Building the Rekin

 Building the Rekin



The background

Waclaw Czerwinski envisaged the PWS 102 Rekin (Shark) as an improvement on his very successful PWS 101. (PWS stands for Podlaska Wytwórnia Samolotóv Sp Ake, Podlasian Aeroplane Factory Co Ltd.) It was to be faster with a less cambered and thinner wing profile. Full-span flaps were combined with the ailerons. The wing planform conformed closely to the ideal elliptical shape. The gull dihedral was retained although this created some problems with hinging the flaps round the bend. The wing now was mounted higher on the fuselage because in circling flight some cross flows usually occur and it becomes important to minimise disturbances on the upper side of the wing in the central area. The cockpit canopy was now fully contoured but not moulded. To ensure good visibility in the forward direction, the shape was somewhat bulbous to minimise distortion.

The prototype Rekin, registered SP – 1126, flew first early in 1939 and proved successful. Two more, PWS 102 Bis, SP 1361 and (possibly) SP 1362  were built with minor improvements, such as increased aileron area, but all three were captured by the Red Army soon after the outbreak of WorldWar 2 when Germany and the USSR divided Poland between themselves. The later fate of these sailplanes is not known.


The prototype, SP- 1361 on first assembly outside the factory. Rigging is not fully completed


 The second prototype


The cockpit interior as it was planned but never finished. Key: A, ­Control column; B, Rudder pedals; C, Dolly wheel release;  E, Flap lever; F, Spoiler handle; G, Tow release strap; K, Trim wheel; L, Canopy latch; M, Pilot’s harness; H, Map pockets; J, Window

Figure 1. The only evidence I now have that my first model of the PWS 102 ever existed.


My first PWS 102

I built a quarter scale, 4.75 metre span model of the Rekin some ten or more years ago. It flew quite successfully, launching by winch and aero tow, and soaring well in thermals and slope lift.

This model was of orthodox construction but skinned entirely with 0.4 mm plywood except for the control surfaces. The thin ply is a little heavier than 1.5 mm medium grade balsa but stands up better to the rough and tumble of regular flying operations. The model weighed a little over 7 kg, mainly because the ply skin covered the whole  wing except for the control surfaces

I never really finished the model. After some flying I began to feel some dissatisfaction, not because there was anything very wrong with but it did not handle quite so sweetly in the air as my earlier PWS 101, and I was always busy with other models. I didn’t add the registration numbers and never bothered to put a pilot in the cockpit.

The main thing it needed was to make it easier to rig it for flying. I had used the traditional vertical steel strip wing joiners mounted in the wing spars, sliding into rectangular brass tubes in the fuselage. Slotting these into place was often somewhat difficult with a big model and when flying I thought there was a tendency for the wings to move slightly back and forth as the joiners flexed.

When there was a choice, I flew other models from my stable and the 102 became a hangar queen. I needed the space and after some years scrapped it, saving only a few bits for re-use.

When I made up my mind to build another Rekin ,  I decided to try some different construction methods, and in particular I completely re-designed the rigging system.

Easy rigging

Many full–sized wooden sailplanes in olden times took hours to rig with mutiple separate steel pins to be aligned and inserted with much wriggling and heaving, and each pin  had to be locked with its own ‘Fokker needle’, which was often a very fiddly business.

One of the most successful wooden sailplanes ever produced was the 18 metre span DFS Weihe (pronounced VYER to rhyme with fire). The Weihe wings can be rigged by two or three people in a few minutes. Figure 2 shows the general idea in sketch form.


Figure 2

To put the wings on, with the fuselage held in an upright position, one wing is presented to the fuselage fittings, front and rear, and a single lever pulled to insert two steel pins linked together and aligned on the same centre like a hinge. The wing tip can then be lowered to the ground. The other wing is presented in the same way and the appropriate lever pulled. Both wing tips can then rest on the ground. The pins are locked with a very simple over-centre arrangement of the levers. To complete the rig, the tips are lifted simultaneously to bring the main upper fittings into alignment and a third steel pin is inserted.


I adopted a similar system for the PWS 102. The accompanying diagram (Figure 3) shows the arrangement and photographs show how the wings were joined and mounted on the main fuselage frames. A long steel rod with a needle point to aid alignment, is pushed through both front and rear attachment points


The special fttings in 1.5 mm steel were made for me professionally but could be made from  scratch by anyone with some simple metal working equipment. When fully rigged the pins were prevented from creeping out of position by the wing root fairings.


Figure 3

The main spars

The traditional way of building a main spar for a model is to lay the lower spar flange on the plan, add the ribs and then the upper flange. The shear webbing comes next as a series of vertical inserts between the ribs. My rigging system required some precision in the alignment of the wing fittings. I needed to have both spars completed so that I could align the wing root joiners very exactly and check that they fitted, before attempting to add ribs, skin etc. With a gull wing it is also advantageous to build both left and right wing spars in the same jig, so that they will be identical. The bend of a gull wing is not a sharp angle. The spar is curved over a distance of three or four feet, and full-scale practice is to laminate the spar by steaming the timber to conform to this bend.

