AJ Madison teaches basic scratchbuilding techniques. .

Starship Modeler - The complete information source for modelers who build sci-fi, fantasy and real space subjects

Scratchbuilding 101

By A.J. Madison - © 1999

Scratch building is not a black art practiced by modelers with hyper-acute cases of "Advanced Modeler Syndrome" with way too much free time on their hands. Scratch building sounds scary, it's not.

A scratch build project will consist of several of the techniques listed below. To start, my advice is to just start carving up styrene. Picture some large sub-assembly you want to build and go ahead and start slicing plastic. You'll make mistakes; once I had to trash an entire Saturday's work because I didn't conceptualize a structure correctly. However, you'll get a real feel for how the model will go together. And when you complete your sub-assembly, you can begin to estimate how long and how hard it will be to complete the subject of interest. This might be important, because your gross estimate may come out in years. This is useful to know before you've invested 3 months on a project you won't or can't finish.

I strongly recommend some "photocopier" magic. Try and produce a set of plans that are the same scale as your intended subject. You can picture subassemblies in "real size" and you won't waste time with tedious numeric conversions that may turn out to be just wrong. Checking your work directly against the plans is the best insurance against building it wrong, especially when your goals include accuracy. (Or be the whole point of the exercise.)

Practice extra care measuring and cutting out parts (larger than 5mm, see the section on details). Especially with a scale model, inaccuracy in part shapes adds up (known as tolerance stacking), and if you have a complex & intricate design, you can reach a stage of assembly to discover that your "fusion thermocouple turbo-cannon laser" won't fit without cutting off half of your "wing assembly." I use a large collection of drafting tools, mostly dividers and the smallest drawing triangles. I try and interpolate between marks on the drafting scale when making or taking measurements as well. Yes, I'm suggesting you mark and cut in between the markings on the ruler. Do not forget to subtract the thickness of the plastic. Yes, it may only be twenty thousands of an inch for one panel, but a cube made of .020" sheet would be a total of forty thousandths out of dimension. You'll definitely be able to see that it's out of square. And in a hobby where millimeters make the difference between accurate and inaccurate, .040" is a millimeter off.

Scratch-building techniques are probably as numerous as the different subjects that are built from them. And if the first cut at building a sub-assembly either fails, or the technique applied appears overly difficult or labor intensive, re-think the application. If the solution is obvious to me, I go ahead and build it. If not, I work on something else or attempt to devise an alternative solution. Without philosophizing too much, there is an elegant solution to every construction problem; sometimes the most efficient answer requires an application of good old fashioned imagination.

But in short, I'd say there's approximately 4 different major variations of technique:


Kit-bashing is possibly the simplest and speediest construction method of the techniques. This involves finding the right or nearly right shape in existing styrene. Don't limit your search to just kits. Lots of packaging is perfectly good styrene.

An example of this involves a very specific rounded convex shape which did not match anything on hand. But I discovered that by cutting the bottom off of an Enterprise lower saucer section and gluing it to the top of the upper saucer section part (after the command decks section was removed) of the same movie Enterprise kit, I got the exact profile I wanted.

Also, Evergreen & Plastruct make a large variety of shapes. Tubes, miniature I-beams, miniature columns & channels, various rod and square stock. I built a pair of warp engines based around sections of Evergreen rod that happened to be the right diameter. Before resorting to the last two techniques below, see if the shape of interest can be built up from simpler elements. Further, subdivide your project into simpler elements. The complete fuselage itself may be very complex, but if the engine intakes are separate components, each sub-assembly may become relatively simple to construct (or find).

"Boxing" it in

Treadheads use this technique regularly, partly because simple techniques produce advanced results (no knock on their skills, hardly) and partly because the final product is straightforward. Sheet styrene is cut in the appropriate polygonal shapes and glued together. The problem with this technique is that the object of interest needs to be planar in form. Good for many AFVs, but useless if you want to scratch a "Globemaster II".

For beginners, I recommend making and attaching gussets (braces, usually triangular in shape, that reinforce the joints where two planes meet) prior to assembly. Initially, it looks like it takes one hand on each part and a third hand to apply the glue or solvent. Gussets eliminates one degree of freedom from alignment. With practice, one can obtain alignment from all of the parts as the build-up proceeds, though all of the parts need to be pre-cut, and assembly takes place before the joints fully cure.

