube, the correct angle being set easily by inserting a piece of 1/8” dowel in the tube to position it before applying cyano. As for the spinner? This is from laminations of thick balsa on a thin ply base, carved and sanded to shape, checking against a simple card template for the correct profile. Before painting, the spinner can either be covered in several layers of medium-weight glass cloth and acrylic varnish (in which case make the balsa shape slightly smaller all round), or soaked in cyano’ and primed.

All-up weight of the prototype was 23oz, which included 2oz of ballast in the nose, added following initial test flights. This gives a wing loading of just under 12oz / sq ft; you might be able to better this if you build light, as less nose ballast will be required - a saving of only 1oz will reduce the wing loading to just over 11oz / sq. ft. However, it flies very well as it is, so perhaps the performance gain would be negligible.

Launching the P-51 is simplicity itself, just push it firmly forwards and it will soar away without any difficulty. Flight performance is very good for a PSS model, it penetrates well in strong winds, and can fly very fast with an unexpectedly high zoom climb from a fast diving entry. The prototype’s been flown so fast that aileron flutter has occurred, so don’t skimp on the material for the torque rods; they need to be stiff. And don’t just run them in a groove in the balsa, they must be in firmly fixed tube bearings to avoid wearing the groove slack and inducing slop. The speed doesn’t decay as quickly as its light weight would indicate, so it seems to be a fairly low-drag airframe. On the other side of the coin it can be flown quite slowly when required, which is a great help when landing or scratching about in poor lift.

Aerobatics are straightforward as the model is quite responsive in pitch and roll, and ‘bank and yank’ turns are accomplished in a ridiculously small radius with full elevator movement. The control movements can, of course, be rated down to give more realistic roll and pitch rates. Before trying any serious manoeuvres, though, you should check out the stall behaviour at a sensible height. Getting the model to stall in normal flight is quite difficult; one wing usually drops first, but the stall is quite benign and recovery is quick - if the surfaces are centred you’ll regain control again almost immediately. Loops are easy; a fast entry will enable a really big, smooth manoeuvre to be completed. Don’t apply too much elevator, or the loop will tighten up and the model may flick out. Approaching the top, release the elevator smoothly and the Mustang will fly itself over; elevator control is then re-applied gently to recover at the starting altitude. Rolls are simply a matter of banging on as much aileron as you like to achieve the desired rate, with a smidgen of down elevator during the inverted phase to make them as near axial as possible. The model will roll from straight and level flight at quite low airspeeds, but getting the same roll rate as when flying faster will need more aileron deflection. Inverted flight needs surprisingly little down elevator and can be sustained better in good lift because the increased drag in this attitude tends to reduce the airspeed.

It’s difficult for designers to be impartial, so you won’t be surprised to read that we’re very pleased with the Mustang; it’s a bit of a pussycat to fly, it performs better than many sport models and can be trimmed to be as agile as you like. Hope you enjoy it, too!

Vortex Plastics, 73, Stonehill Avenue, Birstall, Leicester LE4 4JF.
Tel. 01162 207080. E-mail: vortex_vacforms@hotmail.com
Web: www.vortex-vacforms.co.uk

Name: P-51D Mustang
Aircraft type: PSS warbird
Wingspan: 40''
Wing section: Selig S3021
Wing area: 282 sq. in.
Fuselage length: 321/4''
Tailplane section:   Flat plate
All-up weight: 23oz
Wing loading: 12oz / sq. in.
Rec’d no. channels:  Two
Control functions: Aileron and elevator