CerroBend Casting by Bill Coffey. Reprinted by permission of Benchmark Publications, Ltd.


There are two reasons for making castings in your own shop First, when you need a part in large numbers and it's not available commercially, casting your own is the only alternative to making it one by one. And, second, there is the pride and pleasure to be obtained from doing your own castings. Sometimes modelers assume that they can learn the technique of casting, buy one commercial part, and thereafter produce as many as they want. This is neither feasible or practical, and it's certainly not ethical. Commercial parts are made in high volume and arc therefore relatively cheap, much cheaper than you could achieve in your home workshop considering the time and materials required. I heartily discourage copying commercial parts. Save your casting efforts for parts which are not available commercially, castings from castings will not be accurate. Or make your own patterns. When a number of like parts are required, it is much simpler and more accurate to spend your time making a master pattern of your own than to get involved in the questionable procedure of attempting to copy someone else's casting. Casting your own is most rewarding when you need parts that can't be obtained any other way. If there is a group of modelers in your area that can use the same parts, you can share the expense and work and bring down the cost per part.

Preparing the Master Plan

The first and most important step in the entire casting process is the fabrication of a master pattern suitable for casting. The master pattern should be a "master" in all respects. Consider from the start what kind of casting process it will be used in. Is the master to be used in a vulcanized rubber mold for the manufacture of wax parts for investment casting? Eventually these masters will be duplicated as brass parts from the lost wax investment process. Is the master to be used with silicon rubber molds? These masters will be duplicated in a lower melting point metal alloy. Is the master to be used as an open-faced mold pattern for water putty parts? Once the use planned for the master has been decided, you can select its material.

Casting Methods

Casting can be done by several methods. I will discuss three methods but will give most detail on one. Each of these three methods requires a different type of mold, therefore, the casting technique must be determined prior to mold making. Probably the simplest and least effective of the three is open-faced casting which can, but seldom does, give satisfactory parts. In this method, a single mold is made with the top open for pouring the metal directly into the cavity. You cannot be assured of a complete fill and one side has no detail. It does, however. utilize the least amount of rubber and is the simplest to make. Some very simple parts can be cast by this method. Water putty castings are made in molds of this type and are quite good.

The second method utilizes a two-half mold and a static head pressure to assure complete filling. The two-half rubber mold has either an integral or external reservoir of sufficient height to permit the weight of the metal to force itself into small detail areas. While this system can give very satisfactory parts, some disadvantages must be taken into consideration. A relatively large amount of metal must be melted, poured and allowed to cool to obtain the part each time. This is time consuming and will eventually deteriorate the scrap metal as it is returned to the melt pot. A variation of this method uses a heated vertical feed tube which is capable of being shut off at the mold contact point. The tube is attached to the melt pot and thus a continual supply of melted metal is available. This accomplishes two things -(one) very little metal other than that actually required for the part is chilled, (and two) a tube of several feet can give a head sufficient to fill the most complex mold. In practice, the two-half mold with its sprue opening is placed in contact with the tube outlet. The valve or stopcock is opened and metal flows into the mold. After a static condition is achieved, the stopcock is closed and the mold is set aside to cool. There are two disadvantages to this method. Unless an extremely long tube is used, insufficient head will be developed and loss of detail will occur. Also the system must be cleaned out after a casting session in order to prevent breakage of the glass system. Why use glass?    Well. CerroBend is a funny metal and is contaminated by all other metal. Therefore a glass or stainless steel melt pot and system is required. This applies to any casting or melting method. Why will the glass break on cooling? Here again, it's that funny metal; it expands slightly on cooling, unlike other more normal metals which shrink This expansion, however, has the advantage of assisting in filling small detail.

The third method utilizes centrifugal force to induce a relatively small amount of metal into the mold cavity under pressure thus achieving the same type of pressure casting as does the static head method, only better. In this system, a two-piece round mold is used and the metal is ladled in from a glass or stainless steel melt pot. The advantages of this system are, that only a small amount of metal is actually induced, and that head pressures can be varied by varying the motor speeds. Thus you can fill more complex cavities with a more dense metal by increasing the speed. There are two disadvantages. Time is involved in attaching and removing the tie-down fixtures which are required to prevent the two halves of the mold from spreading under the pressure of the metal. Also, considerably more rubber is required for the molds than in either of the other two methods. This is the method that I have settled on and the one I will cover in detail in this article.
diagram showing mold-making procedure and casting machine
(Click on drawing to enlarge.)

