This article may be copied and distributed freely with the following restrictions: No one may claim it as their own work, and no profit may be made off it without my knowledge and permission. Anyone wishing to use it as part of a Web Page may do so as long as the above restrictions are not violated and the restrictions are clearly stated in reference to them.

    Vacuum Forming is a process that has been used by modelers for many years to form unique parts for or even entire models.

    When I built my son's USS Wichita I started out using the standard plastic sheets to form the upper deck levels with their splinter shields. I first started to build the model as a Cleveland class Light Cruiser (the ship the kit originally came as), but the forming the multiple curved splinter shields slowed progress considerably. After finishing most of the first level deck, I decided to changed the initial planed configuration of a bow and a stern gun to dual stern guns. As this was my sons first R/C Combat ship I came to the conclusion that this arrangement would reduce the potential for him to get into trouble and definitely reduce the chances that he would ram another ship. Plus I have always found a stern gun to be much more useful than a bow gun. The Cleveland turrets were quite small to fit two guns into without making maintenance difficult. So I decided to finish the ship as the USS Wichita, a near sister class. The hull was almost identical, but she had three larger triple 8" gun turrets rather than the Cleveland's five 6" gun turrets. The larger turret would be easier to fit the guns in. Needless to say the rest of the superstructure differed greatly between the two classes.

    Not wanting to start over again on the Wichita's multiple (but different) curved splinter shields I decided to cut molds from plywood and vacuum form the deck levels. While contemplating this I also hit on a great idea for the whole superstructure construction. Rather than my standard foam deckhouses and plastic decks all glued together, why not make each piece easily removable? If I attached a couple of locating pins or dowels (more on this later) to the deck and then drilled matching holes in the foam and plastic decks I could stack everything like one of those baby stacking ring sets. With easily made deck and splinter shield levels I could just unstack the superstructure pieces and replace any damaged ones. This is not only quicker than repairing glued on plastic levels, but I could use a much thinner, and thus lighter plastic, as it no longer needed to be "bullet proof" (or at least as much so as I could make it). For the final design I used the flexible inner rod of a plastic R/C control rod instead of stiff wooden dowels. These are not only almost indestructible in this usage, but they also are designed to accept #2 screws (or threaded wire control rod) into the hollow centers. Thus the method of securing all the layers was automatically incorporated into the setup. Thus the decision to go to a vacuum formed deck made battle damage repair a simpler affair also.  Figure 0 shows the disassembled superstructure.

Figure 0

    The shorter two yellow rods are the ones that position and hold the superstructure.  The longer one is the foremast and also the receiver antenna wire support.  The antenna wire is feed through the center hole and then run back toward the stern, where it is secured.  The deckhouse in the bottom center of the picture is the upper most level.  You can see the dents left by the hold down screw washers.  I plan to add a piece of ply or plastic, cut to fit the top, to spread the load, and glue one of the screw heads into the bottom of the range finder that should be there (not yet fabricated).

    The two turrets on the right show the two piece construction used.  The main turret shown is for the stern guns, and simply has a flat piece glued to the bottom.  This turret slides down over the outside of the barbette and is held in place by friction.  The forward barbettes have a molded bottom that has a raised boss that fits inside the barbette, and are held in place with screws.  The secondary turret in the lower right shows this type of bottom.  The finished bottom piece has a raised rim which slips inside the turret top and is glued in place.  It also has a dimpled center that is drilled out to fit over the matching dowel glued in the main deck.

    The main deck deckhouses and the second level decks are permanently glued in place.  The stacks are PVC pipe with a flat piece of PVC (cut from a larger dia cap) glued across the bottom.  A screw through the center of this piece secures the stacks in place.  The bottom ring (also vacuum formed) slips down over the stack and is held in place by friction.

    Tall parts (relative the the width and depth) can be a problem to form.  As the hot plastic is drawn over the mold it tends to form a web at the corners and center of the sides.  This is because it tends to come down unevenly.  To fix this I wick liquid plastic model glue into the webs central seam.  After this as dried and the web is now effectively one piece, I trim the web down flush with the surface.  Because our models are semi-scale (at least my superstructures are) if the web shape does not allow me to trim it flush, it leave a little of it.  Figure 0A shows such a trimmed (not quite flush) web, just little to left of the turret center.

