A POLISH MAYFLY

POST-WAR PROTOTYPE TRAINER CSS-10A

in 1:50 scale
by "Model Cardboard" (Models by Marek)

by William H. Geoghegan

Photographs by the author




CSS-10A from right rear CSS-10A from front right CSS-10A from top left front

HISTORY

Shortly after World War II ended, an effort began in Poland to rebuild that nation's aircraft industry, which had been left a shambles by seven years of conflict and occupation. In Warsaw, Franciszek Misztal, an aeronautical engineer who was pivotal in several PZL designs before the war, founded the Centralne Studium Samolotowe, or "Central Aircraft Institute," which immediately initiated a number of design and construction projects for civilian training and aerobatic aircraft. The CSS enjoyed only a short existence, however. It was disbanded at Russian insistence in 1950, since the USSR had by that time decided that Poland should abandon pursuit of its own aircraft designs, and instead purchase training aircraft (Zlins) from Czechoslovakia, and build under license aircraft designed elsewhere in the Soviet bloc.

Despite its short lifespan, the Central Aircraft Institute (CSS) was very productive. The CSS team designed, built and flew several prototype aircraft, including the subject of this model, and produced license-built Po-2's at Russian insistence. It also laid the groundwork for the eventual re-establishment of the Polish Aircraft Factory (PZL) and an Aircraft Institute that later served as the principal research and development facility for a reborn Polish aircraft industry.

The CSS-10A was designed as a two-seat tandem trainer. The student pilot was seated forward, as usual, and the instructor was seated in a slightly elevated aft cockpit that provided a clear field of view over the student's head. The plane was built in 1950, but was not actually flown until October 1951, about a year after CSS had been disbanded. Although opinions of the plane were positive, it never went into production. Like its successor, the CSS-11 acrobatic aircraft, it fell victim to Russian insistence that Poland purchase its light aircraft from other Soviet bloc countries rather than produce its own. In the end, the CSS-10A flew only 53 times, with total flight time amounting to only 28 hours and 28 minutes.

CHARACTERISTICS


THE MODEL

The model is distributed in electronic format on a CD-ROM that contains four other civilian aircraft of Polish and Czech origin. The planes in this set are not widely known outside of Europe (nor even within their home country as it turns out), and they are by no means generally available as models in any medium. The subject of this review could have been used as the subject of a "Mystery Aircraft" contest with almost no chance of correct identification.

CSS-10A Parts Sheets

The model is formatted for printing on two sheets of A4 cardstock, though legal sized stock could easily be substituted. I managed to print it on standard 8½ by 11 inch paper by printing two copies of each sheet, with the second of each pair printed after rotating the image 180°. The first sheet contains drawings of the plane (see illustration), basic construction diagrams, and most of the internal framing and formers that need to be backed on heavy card stock. The second sheet contains most of the externally visible components, including a number that also need to be backed on heavy stock.

The design of the kit is excellent, with very finely drawn panel lines and rivet detail. Overall, the level of detail is just about perfect for a kit of this size and scale. One nice touch that card modelers will certainly appreciate is the fact that the lines on which parts are to be cut are very thin. This helps to minimize cutting errors, and it goes a long way toward eliminating the heavy black lines that can appear where two edges join together. The fit of the parts is excellent, and I ran into no alignment problems with panel lines, lettering, etc. The only exceptions were in one or two places where fit was affected by the way I happened to shape a particular part; but these were easily remedied with minor trimming.

The fuselage is designed for butt joint construction (see construction photos later in this review) rather than the connecting strip method. The fuselage consists of five discrete segments, two of which contain cockpits. The gull wings require two panels on each side, joined in this case by connecting strips. Interior detail is minimal. Each of the two tandem cockpits is equipped with a simple instrument panel, a cockpit floor with simulated pedals, and a simple seat with printed harness. The instructions show no control column, although the location for attaching one is shown on the floor. Additional detailing might be possible, but the lack of technical information on the plane would make it speculative at best.