The First Half of The Mold

Now to the making of a two-piece round centrifugal mold. First determine the diameter of the spin table (see the diagram and discussion on the casting machine) and this determines the diameter of the mold. Two pieces of one-half inch plywood are sawed and sanded around a center hole to a perfect circle of the proper diameter. With the two pieces bolted together through the center hole, three 3/16 inch holes are drilled through both pieces approximately 3/4 inch from the edge and on 120 degree centers. Mark the pieces of plywood, "top" and "bottom" and index mark both the plywood and the spin table so that all parts are related. These holes are for the mold hold-down bolts and also provide spacer support while pouring the molds. Now, take that bottom piece of plywood and run bolts of the correct size and sufficient length from the bottom up. Place metal spacers over the bolts equal to one-half of the mold thickness. Invert, using bolts to provide the proper spacer alignment. Using gummed paper-tape, wrap the plywood with several layers with the lower edge of the tape even against the flat surface upon which the spacers rest. When the tape is dry, restore to the upright position and place several ordinary staples through the tape and into the plywood. You now have a housing of proper height in which clay can be formed.

While clay can be pressed into the resultant "can", it is easier to melt the clay (hobby shop plasticine) and pour it into the "can". The "can" should be filled with clay so that the upper surface is even with the tape and spacers. Set this aside until the clay has completely hardened. While this is occurring, lay out a circle on paper of the proper diameter and lay the patterns in place to determine the proper spacing. Attempt to balance relative part mass by distributing equal sized parts diametrically opposite each other. Don't forget to provide space for the tie-down bolts and spacers. Also, fabricate a pouring sprue blank equal to the diameter and a length of one-half mold thickness plus 1/16 inch.

Now, using the plan devised on the paper circle, start embedding the patterns to their parting lines in the clay. Be very careful to obtain sharp parting lines between the clay and the pattern. The success of the mold may depend entirely upon the care taken at this time. When all of the patterns are in place in the clay, slide the bolts through the bottom plywood and the spacers. Place another set of spacers on the upper ends of the bolts. Thus, there are now spacers equal to the full thickness of the mold. Using rubber cement, glue the pouring sprue to the under side of the top plywood in the exact center. Allow to dry for a few minutes and then slide the top plywood over the bolts. Press down until the plywood is bottomed out against the spacers. Now remove the top plywood, upper spacers and bolts. Peel off the tape "can" wrapper. An impression will now be in the clay where the pouring gate blank is to be positioned. Remove the pouring gate blank from the top plywood with lacquer thinner and position in the proper position on the clay. Sprues can now be cut and installed from the part to the pouring gate blank The end of the sprue that butts against the part or pattern should be turned to a long taper. This permits the break off of the part from the sprue with little clean up. If the opening from the sprue to the part later proves to be too small, it can be opened up with a razor blade or knife.

The upper mold half should now be ready to pour. Check once again that all patterns, sprues and gate are firmly seated into the clay to the parting lines - make sure the parting lines are clean. Replace the bolts through the bottom plywood, spacers, and add the upper spacers and top plywood. Wrap the two plywood pieces with gummed paper tape. Staple the tape to the bottom plywood and remove the top plywood. Once again we have an open top "can" with tape sides and a clay bottom (with patterns, sprues and a gate protruding). Three bolts with spacers are also protruding. See the drawing for Step 1.
Now is the time for rubber. I leave the choice of rubber and the mix formula to you. However, after many years of working with the various types, I recommend General Electric RTV-60 as a general purpose molding rubber. RTV-11 is excellent for some items and surface coatings.

The rubber must be weighed very accurately in order to determine the proper amount of catalyst to be added. The best results that I have experienced were obtained with 5 drops of the Thermolite-12 (furnished with the RTV-60) to an ounce of rubber. Use the medicine dropper furnished with the RTV to assure even drop sizes. The rubber of the proper amount is placed in a disposable container, weighed, and the catalyst is thoroughly mixed in. If a vacuum pot is available, vacuum the mix to remove all air bubbles. These bubbles are so small that they cannot be detected visually, however, they do affect the mold. Vacuuming increases the amount of rubber space required by a factor of four so be sure that an adequate sized container is used. The rubber will foam up and then return to its original position as the bubbles burst A common pressure cooker can be hooked to the intake of an air compressor. This will usually provide sufficient vacuum to destroy most bubbles. If a vacuum is not available, use care in mixing to prevent excessive air induction, let the mixture set for a few minutes and break what bubbles do arise to the surface. The rubber will remain fluid for a considerable period of time; however, work fast in all operations.

