The pour on August 26 went smoothly. We followed several suggestions offered both by Colonial Williamsburg tradespeople and experienced founders from outside the Foundation. Before the pour, the mold was fired in a modern pottery kiln at the College of William and Mary (Brad, thank you!) to approximately 1500-1550 degrees Fahrenheit. This resulted in a much more “ceramic-like” material, with all the fiber and any remaining wax burnt out of it. That eliminated several variables. We eliminated two others by using a modern flux and degasser. When the furnace was tapped, the flow of bronze appeared to be less viscous than that of our first attempt, and the material itself was a bright, gold color. A bit of the initial flow was diverted from the mold cavity.

The mold filled nicely, and the observed shrinkage looked good. The next day, Thursday afternoon, we excavated the mold and broke it away to reveal a smooth-surface casting with nice detail. On Friday afternoon, we sawed off the deadhead. While the surface revealed was a vast improvement over the first casting, it still had a ring of small (pin-head size) holes just below the surface and some small shrinkage gaps. Slicing the deadhead a couple of inches above the first cut revealed more porosity.

Although we won’t know until we clean the surface of the mortar-portion of the casting, our hope is that the defects diminish toward the bottom of the casting. If so, we may have a mortar. We’ll find out over the next couple of weeks and let you know.

Once we are satisfied with these experimental castings, we will begin “reverse-engineering” the modern technologies we are using in an effort to complete the entire process using only period methods.

Watch The Second Pour.

We are in the process of completing the mold for a second coehorn pour, which is scheduled for August 26, weather permitting.

As a result of our own analyses, suggestions we’ve received via the blog, and discussions with several experienced and knowledgeable folks, we think we have discovered the source of at least some of our problems and are working to correct them.

The proof will be in the pour!

Listen to the podcast "The Joy of Discovery" to hear more about the upcoming pour.

Further investigation of the “porosity” has revealed that many of the pockets have bits of material in them—some are likely loam from the mold, others could be dross. The furnace is tapped at the bottom of the melt, so I don’t think it likely we are being contaminated by the stuff floating on top of the metal.

But, we did not divert any of the run, and that could mean that we were washing out stuff from the furnace itself, picking up bits from the trough that runs from the tap hole to the mold, or that the mold itself may have spalled off bits as the metal heated and filled it.

Diverting the flow could possibly take care of the first two of these potential problems and is certainly something we’ll look into. We’re also looking into whether or not we fired the mold to a high enough temperature to dry/harden it properly and whether or not the mold-making mixtures we were using resulted in a porous enough shell to permit gases to escape.

As for degassing, we’d like to stick to the techniques used in the 18th century—but, again, we’ll need to experiment. Using a modern degasser in a test pour could help us determine whether or not that is the/a problem.

We’re experimenting with some of the mold making questions now and hope to get things ready for another test pour -- a coehorn again -- sometime over the next couple of months.

After turning down the surface of the casting several tenths of an inch, we’re finding even more porosity, including some large, and relatively speaking, deep holes. While this is disappointing, discovering such problems was the reason for this initial pour.

We currently are trying to determine whether the gas causing the porosity was in the bronze or the mold. If the latter, then the solution will be fairly simple: burning out/drying the mold more thoroughly. If the problem lies with gas in the molten metal, then dealing with it is more complex.

It is my understanding that the main culprit is hydrogen actually dissolved in the bronze. As the bronze cools, hydrogen bubbles can form, leading to porosity. Vibrating the molten metal will not solve this problem, but there were eighteenth-century means of controlling it.

“Poling,” mixing of the molten bronze with fresh wood sticks, is critical, both in terms of its timing and duration. Preventing the absorption of water vapor and fluxing are also important. We need to look carefully at all three (and quite possibly something else we have yet to identify).

The first process under scrutiny probably will be the poling. We did this to some extent, but because we were concerned that the bronze was taking longer to melt than anticipated, we were hesitant to open the furnace door and allow the temperature to drop. This limited the amount and timing of the stirring we did. We now know that we can be more flexible, and we’re hoping this will help.

Another possibility is that the mold was not as porous as required. This would prevent gases from escaping. When we make the mold for our next pour (another coehorn), we may experiment with including more fiber in the loam.