UIlustration of backing reinforced parts

The instructions call for the use of backing material between 0.5 and 0.8mm thick. It is important to stay within this range to avoid strength and fit problems later on. Despite the model's small size, the fit of the parts is very tight; and there are several places where parts are formed by sandwiching multiple layers of backed parts. For example, the wheels are formed of five such layers; and using backing that is too thick will make the tires appear oversized, and cause problems fitting the wheel skirts. If possible, stick with a strong cardboard approximately 0.5 or 0.6mm thick.

The two windscreens in the model have printed glazing. It would not be difficult to substitute very thin transparency material of some type; in fact, it greatly improves the appearance of the finished model. I also recommend printing at least one set of parts on good quality bond paper. Several of the components are too small to be formed from cardstock of the usual .0075" to .0090" thickness. Finally, print at least one extra set of parts for backup. You will make mistakes, and this will save you having to drop your work and run back to the computer to produce a replacement.

I need to make one important admission at the outset. I did not use my normal cardstock for this model, thinking instead to try an experiment with another type of paper that I had some success with in making small replacement components for another kit. The paper I used was Great White's "Matte Finish Imaging & Photo Paper." This is a 37 lb (138 g/m²) coated bond, approximately equal in weight to 60 lb index stock or 50 lb cover stock. It has a very bright white finish, and renders fine detail and subtle colors extremely well. It also has the advantage of holding its shape when rolled; and I thought it would make a good choice for this relatively small model with its subtle coloration and fine detail. Unfortunately, the paper was a little more fragile than I had anticipated; and the coated surface, while holding ink extremely well, picked up dirt just as easily. It was also nearly impossible to wipe off a glue smear without leaving a tell-tale trace. It was an interesting experiment, but in the future I'll stick with card stock for my models, with special papers like this one limited to individual components where it would make a positive difference.

CONSTRUCTION

The usual rule of thumb with card models is to let the part numbers be your guide to the construction sequence. I'm not sure why, but this almost always means building an aircraft fuselage from nose to tail, in that order, if the typical numbering is followed. My own preference is to begin near the middle of the fuselage, starting with the larger midship sections and the cockpit interior (if provided). This gives a larger base from which to install the nose detail, which almost always seems to require very careful shaping and fitting to look right.

In this case I began with sections 6 and 7, which together form the aft portion of the fuselage that begins immediately behind the instructor's cockpit. Although the instructions do not call for them, I used narrow connecting strips for sealing the longitudinal seam along the bottom of the fuselage sections. The strips were cut slightly shorter than the section covering to allow clearance for the bulkheads. Dryfit and adjust each assembly (covering and two bulkheads) before gluing.

Section 5 comes next. It not only contains the instructor's cockpit, but the forward part has reverse curves to provide contouring behind the headrest in the student's cockpit. Work the overall shape into the covering before cutting Partially assembled fuselage with subassemblies and components to be added out the cockpit opening. Cut out and install the cockpit seat and floor to the bulkhead. It would add a little to the 3-dimensional quality of the assembly to use a small scrap of cardboard to separate the seat back from the bulkhead, rather than gluing it flush to the bulkhead surface. I also used a short piece of brass wire, bent to the probably shape, and painted gray with a black simulated grip, to represent the control column. Glue the forward bulkhead in place; install the instrument panel (which helps to maintain the contouring in the forward portion); and then glue the rear bulkhead in place, being very careful with alignment. Section 4 (student cockpit) can be completed in the same fashion.

The above illustration shows the assembled sections 6 and 7, the partially complete section 5 (with the aft bulkhead and seat ready to install), the covering for section 4 (student cockpit), and the aft bulkhead and cockpit interior for the student position.

Fuselage assembly

Section 3 can be installed at this point. Note that it has a connecting strip at the forward end to accommodate the sharp taper in the forward cowling and radiator (parts 2 and 2a). Note that section 3 should be slightly larger than section 2 where they join. The assembly up to this point can be finished off with the installation of parts 2 and 2a (careful dryfitting is needed here), the rear cockpit fairing (part 9) and the black coaming that completes the two cockpit openings (parts 4f and 5d).