Using an ordinary clean water color brush, paint all patterns and the clay surface with a thick coat of rubber, being sure to thoroughly work the rubber into all surface detail and holes. The rubber will tend to settle out and cover things, but sometimes air will be trapped and must be worked out, especially on intricate surface details. Once the surface is completely covered and has settled, the remainder of the rubber can be poured. Pour a sufficient quantity to bring the level up to the top of the spacers and the pouring gate. Add a slight amount more spread evenly over the surface. Allow to set for about one-half an hour, breaking all bubbles as they appear on the surface. The entire mold can be vacuumed; however, care must be taken to prevent over-foam. Install the top plywood over the bolts and inside the tape band. Place nuts on the bolts and tighten until the spacers are tight against the upper plywood. See the drawing for Step 2. You will have a slight amount of excess rubber squeezed out through various openings and cracks. Actually you are applying pressure to the rubber. The pressure causes the remaining air bubbles to decrease in size, and forces the rubber tightly against the patterns. Set aside for at least twenty-four hours. This is longer than required, but remember that haste makes waste, so regardless of how much you would like to see the result, be patient.
photo showing a mold and finished parts
(Click on photo to enlarge.)

Twenty-four Hours Later

Remove the tape from the sides of the mold, loosen and remove the three bolts. The plywood should peel from the rubber surface. Be careful not to disturb the patterns any more than possible. Now peel the rubber disc from the plywood and clay. Some of the patterns will stick in the clay and some will remain on the rubber. Clean all the clay from the rubber surface and trim any rubber leakage. Clean the patterns and reinstall them in the rubber half, refilling them in the same locations with the rubber tightly against the patterns. Replace all the sprues in the exact location in the rubber that they occupied during the pouring of the first half. The pouring gate will probably remain in the rubber, but if it does not, carefully re-install. Now we have to work upside down. Place the top plywood piece upside down with the holes through it. Carefully re-install the rubber disk over the bolts. The spacers should have remained in the rubber disk Place a second set of spacers on the bolts and install the bottom plywood in its original position. Again make the sides of the "can" from tape, stapling to the top plywood piece which is now on the bottom. When completed, remove the bottom plywood (which is on the top) and check to see that none of the patterns have been disturbed in the rubber. See the drawing for Step 3.
We must now apply a parting agent to the first half of the rubber mold in order to permit the pouring of the second half. Some RTV manufacturers recommend a 5 percent mixture of common liquid household detergent, mixed with water, while others recommend a grease type parting agent The only thing I have found usable and easily applied is an aerosol spray of Teflon applied in a very thin coat over the rubber and patterns. This product is FluorGlide made by Chem Plast Corporation of New Jersey. Spray the entire rubber mold face and the patterns with a very light coat it will disappear into the rubber, but it is there. Excessive amounts will cause trouble and if it can be seen as a complete white coating, you have over-sprayed. Pour the second half using the same techniques and procedures used when doing the first half. See the drawing for Step 4.

Let the second half cure a sufficient amount of time, then separate. The patterns, sprues, and gate can now be removed and a close examination of mold halves made. Slight trimming in the gate-sprue area may be necessary due to rubber flow. Be careful not to cut into the pattern recess. The mold becomes stronger if you let it age over about a week However, most modelers will do a trial shot right away.

photo showing casting machine made from kitchen mixer
(Click on photo to enlarge.)

Casting Machine

Any motor unit can be used as a drive; however, the RPM of the spin table must be limited to prevent centrifugal force from forcing metal between the mold halves. After investigating several setups and systems, I decided that the simplest would probably be the best. I got an old Mixmaster with a burned-out speed control. This was adapted for used by replacing the variable speed control with a set of fixed resistors to give three speeds. For most operations, at least with a four-inch diameter mold, full speed will give proper filling of the mold. The other speeds will be used with larger molds or with larger parts.

The spin table is a four-inch, inside diameter, aluminum can fitted with a drive for the Mixmaster. The shaft is a bolt with the head removed and a slot ground in the side to engage the ball in the Mixmaster drive. A cover plate with a central pouring hole fits inside the can. The spin table and the cover are drilled to the same hole pattern as used in the plywood form blocks. The cover is essential to prevent the mold halves from separating and spraying metal all over. See drawings for details.


Now you have a mold and a casting machine so let's make some castings. Melt down a pig of CerroBend (don't get it too hot, but be sure that it is all melted, the temperature should be about 190 to 200 degrees). Put your ladle where it will be heating up. Put the mold in the spin table, install the cover and snug up the bolts. Incidentally, the double spacers should be removed and a single spacer installed so that it is between both mold halves. Well, shall we give her a go? Turn on the motor and get the table up to speed Using a small funnel, pour in a spoonful of metal through the cover hole. Goofed already, huh ... too much metal so you threw a little? Well, that's something you have to learn, how much metal the molds will hold. Pouring fast will help molds to fill, but you do take a chance of putting in an over-abundance of metal. Right now would be a good time to put an open top wooden box around the casting machine to contain any metal overflow. It will happen from time to time. The casting should be set by now (a couple of minutes) so shut off the motor and after it has stopped, remove the cover and remove the mold. Did you get some flash in the can? You must not have had the tie-down tight enough. Again, this is something that you can only learn by doing.