Figure 0A

Figure 0B

    Figure 0B shows the patterns used for the parts in Figures 0 and 0A above.  The two boards with the holes in the upper left are the patterns for the secondary turret bases, with the turret patterns next to them.  The center pattern at the top is the main turret for the Wichita, and the one to the right is for the Louisville (Note the holes drilled through the patterns to allow vacuum to reach the centers of the broad surfaces).  The pattern in the center is for the rear superstructure.  The sunken sections on either end are for the molded in cutting guides.  Both levels of the deckhouses for this area are permanently installed and the cutouts go down between the foreward and stern deckhouses in the top level.  If you look closely you can see the series of holes around the edges of the sunken sections to allow the vacuum to get into the corners.  The dimples in the two patterns at the bottom and the one in the middle right form the locating markers for drilling the locating holes for the forward plastic hold down rods and the mast.  They are made by just kissing the surface with the point of a drill and then drilling a small hole for the vacuum in the center all the way through the pattern.  The pattern at the lower right shows the top of the gun splinter shield in the figure at the end of the article.

    One down side to all this was, of course, that while making the molds was much faster than cutting, forming, and gluing together all the decks and shields, I had no vacuum forming machine!! Building or buying one became necessary. After looking at the +$100 price of even a small machine that was basically a table with holes (you had to supply the heater and vacuum source), I decided to go the building route. I had, a few years before, purchased a book on building, relatively, simple vacuum forming machines (actually a general guide with several specific examples), out of general curiosity. I dug out the book and reread it.

    The machine I built is based on those general guidelines. I did have the advantage of already having a decently powerful shop vac to use as the main suction source. In addition I needed a heat source to soften the plastic in preparation to forming it over the mold. Trips to a couple yard sales resulted in the purchase of an ugly, but serviceable, toaster oven. I did not think my wife would appreciate me using the kitchen oven (especially should I accidentally get molten plastic in it!) and this allowed me to setup everything in my workshop.

    The size of the machine was dictated by several criteria. The first was the size of the available plastic sheets at the local hobby shop. The size of the frame used to hold the sheet is large enough so that 1/4" on all four sides could be gripped by it, but no smaller. The molding table has to be small enough for the opening in the holding frame to just clear. These dimensions ended up being: 5 ¼" by 10 7/8" for the frame opening and 5" by 10 3/8" for the table top. Thin rectangular cross section molding was used for the sides of the molding box (stood on edge). The thickness of these subtracted from the table size dictates the maximum size of the grid of holes. The border of the table was fitted with a high temperature RTV gasket for the hot plastic to seal against as the air underneath is evacuated. This seal sits on top of the table directly over the spot occupied by the molding strips underneath.

    There are two general sources of vacuum commonly used for vacuum forming plastic sheets. The main is a source such as a vacuum cleaner, or one built from just the operating "guts" used in them. These are good for many uses, but to get crisp edges and surface detail, a source of higher vacuum is needed. For the pieces I made for the original Wichita just the shop vac worked. The edges where the plastic met the table were rounded, not tight, but after trimming did not look to bad. For my first set these were usable, but I plan to add another source later. The pictures of the machine show my first , failed, attempt at a second vacuum source. I have started on a better source and will describe it later.

    The machine itself is shown in Figures 1 and 2. The large gray device on the side is a failed attempt to use a (no longer made) cheap toy vacuum forming machine's vacuum source to supplement the shop vac. The frame used to hold the plastic sheet during heating and molding is sitting on top of the machine in Figure 1. The machine itself is just a box the top of which is the molding table, and a lower skirt to hold the vacuum inlet pieces. The plywood top of the skirt forms the bottom of the box. The skirt sides are standard dimensional lumber. The box sides are cut from clear pine rectangular cross section molding, and the top is a piece of 3/8" plywood. The holes in the top are 3/32" set on ¼" centers.