Rudder and Stabilizer

The rudder assembly is straightforward. Score the rudder at the leading edge and fold it around former K, which helps to insure the proper cross-section and positioning. The bottom of the rudder is contoured to match the cross section at the aft end of the fuselage assembly. Put the curvature into the lower rear section before gluing, and then apply a very thin bead of glue around most of the trailing edge, leaving the bottom unglued for the moment. The less glue the better; you want the trailing edge to preserve an angle. When the glue has set, gently form the bottom of the rudder around former K, forming a smooth, rounded surface at the joint. It can be done -- but have a sheet of spare parts handy. Glue the rudder in place, and attach the fairing that blends rudder to fuselage.

The instructions call for a small length of wire to position and reinforce the stabilizer. I used a length of .020" piano wire from the scrap box. Brass would work, but I would recommend something stiffer. Although the drawings don't show this, the wire has to go through the rudder fairing as well as the fuselage. You can make the necessary hole with a pin or needle, though I prefer using a fine drill bit in a pin vise. Install the wire and make sure it's perfectly horizontal before gluing it in place with slow-cure cyanoacrylate, if available. (Do NOT use fast cure CYA. It will saturate the paper and darken the colors noticeably.)

Stabilizer surfaces ready to be attached

Score and fold the stabilizer halves; glue the trailing edges with a very thin bead of glue, and attach them, using the outlines on the fuselage as a guide. It is not necessary to glue them to the wire. Then attach the stabilizer fairings, making sure that you make all the short lateral cuts that are indicated. These are essential if the fairing is to do its job.

By the way, at this point, you will have a number of exposed cut edges to deal with: the trailing edge of the rudder and stabilizer, and the trim tabs. Use your favorite corrective medium to fix them now, while the model is easy to handle. I use Faber-Castell water color pencils for this. They can be purchased in a wide range of fine color gradations, and they can be applied with a 0, 00, or 000 acrylic brush dipped in water and wiped over the pencil point. When the point's used up, sharpen it. You can purchase these singly in most art supply stores, and this allows you to buy only the colors you need for the models you intend to build.

Wings

The softly angled gull wing is by far the most challenging part of this model. I installed parts of it twice -- ripping out the root section once before I was satisfied with its position, and then ripping out and reinstalling the outer panels because they developed a nasty twist as a result of the gluing sequence I used. There are many ways to ruin the wing structure. I discovered a few of them; with luck, you won't have to discover the rest! And, by the way, the one real advantage of the otherwise disappointing coated paper I used became apparent at this point. Removing a glued part was easy; the glue attached to the coating, which came off when the part was removed, leaving a perfectly good gluing surface in the paper base. One mistake is okay; a second is fatal, however.

There are two ways to proceed: build the entire wing assembly (spar plus four panels) and attach the result to the fuselage; or glue the spar to the fuselage and then attach the wing panels. The latter is more difficult, but it will work. The former is easier, but it can lead to significant problems. The fit of the parts is so tight that the slightest error will render the assembled wing unusable. Rather than do that, install the spar assembly first, and then glue the panels in place, working from wing root to wing tip.

Build the main spar first (12A and 12B). Cut only enough of an opening in fuselage section 4 to accommodate the double thickness of this assembly. If you installed the floor in the student cockpit, then the center section of the main spar will have to be trimmed significantly to keep it from pushing the floor out of alignment. Make sure everything fits correctly before going any further. When it does, add the rib formers to the spar. The inner set will fit very tightly against the skin of sections 4 and 5. You might want to do some dryfitting here to make sure that they are not so tight as to deform the fuselage. When ready, glue the spar in place, being very careful to insure the proper horizontal alignment.

The leading edge of the wing panels should not be scored, but rather curved using decreasing diameters of wood dowels (or plastic tubing, etc.), attempting to match as closely as possible the shape of the wing ribs. Glue the connecting strip to the inner panel of each wing. Note that the inboard panels have integral fairings. These need to be scored where they will bend, and notched where indicated.