Open the mold and let's see what you got. Probably only a partial fill or nothing in the sprues. Can be one of several things, but try another shot first because the mold has to heat up before it will shoot correctly. If it still does not work, there is a possibility that you have tightened the tie down bolts too tight and have closed off the sprues feeding the parts. Try again, not quite so tight If this doesn't help but gives flash in the can between the mold halves, it would appear that the sprues are too small. Open them up a little. There are a couple of other things that can help to improve the filling of a mold. By placing the parts at a slight angle to the pouring gate to take advantage of the velocity of the direction of rotation, parts will fill better. Once in awhile a part just will not fill no matter what you do. This is usually an air lock A knife cut from the part to the outside of the mold will usually cure this problem. The cut will permit the air to escape, but not the metal.

Now you should be getting parts, but all of a sudden they look peculiar. If you ignore this a few shots later you will discover that the metal is still molten in the mold after a normal spin period. The problem here is that the silicon rubber is a good insulator, and consequently will absorb and retain the heat from the metal. Eventually, it will get hot enough that the metal chills slowly. When metal chills slowly the result is large metal grain. If you don't let her cool off a bit, pretty soon the metal won't chill in the normal time period and when the mold is opened the metal will run out on the bench.

A couple of other items which may help out. By placing the deepest fill in the lower mold half, the air space in the upper mold is minimized resulting in better filling. A coating of baby powder on the mold halves (shake off the excess) will give a matte appearance to the castings. Too much powder will contaminate remelted metal however.

Resin Molds

The durability of RTV molds, while good, is limited and they wear out in due time. In an effort to increase mold durability, an attempt was made to use metal filled resin molds. While this provides excellent castings, it does have some limitations. The mold is not flexible, so straight draws are required with draft, and the metal filled resin molds must be provided with release agents to preclude them bonding to patterns and each other. In addition, the mold must be heat treated to at least 50 degrees over casting temperatures.

The mold material used is Devcon Aluminum Filled Epoxy, type F-2 and the mold making process utilizes the same steps as outlined for RTV molds except mold release must be used on the clay patterns in each step. A wax type mold release is recommended rather than a spray type. Apply a thin layer of wax, allow it to dry for about five minutes and buff thoroughly, especially in small detail, cracks, joints, etc. Be absolutely sure this is adequate draft for all patterns, sprues and gates and that all cracks the F-2 can run into are filled, otherwise the pattern is going to be buried in the mold material. When this happens, it's almost impossible to get it out. I know, I've done it. When the mold halves are complete, the pattern removed and everything cleaned up, heat treat the mold in the oven for at least an hour at 50o F over casting temperature. The first castings usually provide excellent parts due to the rapid chilling and resultant small grain size. Resin molds may retain some heat and eventually have large grain marks, so let them cool off. These molds should last forever with proper care.

Modified Lower Cost Molds

The price of RTV has increased drastically since the original version of this article first appeared in the September 1966 FINELINES. The following technique may be used to reduce the mold cost and provide additional mold flexibility for extraction of complex parts.

Prepare the mold form with plywood disks, clay and paper tape in preparation for the first half-pour of RTV, as described previously. Carefully paint RTV over all parts and the clay gradually building up a layer of about 1/8 inch thickness. Let the layer cure thoroughly leaving the paper tape and spacers in place. Mix a batch of plaster or Hydrocal and pour a layer into the mold cavity, working carefully down on the RTV layer. Reinforce with some nylon mesh material (your wife can help out here) and pour the balance of the container full. Install the plywood disk and tighten up the tie-down bolts. Disassemble and repeat for the second half. When completed you have a plaster slug with an RTV split liner where the actual casting takes place. In use, the RTV may be removed from the plaster and being thin allows the RTV mold to be peeled from extremely complex parts. The plaster also provides a heat sink drawing heat from the RTV molds and providing a longer casting time before the RTV heats to a point where the castings will not chill. Care must be exercised to prevent tearing the thin RTV liners or breaking the plaster mold backings. The plaster backings must be thoroughly dry prior to use.

This article originally appeared in the Nov/Dec 1980 Narrow Gauge and Short Line Gazette, pp 28-32.  Copyright 1980 by Benchmark Publications, Ltd.  Used by permission.