Figure 1

Figure 2

    If you are going to use just one vacuum source the plumbing required is simply an adapter between the source and the box. In the pictures of mine you will see a black piece going into the bottom of the box. This is a modified sump pump check valve. It closes when the box internal pressure approaches that in the vacuum cleaner inlet. This happens when the vacuum has lowered the pressure as much as it can. With the valve closed a second higher source of vacuum can be used. Without this valve the second source would simply be trying to suck air out of the hose, which means it would just pull more air through the vacuum, and not lower the pressure any. When used on a sump pump the check valve has an internal rubber flapper that falls down (via gravity) when the water flow from the pump stops and prevents any water still in the pipes from flowing back into the sump. In this usage the threaded end is on the bottom and screws into the pump housing. In the vacuum former usage the valve is inverted so that the piping can enter from the bottom. A tension type spring is used to pull the flapper up when the pressure from the vacuum and the internal box pressure match. To install the spring a pilot hole is drilled in the hard plastic center of the flapper (from the threaded end) and a small eyebolt is screwed in. Once the piping is installed a dowel is placed across the opening (set in notches cut in the retaining nut) and the spring is attached between the eyebolt and the dowel.

Figure 3

    The vacuum cleaner inlet is built from PVC pipe sections, plus the modified valve (Figure 3). I turned a PVC coupling so that the inside of the end of the vacuum cleaner hose was a nice tight fit. I also turned another coupling to fit snugly over the normal outlet of the check valve. This later could be dispensed with by using the standard rubber hose and hose clamps to tie the valve and a PVC outlet pipe together in the normal sump pump usage. In this case the vacuum cleaner end of the inlet will have to be secured to the skirt to prevent twisting when attaching the hose. All the PVC joints were glued using PVC primer and glue. As it comes the pump end of the check valve is threaded to mate with the pump housing. This is a standard PVC thread and I cut down a matching coupling to be used as a nut to secure it to the bottom of the forming box (plywood skirt top). This nut is long enough to be project past the end of the check valve threads, with enough extra length for a notch to fit the dowel used to support one end of the spring which holds the valve closed. Make the nut as short as possible as its' height is the main factor in the height of the forming box. The internal volume of the box should be as small as possible so the vacuum cleaner has less air to suck out. The smaller the volume the less time required to empty it.

    The assembly of the machine started with the skirt and its' plywood top. The pipe, elbow, and check valve were glued together and installed. The check valve was then sealed with RTV where it protruded through the plywood skirt top and secured with the cut down nut. After the RTV has set for 24 hours the vacuum cleaner inlet adapter is glued on. The adapter should fit flat against the surface of the skirt, so that the pipe will not be pushed into the skirt when the hose is attached. The spring and dowel are now installed in the check valve. This completes the skirt assembly, if only one vacuum source is used.

    If a second vacuum source is used, or planed, an inlet nipple for it should be installed before the forming table box is installed. For mine I used my lathe to drill an axial hole down the center of a standard bolt, and then turned the end down for a tight slip fit for the inlet hose (Figure 4). The bolt was installed and secured with a standard nut. The bolt was then sealed with RTV. For those without a lathe the central hole would be a little tricky to make. Look around for a standard nipple you could adapt to this use. If you do not yet have the second source, block the second inlet for now.

    One note on the spring: You may need to try a couple different springs. You may also have to trim the length of the spring. What is needed is for the flapper to open when the vacuum cleaner is turned on, but still close completely when no air flow is present. This is, of course, with the valve inverted.

Figure 4 - The inlet for the second vacuum source.

    The forming table was constructed with rectangular pine molding strips and a 3/8" model aircraft plywood top. Any sufficiently stiff and smooth heat resistant material can be used instead of the ply. The table was cut out and the hole locations marked by drawing a grid on the bottom. Where the lines crossed the holes were drilled. The holes where drilled with a 3/32 bit. A sacrificial board was clamped on the drill press table and the forming table was set on this with the bottom up. This way any splinters the drill may raise when it enters the wood will be broken off of the bottom of the table. With the top held against the sacrificial board the drill will cleanly exit the ply and enter the board without risk of breaking a splinter out. The top of the table needs to be as smooth as possible for best results.

    With the many holes drilled I cleaned up any loose splinters on the bottom and cleaned any debris from the holes. The top was lightly sanded and the holes cleaned out again.

    With the table drilled the pine molding was cut and glued to the bottom of the table. Take care to make tight joints at all the mating surfaces. With the glue dry a bead of RTV was run at all the seams to seal any gaps. After this set overnight, a generous bead of RTV was run along the bottom of the molding and the table was pressed onto the skirt, no other fasteners were used. This allows me to remove the table assembly in the future if needed. Finally the excess RTV around the outside of the table/skirt joint was trimmed and another RTV bead run along the seam to insure a good seal.