I made a jig for the spars from square aluminium tubing bought from the local hardware store. The plans for the wing were laid flat on the bench with a sheet of thin polythene (any other suitable material will do) to prevent the glue sticking to the drawing. The tubing, also protected with polythene, was fixed rigidly (drilled and nailed down) exactly to the outlines. Two rib bays where the gull bend occurs were left open to allow the wood to take its natural curve here. On the PWS the outer, elliptical section of the wing requires a change of taper which had to be observed when aligning the jig. Three separate sections of tubing were needed here for the upper spar flange while the lower flange remained straight. (Figure 4) The 1 mm thick plywood webbing, with the grain of the outer ply running diagonally, was laid flat into the channel between the tubes. This requires some accurate cutting of the plywood, and the inevitable joints in the web were scarfed. If the scarfing seems too difficult, a butt joint will do but after the laminations are glued in, the butt joints should be backed up with an overlapping insert, between the flanges. For all this work I used slow curing epoxy glue.

Figure 4. The spar laminations for the upper flange are shown here clamped against the aluminium tubing which forms the jig, nailed down to the bench top or building board. The spar web of 1 mm plywood is in place between the tubes. On the left are the lamination strips ready for the lower spar flange.

 Sufficient spar laminations of 3 mm (1/8th inch) thick spruce or pine were cut from stock pine or spruce strips, and tapered as necessary with a band saw. The timber for the spars was chosen with care. Short grain or other oddities in the wood must be avoided. Grain with closely packed growth rings is preferred. If strips of suitable length cannot be found, scarf joints become essential and it is wise to take special care over this. A good scarf joint should be as strong as the un-jointed timber. The ratio of a scarf joint to the thickness of the timber should about 1:12 or 1:15. For models we may not need to be quite so careful as full scale practice but these are standards to aim at.

To lighten the spars, the number of laminations was progressively reduced from the roots to the tips and the spars were further lightened by planing towards the tips. The laminations for one spar are glued to the webbing and to one another, weighted down and clamped to the aluminium and left for the glue to set.

Providing one edge of each lamination is perfectly smooth, the other edge need not be perfect. At a later stage both spars are planed together, removing saw marks and bringing them to the correct cross section dimensions and smoothness.


 Figure 5 The twin laminated spars assembled with the drawing used to set up the jigging.


Figure 6 and 7 show the steel fittings bolted to the spar roots where the laminations were solid, and Figure 8 shows the spars joined as they would be when  the model was rigged for flying. Figure 7 shows how the fuselage fittings were arranged to take the wings.


Figure 6. The root fittings bolted and epoxy glued to the main spars



 Figure 7 The main spars joined with the fuselage cross member in place.


Figure 8. The fuselage fittings corresponding to the spar fittings.

Building the wings

Building wings from the outer skin inwards is the modern method in full-sized and model sailplane manufacture. A very accurate female mould is made first, then the entire skin for one surface, of fibre reinforced plastic, is laid up, vacuum bagged and cured at high temperature in an autoclave. Spars and all internal fittings are fitted while the skin is still in the mould, after which the other surface skin, also pre-moulded and cooked, is glued on.


For the Rekin  I adopted a somewhat similar method, but with wooden skins and no autoclave. A female form for the underside was cut from plastic foam with the usual hot wire bow and templates. To accommodate the bend in the wings and the elliptical taper of the outer parts, the form had to be made in several sections. Accurate cutting was important but it was not necessary to finish and polish the surface as it would be for a plastic sailplane.

With a sheet of polstyrene to protect the form, the entire plywood skin for the underside,including the ailerons and flaps, marked with the position of all spars and ribs, was made up with all the necessary joints completed and laid into the form. The main spars then were glued and weighted down in position. (Figure 9)

Figure 9. The 0.4 mm plywood wing skins in place, joined and marked out for the addition of ribs and spars. The main spar is glued in position.

 Figure 10 The spars and all ribs fitted including the ailerons and flaps.

All the necessary ribs and other details and internal arrangements for servos and electrical connections to the fuselage were made (Figure 10), before laying the upper surface skin. At this stage the ailerons and flaps could be cut free. To lighten them and for the sake of appearance, the unwanted plywood was cut out leaving the control surfaces with plywood stiffening at all the joints, but preserving the translucent appearance of the eventual fabric covering (Figure 11).


Figure 10. A shaped balsa wood leading edge fairing was added after the wing skinning was complete. To accommodate the change of wing thickness at the root, the skin was laid in several chordwise strips.

The rest

The fuselage and tail unit were built in a very orthodox fashion which should not need detailed description. A check on the assembly of the wing fittings was made before gluing, as shown in Figure 11. The fuselage was built in two halves, the so called ‘crutch’ type of construction with half formers of plywood and 0.4mm plywood skins.  Skinning of the front where the curvature is in three dimensions, is shown in Figures 12  & 13. Assembly of the model, less the cockpit canopy, is shown in Figure 14, and some further photographs show the finished Rekin.


Figure 11 Trial assembly of the main fuselage formers

Figure 12. The ‘crutch method of assembling the fuselage. This shows the port side with the nose partly planked