For more information, see The Art of Gunfounding, edited by Carel de Beer (Jean Boudriot Publications, Rotherfield, England: 1991). It is one of the principal sources of information we are using for the casting processes, and the source of the explanation given above.

See images of the coehorn's surface in project multimedia.

The mortar we cast is a reproduction of an eighteenth-century one in the Colonial Williamsburg collection. Military forces used coehorn mortars extensively during the eighteenth century, but our original was privately owned and probably was used for fireworks or signaling. We chose it as a test piece because it is relatively small and therefore a good place to start with learning the model- and mold-technology and testing the furnace.

We now are in the process of turning the outside of the casting. As we do so, we are realizing that there is some porosity (small bubbles) at the surface of the casting caused by gas in the bronze. If machining the surface removes most of these, then we probably will bore the mortar and, eventually, test fire it.

If the resulting surface is not satisfactory, then we probably will not take the time to bore this particular piece. In either case, it is serving its primary purpose: indicating to us what we need to do to produce clean castings.

We now know that we need to determine how to eliminate gas in the molten metal—and we are continuing to study the original documents to see how better to do that. We’ll post a photo of the mortar casting after it is turned, so that everyone can see the results.

On Wednesday afternoon, we dug out the sand and dirt around the mold, raised it out of the casting pit with block and tackle, and set the mold on the ground.

The founders removed the iron bands and straps reinforcing it and cracked the ceramic mold away from the casting. What lay inside is, to all appearances, a still-very-warm, but beautiful casting with excellent surface and detail.

We won't know for sure about its soundness until the deadhead is sawn off, the surface skimmed, and the bore drilled, but so far, looks like SUCCESS!

A second coehorn model is in the works for the Metalworking for Revolution conference starting this weekend. Not sure when we will attempt another casting. A lot depends upon a detailed inspection of the furnace and a determination of what, if any, repair work is required before the next melt. Stay tuned!

View video of the pourAt 2:07 p.m. Tuesday, the Founders pushed open the tap hole plug of the furnace and a smooth, red-hot stream of molten bronze flowed into the coehorn mortar mold.

The bronze had melted shortly before, after being heated, slowly at first, through the night and this morning. We won’t know exactly how the casting looks until we dig it up and break away the mold tomorrow afternoon, but the heating and pouring process went without hitch.

A number of folks lent a hand out at the site over the night and today. Thanks to them and all the other tradespeople and others who have worked on the project.

Masonry Trades, with assistance from Rural Trades, have finished building the casting furnace. Historic Trades blacksmiths provided the ironwork and carpenters the woodwork for the molding pit.

The furnace currently is being “pre-heated” to cure the masonry work. The toolmakers have made a number of implements and parts needed for the pattern-making process.

The Geddy Foundry staff has been involved in the furnace design, development of tools, and researching mold making. They currently are completing the mold for a coehorn mortar, a smaller barrel that we will attempt to cast to test out the furnace, mold making, and other procedures.

The wheelwrights have cut out the carriage trails and other wooden components and have one carriage wheel about ready to tire.

The gunsmiths have finished making the accompanying wall gun -- an oversized musket that was part of the cannon’s equipment. It can be seen at the Gunshop.

The weavers are preparing to weave the canvas required for the covers, ammunition, etc. Before the project is complete, harness makers, tailors, and other Historic Trades shops will be involved.

Current activity is best observed at the Geddy Foundry, Wheelwrights Shop, and furnace at Great Hopes.

Cannons stand as monuments to battle, as well as to the hands that made them: a confederation of specialized trades collaborating to make a formidable weapon.

Each tradesman brings knowledge and expertise as indispensable as the next. A brickmaker builds a furnace, a founder designs a mold to receive a stream of melted bronze, a gunsmith brings his skill to bear, and a wheelwright constructs a carriage to convey the barrel’s heft.

In summer of 2008, Colonial Williamsburg’s Historic Trades began an effort to do the same, recreating a light infantry three-pounder gun and the carriage to carry it. This blog is a record of the process. Check often to see photos, videos, and updates from tradespeople as they bring the cannon to completion.