    Originally the Vacuum Former was used without a seal around the top of the table, but I was not able to get a good consistent seal between it and the heated plastic sheet. To correct this I added the seal shown in the photographs. I used High Temperature Permetex gasket making RTV for the seal. A smooth nicely defined seal is required, not possible by just running a bead straight from the tube (at least not by me). So I built up dams on either side and laid the RTV down between them (as shown in the upper drawing in Figure 4A).

Figure 4A

    The dam was built up of basswood strips 1/4 inch thick and 5/8 inch wide. One side of the dam consisted of the strips, laid on their sides, cut to form a rectangle with the rim 1/4 inch in from the outside edge of the forming table. The other side was made from the strips laid vertically against the side the molding forming the sides of the table box. These were set so that the upper edge was on the same level as the tops of the pieces forming the inner rectangle. A straight edge laid across the inner rectangle was used in positioning the outer dam to the correct height. All the strips were secured with small brads pushed into predrilled holes a little smaller than the brad diameter. Once the dams were finished, the parts were marked and removed. A sheet of wax paper was laid on the table with generous amounts of overhang. The inner dam strips were then reinstalled on top of the paper. The wax paper was then folded back tightly over the strips and secured with tape. Similarly another sheet was positioned around the outer perimeter of the table and the outer dam strips reinstalled. After the wax paper had folded and secured over these edges, a channel all around the periphery of the table was formed with wax paper protected sides. The RTV, for the seal, was then run completely into this channel with a very generous bed left above the tops of the dams.

    Several days were allowed to insure that the RTV was completely set, then the dams and wax paper were removed. The dam strips were reinstalled and a sharp Xacto blade was run, horizontally, along the top of the dams to trim the excess off. By using the dam strips as a guide an even flat top was thus cut on the seal. I saved the strips in case I should need to redo the seal in the future. The bottom drawing in Figure 4A shows the completed seal. In fact there is one area where I did not lay down quite enough RTV and the top of the seal was below the dam edges and it is uneven there. I was pressed for time in finishing the ship, so I used it as is and made the deck levels I needed. Before I make some replacement sets I will replace the seal.

    A view of the frame used to hold the plastic during heating, is shown above.  The frame is made from 3/4 X 3/4 aluminum angle, with more angle cut down on one leg used for the clamps.  The cutoff leg was then layered onto the main frame bottom leg to thicken it for the clamp screw threads. The frame corners were formed by cutting the lower leg and bending the vertical leg at that point, using a vise to both hold the angle and also as a form to bend around, insuring a straight properly aligned bend. To cut the lower leg I first drilled a hole in the bottom leg at the bend location larger than the saw blade width and right in the corner of the angle. This prevents a crack from forming in this area when the bend is made. The lower leg was then cut with the cut finishing in the hole.

    At one end a tab was left on the vertical leg. When the frame bends were all made this tab was folded over onto the mating vertical leg and a rivet was used to hold the joint together. The assembly was laid on a flat surface (my drill press table) during the riveting to insure everything lined up. A flat piece of stainless steel was then riveted on for a handle. I used stainless simply because I had some scrap available.

    The frame was now placed on the toaster oven tray and three holes drilled through both the tray and bottom leg for supports for the frame during heating. Do this before installing the doubling strips below. Those strips will be the surface the supports rest on, and the holes in the lower leg will act as locating points.

    To make the clamping strips one leg was cut off a length of the remaining angle, leaving a stub about the thickness of the angle to act as the clamping edge. I used my metal cutting bandsaw, but a jigsaw or wood working band saw could have been used, with care taken. The aluminum is thin and a fairly with soft grade alloy. The cut edge was then smoothed on both the clamp and the cut off strip. The strip was then layered onto top of the lower leg of the frame to increase its' thickness in anticipation of threading the leg to hold the clamping screws. I secured the strip with epoxy, but it tends to melt slightly during heating. Regular pop rivets would have been a better choice.

    The clamp strips were positioned with the stub of the leg against the frame, and the holes for the clamps were drilled through both pieces with the proper size drill for the tap used. The clamp strips were removed and the holes enlarged for clearance with the screws. The holes in the frame were taped to accept the screws. The screw holes are positioned so that with all the screws in place the plastic sheet will just fit in place. The screws, thus, act to align the sheet during installation.

    In use the screws on the two long sides and on end are loosened and the screws and clamp on the other end removed. The plastic sheet is slid into place, the one clamp strip installed and all the screws tightened. The grip should be firm, but not so tight as to mark the plastic.