Wing root assembly in place; outer panels ready to attach

Install the wing root panels as follows: Shape and dryfit the panel until it fits snugly against the fuselage, the trailing edges mate exactly, and the connecting strip's midline is directly over the rib. When ready, shape the panel around the rib, and put a very thin bead of glue along the trailing edge only -- not on the ribs or anything that touches the fuselage -- and hold it in place until the glue sets. You can then apply sparing amounts of glue to the portion of the panel that attaches to the bottom of the fuselage (it abuts the opposie panel at the centerline); and then the fairing and the short curved section on the underside of the wing that joins the fuselage. You should need no glue on the top of the inner rib. Apply the glue to the fairing and use the rib as a template to make sure the panel is shaped properly. Once this is done, a small amount of glue can be applied around the edge of the outer rib to tack the panel in place. Repeat on the other side.

The outer panels should be done in a similar fashion, except that I would recommend creating narrow gluing strips from thin paper and positioning them about 0.5mm from the trailing edge. This will help maintain a realistic airfoil (see photo). Outer wing panel showing gluing strip Dryfit and make any adjustments that might be needed to insure a tight fit and proper alignment, and then glue only the trailing edge. Put the panel on a flat surface and use a book or some other flat weight to keep the joint perfectly straight as the glue dries. This will help prevent warping. When dry, you can slip the panel like a sleeve over the spar and connecting strip on the rib and inner panel, check alignment, and then apply a small amount of glue to the connecting strip. A small amount of glue can be used on the spar, but be very careful just to tack the wing covering in place. If you apply pressure, you can end up with the covering deformed and the spar showing through.

The most serious mistake I made when I tried assembling the wing for the first time was to begin by gluing the outer panel to the connecting strip that joins it to the inner panel. Even though it mated properly and the panel lines matched up precisely, by the time the trailing edge was glued together, the wing had developed a very serious twist. This was all due to a very small misalignment (a fraction of a millimeter) introduced when installing the inner panel. By the time that tiny error had propagated itself to the end of the wing, it had become a very big and extremely noticeable error. By creating the wing covering as a sleeve, it is possible to fit it onto the spar and connect it to the inner panel in a manner than insures proper positioning and alignment of the leading and trailing edge, even if this results in a minor mismatch of panel markings where the inner and outer panels join.

Remaining Steps

At this point, all that remains is the tail skid, landing gear, propeller and spinner, exhaust ports, and venturis for the instructor and student air speed indicators. None of these items is particularly complex, but they involve some minuscule components.

The landing gear is built as a sandwich of wheel disks glued directly onto a reinforced strut core. It would be a good idea to glue together the outer and inner parts of the Landing gear and tail skid assemblies and fairings sandwich separately first, and sand them to shape before gluing them to the center strut. Once shaped, paint them very dark gray or black, and then add the wheel spat (which covers only the inboard side), the outer cover of the strut, and the wheel spat fairing. The strut is longer than the covering (this may be intentional), and I found that I could get a stronger attachment to the wing by cutting a small slot to receive the strut at the point of attachment. It's probably a good idea to use slow-cure CYA adhesive for this. When dry, attach the upper fairing. The illustration provided here shows the landing gear before final touch-up and gluing to the wing.

The tail skid, also shown in the above illustration, is extremely delicate. Using a very sharp blade (a new #11 X-acto, or #11 scalpel), cut out the inside of the outside pieces first, then the outline of all three components. Once glued, the inside of the reinforced inner piece can be cut out, and the assembly mounted to the fuselage. Use CYA adhesive for this.

The exhaust ports are supposed to be 0.7mm diameter. I used very thin tracing "vellum" (.002" thick) to roll a thin tube around .020" piano wire. I painted the tube gunmetal black and then cut off a supply of short lengths (about 2mm) by rolling the knife blade over the paper and wire. These were glued in place where indicated on the nose covering.

The propeller is straightforward. Form and glue the spinner after cutting small holes to receive the wire on which the blades will be mounted. Insert the wire; add the blades; take care of any touch-up painting that's needed; and set it aside. Glue together parts 1a, and sand them to a sharp taper, to fit with Part 1. Then glue the propeller assembly in place on part 1, and glue the resulting component to the fuselage at the location marked. There are a few small cooling intakes that can be added at this time.