A variation of this technique reduces to a simple lamination - stacking sheets on top of each other, then sanding or carving the exact shape you want. For example, say you wanted to replace warp nacelle support pylons on the Enterprise-A with curved leading and trailing edges. (ie, the pylon is still flat when viewed on edge, but its "elevation" view has the outside edges curved. Imagine that Federation Engineers discovered that pylon streamlining decreased warp-field turbulance). All of the warp pylon details can be laminated from plain and textured sheet (namely Evergreen grooved sheet styrene). To build this example, I would take some 60 thou sheet and cut it in desired shape. However, there are forward and aft intake vents on the pylon, and the center sheet would be notched, as appropriate, fore and aft to account for these intakes and exhausts. Next, I would apply grooved sheet to replicate the plasma flush vents on the sides of the pylon. Since the grooves would show on the lamination edges, I would cut the grooved sheet slightly wider than the final vent openings, then fill in around the textured parts with plain sheet. Finally, a layer of 10 thou sheet, with the appropriate holes cut in them for the flush vents is applied. The gentle fore and aft tapers (airfoil like) on the pylon cross section would be replicated by careful sanding.

The lamination technique can be elaborated upon as needed to create some very intricate and complex shapes. While it doesn't call for extra-ordinary construction technique, it does involve some imagination and a lot of planning.

"Wrap it Up"

I would call this die-cut Balsa wood model airplane technique one. Its a variation of "boxing it in" and is also known to ship modelers as "plank on frame". This technique is good for building the main sections of a true airfoil and (ocean going) ship hull shapes. Your shape has a curve in one axis, and not necessarily circular, but a shape that a sheet of styrene can assume without cracking, tearing, or buckling. You start by building a frame with some center spars (or keel) with shaped ribs at intervals either dictated by the real subject, or sufficient to keep the covering sheet of styrene from caving in. Depends upon your subject, but avoid ribs more than 1" (~2.5 cm) apart. When your frame is complete, wrap it in sheet styrene. 5 thou sheet is very flexible but you can easily sand through it. 10 thou sheet is thicker but not as flexible. An alternative solution is two layers of 5 thou, but care is needed gluing the layers together, lest the outer layer become chinked by the inner one. Provided the curve is gentle, you can use several layers to build up to the final shape.

My Akyazi Starship warp pylons (See it in the Dominion War Contest) consists of six layers of sheet styrene (three top, three bottom) attached to a curved frame to obtain the final desired shape. The three layers consist of 1. support layer for layer 2, 2. Texture layer for the plasma flush vents, 3. Final exterior skin.

I have also used this technique on a "Classic" or pre-refit Enterprise project in 1/640 scale. In this example, I cut the inaccurate hull rim off the primary hull lower part. I then produced a series of gussets with the correct rake angle that also held the primary hull upper and lower kit parts the correct distance apart. I sandwiched the gussets between the two primary hull parts. I then "skinned" the hull rim with 10 thou sheet, attaching the sheet to the gussets, then trimming the excess down to the respective top and bottom of the saucer upper and lower parts.

"A+B Epoxy" to the Rescue

Call this die-cut Balsa wood model airplane technique two. It is also the technique of last resort, because it can solve nearly any problem composed of compound curves. The first time I saw this technique was in a FineScale Modeler article on scratch-building a then unavailable 1/72 scale X-29. Once again, one builds a frame, composed of a series of ribs (and longerons if so desired or necessary). Not unlike the previous method, the frame roughs out the final shape. However, because of the compound curves, you can't just cover the object with sheet styrene. Instead, you use A+B Epoxy (if unavailable in your area, Milliput or plumber's waterproof "putty" epoxy will work, albeit at greater expense). In the X-29 article mentioned above, it was necessary to pre-fill the area between (inside) the ribs with balsa wood to reduce the weight of the model (consider 1/4" to 1/2" external layer of A+B sufficient). Once that's done, fill the ribs with A+B. Try to knead the bubbles and voids out of the A+B otherwise you'll need to putty those voids after you achieve the final shape. Once the A+B has cured (about 24 hours), start sanding with a low numbered grit sandpaper. 60 or 80 grit is fine until you've roughed out the general shape. Use progressively finer grits until the ribs start showing through the A+B. This sounds time-intensive, its not. I sanded out the final shape on one subject in a single evening. Admittedly, I spent a great deal of time later wet-sanding the project to a glass smooth finish that was also absolutely dead-on in profile, so you can put in as many hours as you want during this stage of the detailing.