    A close up view of the frame and clamp is shown below.  On the left is the frame with its' doubled bottom leg and the right shows the clamp made from the cut down angle.

    Above is the toaster oven I use to heat the plastic.  The piece on top is the cooking tray that came with it.  You will notice the three long screws attached to the tray.  These fit matching holes in the frame and hold the frame off of the tray.  I have removed the glass front door originally attached to the tray for quicker access to the frame during the transfer to the vacuum former.  The plastic is clamped in the frame, placed in the preheated oven (set at about 350 deg.), and heated until it droops about 1 inch in the center.  For plastic 0.015" thick and above the heating element is on the top, and for 0.010" thick plastic it is on the bottom (it tends to burn through the thinner plastic if above it).

    Experiment with your oven to find the right settings for each thickness. For some of the thinner sheets I turn off the heating elements shortly after I insert the plastic and let the residual heat soften the sheet. Each oven will behave differently.

       To use the machine the oven is preheated and the frame with the sheet inserted. When the sheet droops down about 1 inch the vacuum is turned on and on and the plastic and frame removed from the oven and dropped onto the table and the assembly pressed down on the table seal. If a second vacuum source is used its' inlet valve is opened (or the source turned on) when the vacuum has exhausted as much air as it can.

    Do not let the vacuum run too long when hooked to the table. The vacuum motor is cooled by the air flow from the vacuum that is directed around it before it is exhausted to the outside. With it hooked to the table the overall airflow is restricted and the motor can overheat.

    For the secondary, or high, vacuum source I will use a compressor from an old refrigerator, and one of the portable air tanks sold at the automotive stores.  A picture of the compressor is shown below.

    Allow the plastic to cool and then remove it and the frame from the table. The pattern will more than likely stay inside the plastic. Each pattern I use has a hole for a woodscrew in the bottom. I install the screw and use it as a handle to remove the pattern. Use care the plastic is not that strong and can easily be damaged.

    Above is a picture of one of the patterns I made for the Wichita's superstructure.  It is for a gun tub splinter shield.  The main body is made from 3/8" plywood.  The 3/8" plywood is the same thickness as the desired height of the finish splinter shield.  A copy of the footprint for the tub from the plans was glued on and the ply cut to the outline.  NOTE: The pattern will be placed on the table with what will be the bottom of the finished piece up!  This is a view of the bottom (table side) of the pattern.  The 1/16" balsa pieces serve two purposes. The first is to provide vents so that the vacuum from the area directly under the pattern can reach the underside of the softened plastic sheet, to pull it down around the pattern.  The second is to raise the pattern up so that the molded piece will have a little excess available to trim the plastic down to the final height, and thus remove the rounded fillet where the corner between the pattern and table was.  The shop vac alone does not have enough suction to produce a crisp edge at this junction.  Once I install a good secondary vacuum source that can pull the plastic down all the way, I will remove the balsa and cut shallow groves out to the edge.
    When making a pattern one important consideration must be taken into account.  That is the pattern must be slightly smaller (or at a minimum the same size) at the top, in relation to its' final position on the table.  this is so the pattern can be removed from the molded piece.  If the top is larger the table side of the molded sheet will be to narrow for the pattern to be removed!

    This is a picture of the completed forward superstructure.  The three upper levels are the stacked foam deckhouse and vacuum formed deck levels.  The two screws at the top hold the sandwich in place.  They are screwed into the inner sleeve from a flexible control rod assembly, that is also attached by screws to the underside of the permanent superstructure.  Another piece of this inner part is used as the mast on the left of the superstructure.  The hollow interior of the mast acts as a support for the  receiver antenna, which is fed through it.
    The vacuum formed deck levels have dimples formed where the securing and mast/antenna rods will eventually pass through them.  these are drilled out before the parts are installed.  These preformed dimples insure that all the parts will line up every time.
    The dowels locate and hold the vacuum formed secondary 5" mounts.  These are going to be modified so that the mounts will be screwed or otherwise held in place.  Presently the secondaries just slide down over the dowels and tend to fall off easily.  The mounts are made in two pieces the mount itself and a floor with the locating dimple.  These are cut from the plastic sheets (with a little left around the mating edges), the locating hole in the floor drilled, the two parts glued together, and then the excess trimmed.

Return to Construction Articles