Install the windscreens next. Rather than a printed windscreen, I used a spare set of parts as a pattern to cut out thin glazing from some transparency material in the scrapbox. This will probably need to be trimmed to fit the curve of the fuselage on your example of the model. When you have the glazing right in terms of fit to the outline marked on the fuselage and correct profile angle, glue it in place bit by bit, starting from the center. Use slow-cure CYA or watch crystal cement for this. When dry, glue the printed framing to the front of the glazing.

The only items remaining at this point are the air speed indicator venturis. I found the printed parts too small to be usable, even with the relatively thin material I was working with. I tried carving a toothpick to shape without success (too narrow at the center); and I was unable to form any kind of a paper cone that would work (though I might have been successful had I kept at it longer). I finally settled on rolling a very narrow tube, cutting out a pair of very small diamond-shaped holes at the midpoint, and then crimping the tube to approximate a pair of cones joined point to point. I put the mounting bracket around the center and glued it in place to the fuselage. Surprisingly, the final result bears up under all but the very closest scrutiny.

A little final touch-up with watercolors, and correction of any alignment problems that may have crept in, and the model is done.

Conclusion

This was not a particularly easy model to build, not because of any inadequacy in the design, but rather because of the necessarily small size and fragility of many components, because of the gull wings and dual cockpits that had to be CSS-10A -- Side view from right modeled, and because of my stubborn insistence on using a construction material that was inappropriate for this type of project. I have built kits in smaller scales, some of which have had even more parts than the CSS-10A; but none seemed as challenging. In retrospect, much of the difficulty was due to the very precision of the kit's design. The fit is extremely precise and tight -- as it should be -- and the scale does not provide much tolerance for error. This is one of those kits that you want to build twice: the first time for practice, and the second time to "get it right." That's one of the advantages of kits in electronic format -- an unlimited supply of spare parts for correction of errors, or even for complete rebuilds.

Finished model from top left rear

Despite the problems I ran into (mostly of my own making, I have to admit), I am very pleased with the result. I like subjects that are historically unique or otherwise a little offbeat; and this one qualifies on both counts. This is the first design by Marek Pacynski that I have attempted; and I am thoroughly impressed by the quality of the design, the level of detail, and the overall packaging. The lack of written instructions turned out not to be a problem, since the diagrams provided almost all the information I needed. A little thought takes care of the rest.

Completed model from right front

As I said, this is the first of Marek Pacynski's models that I have tried. It will not be the last. There's already a PZL P-11c underway on the workbench (can't get away from those gull wings!), and more already printed and waiting to go. But the CSS-10A is still unique: a one-of-a-kind aircraft that symbolized Poland's effort to create a renaissance of its pre-war aircraft design and manufacturing capabilities, and that became emblematic of the difficulties the nation faced in doing so. The result was a beautiful little plane that had not much more than a mayfly's experience with flight -- 28 hours and 28 minutes in the air before it succumbed to dictated irrelevance.


Summary

Model: CSS-10A (prototype design by the Centralne Studium Samolotowe)
Kit: CSS-10A by Model Cardboard
Distributed by Models by Marek (DeWayne Barnett)
Designer: Marek Pacynski
Price: $10.00 (CD-ROM with four other models of the same period)
Scale: 1:50
Dimensions: Wingspan: 8½" (21.4cm)
Length: 5¾" (14.7cm)
Number of Parts: 80
Difficulty: 3 on a scale of 1 to 5
Time Required: About 15-20 hours
Instructions: None
Diagrams: Good: shows layout of all major parts and detail on critical subassemblies
Fit: Excellent
Coloring and Artwork: Excellent, though results will depend on printer and paper used
Printing: User determined
Resources: Extremely limited: one short reference (in Polish) on a Polish aviation history site at www.enter.net.pl/www/szybowce/histlotn.htm. Additional information was provided by the designer, with translation assistance was provided by Miroslaw ("Mike") Krol. My thanks to both of them.