This technique also works amazingly well with profile correction. I can wet-stand the A+B until it is actually translucent. It neither detached itself from the kit part, nor chipped away at the edge of the A+B where it feathered into the model. And A+B takes scribes well, though not as nicely as Evergreen sheet.

A necessary alternative to building a styrene framework is using some bulk material, especially when the subject is very large: 50cm and larger. A modeler showcased by FineScale in the mid to late 80's, did beautiful Sci-Fi ships using plexiglas for his armatures. Other alternatives include basswood, balsa, and pine, each requiring some wood working skills. Another alternative is the foam used to insulate basement exteriors. One skins these forms with materials like bondo, fiberglas, spackle, or liquid (bulk) epoxy.

Finally, a note about epoxy. An increasing number of people are showing up with epoxy intolerance and furthermore, they discover their chemical intolerance by waking up in the intensive care ward 2-3 days after their last contact with it. I strongly, strongly recommend never touching uncured epoxy directly; use a tool or gloves. Avoid breathing the fumes when mixing it and the dust when abrading it.

"The Details Have It"

In "Sci-Fi and Fantasy Models" Issue #29, Tim Ketzer describes his Millennium Falcon FX miniature replication project. He pointed out an important down side to scratch-building: the initial shapes built up rather quickly, followed by month after month of adding details. And because it's a scratch build, unless you're kit-bashing, you can't reach into the kit box for detail parts and apply them, especially when its an accurization project.

Admittedly, the second trilogy Star Wars FX miniatures are greeblie intense, so not all projects will have discouragingly long periods of slow progress, but the details will make or break your project. And not necessarily in terms of being a contest winner, but rather in terms of completion. Nor is my point to discourage. Super-detailing an injection molded aircraft kit cockpit may take a couple of weeks, perhaps a month depending upon the level of detail and scale. This is not that painful when the last step to the sub-assembly is gluing the two fuselage halves together and you're done. BUT if your last step includes creating the other hull half, the project may take a very, very long time. Be creative with details, both in their scope and construction. Avoid Acute Advanced Modeler's Syndrome.

For example, think about replacing the photorp launcher part on the Enterprise-A. I have references that show details INSIDE the tubes. Consider that at 1/535 scale the tubes themselves are approximately 1 millimeter high and 2 mm wide. Just replicating the rectangular box with a hollow is a challenge, never mind details that require a jeweler's loupe just to see. For my launch tubes I used 60 thou sheet cut into small stubs and a no. 11 blade to dig out some of the material in the end of the stub, leaving a jagged hollow. A drop of Tenax on the jags mildly dissolved the styrene, leaving a smooth, rectangular hollow. Alas, it does not reach the bottom of the stub, but after some black paint, and the red outline, no one will notice.

Similarly, use your imagination and creativity on those details you've decided must be replicated. On my Akyazi project, I would estimate about half of the details under 25 square-mm I had scribed out or made with strip styrene should have been represented with paint, technical pen, or decal sheet. That would have eliminated a significant amount of construction there. And when committed to constructing greeblies, anything with dimensions under 5mm, I just chop out freehand from scrap, albeit the "correct" thickness sheet. Any gaps or lack of orthogonality is hidden after applying Tenax or super-glue. This is how I constructed the previously mentioned Enterprise-A photorp launcher ribs. A little sanding and filing when you're done to line up the exterior edges, and it will be a gem of a part, and other modelers will think you're a scratch-building god or goddess.


A scratch build project will consist of several of the techniques listed above. The main things to remember before embarking on such a project are:

  • Divide and conquer sub-assemblies.
  • Plan before building.
  • Contemplate and imagine alternatives when faced with a very difficult solution to a construction problem.
  • Take care cutting out parts, BUT you can impress yourself with what you can freehand and still have it come out looking perfect. So don't get hung up on small details, if you can build the major shapes the details are usually easy to coax in later.

Go back up | Tech Library Main Page | Starship Modeler Home | Site Map | Feedback

This page copyright © 1997-9 Starship Modeler™.
Last updated on 2 September 1999.