We have often been asked about the industrial past of the Felbridge area. Whilst we have had talks covering Warren Furnace (within Furnace Wood) and touched upon part of the mould making process during our talk on rope, we have not covered the gunfounding process which took place at Warren Furnace.
This is not meant to be a technical paper detailing the gunfounding process, rather it is a document that covers the processes required and provides descriptions and explanations to a series of period illustrations that were used to accompany the talk.
There are three extensive illustrative sources for gunfounding, the first is a series of 152 engraved plates that accompanied David Emmanuel Musly’s Traité ď Artillerie completed between 1760 and 1766 inHolland. The second is a series of 50 wonderful paintings by Jan or Pieter Verbruggen which were painted sometime between 1776 and 1782 at the Woolwich gun foundry. The third is Isaac Landmann’s sketches in his ‘commonplace book’ regarding gunfounding at Woolwich Brass Foundry (1793-1796) whilst he was Professor of Artillery and Fortification at the Royal Military Academy at Woolwich.
The purpose of this talk is to demonstrate the complexity involved in the manufacture of cast guns as well as the wide range of skills required and to reflect on the scattered evidence relating to Warren Furnace and what this tells us about the processes and methods that were taking place within Felbridge at a similar period.
The founding process requires the making of a mould which is then placed muzzle upwards below the furnace such that the molten metal can be poured into it. The casting is then taken out of the pit and broken out of the mould and then finished to complete the gun. We will now look at these processes in more detail concentrating on the manufacture of cannon rather than large mortars which have a different mould preparation.
For cannon and small mortar, the barrel model was built up on a wooden spindle on a brick firebox. The spindle was rotated and rye-straw rope was wound tightly onto it. When the straw was of sufficient depth a clay layer of about an inch was added. A small fire was lit beneath the spindle to slowly dry the model as it was being built up. A template called a strickle was used to form the shape of the barrel. Melted wax was then poured onto the mould and the strickle used to make the fillets and reinforcing rings highly defined.
The detailed ornaments were made using the lost wax process to make the models that were then attached to the model barrel using skewers. Lifting rings called dolphins were added to guns for land use, Raby was the only gunfounder who had perfected the founding of iron cannon including the dolphins in the main casting. The trunnion models were also added. This completed the barrel model. The model was longer than the eventual barrel as it had a feeding head model at the muzzle end, this compensated for the shrinkage of the metal as it cooled in the mould and also provided a headspace into which impurities could rise.
The first layer of the mould was critical as this was the one against which the metal would eventually be cast. A mixture of sieved clay and fine sand was applied in thin layers and allowed to dry slowly building up the mould. Once it was about an inch thick, it might be reinforced by winding hemp around the model. The iron skewers would be removed at this stage, the wax ornaments would now be held in place by the clay of the mould. A further layer of coarser clay was added over a constant fire, making the mould up to its eventual thickness. This varied with the size of the gun, being up to about 3 inches for the larger cannon.
The mould was then reinforced with iron staves and hoops. All of this work to construct the model and then the mould around it has taken place on the turning frame. The mould was now removed from the turning frame to enable the model to be removed from the centre.
To remove the model, the tapered spindle was firstly struck out and then the rope was pulled out. A fire could be lit within the mould to melt the wax models and to make the clay brittle and easier to remove. Pieces of the outer clay of the model lining have been found scattered on the fields at Felcot Farm, on Hedgecourt Common. A hole was added into the feeding head to provide a filling point for the mould, this was known as the sprue hole. The mould would be checked and any defects addressed before a thin layer of carbon was applied to prevent the metal sticking to the clay of the mould.
The mould was still open at the breech end and a cascable mould would be made. This mould had to be particularly strong as it would take the weight of the molten metal during casting. The process followed was very similar to the one used to make the barrel mould, a model was made on an iron spindle and shaped before a mould was made round it and the model removed. The cascable mould would also be reinforced but this was done either with iron staves and hoops like the barrel mould or by making it fit tightly into a metal container.
Preparing the casting pit
The moulds were baked by fire until they were red hot, this made the clay of the mould porous such that gases could escape through, unfortunately it also made it fragile and hence the larger moulds were baked in or very near the casting pit to minimise later transportation.
The depth of the casting pit would be adjusted to suit the barrel length such that the sprue hole in the feeding head was a short distance below the tapping hole of the furnace. The casting pit for single moulds was similar to a wooden barrel sunk into the ground with wooden staves around the sides held in place by iron hoops and an adjustable wooden platform to provide the floor at the correct height. Multiple moulds would be placed within a rectangular brick built pit with the largest mould nearest to the furnace.
It was very important that moisture was not allowed to enter the moulds prior to casting; in 1716 this had been the cause of an explosion in a foundry at Moorfields, London which killed or wounded 17 men. There was an increased risk of moisture ingress where the casting pit was below the local water table requiring extensive waterproofing and in some cases pumping mechanisms to remove water from below the pit.
The cascable moulds were placed in the pit first and the barrel mould lowered onto it. In smaller pits it would be necessary to attach the cascable mould to the barrel mould before the pair were lowered into the pit. The openings of the mould were covered to prevent any debris falling in. Earth was packed tightly around the moulds in layers about 6 inches thick with each layer being rammed down until it was solid. Speed was very important as the earth contained moisture which would permeate into the mould.
Loam channels were added to the top of the packed earth down which the molten metal would be directed to the top of each mould. The loam channels would be baked solid using a charcoal fire.
Bronze cannon were cast solid and the barrel bored out afterwards. Iron guns were often cast hollow with a core, called a newel, placed in the mould. The newel was made up of iron wire wound on an iron rod and all covered in clay, it was held in place with iron rods which became incorporated into the body of the barrel. The barrel would then only require reaming (up to 6mm removed) to obtain the required size, although the presence of a core tended to affect the metal and the casting often had impurities near the core. In 1771, the Government insisted that iron ordnance had to be cast solid requiring boring and thus considerably more power. It is not known if Raby cast his bronze mortars with a newel, the Verbruggens at Woolwich were casting theirs solid, but Musly stated in his treatise that he believed they should be cast hollow as this increased the hardness of the metal that later formed the inner wall of the bore.
Ordnance was made of both iron and bronze (the latter known as brass in the 18th century). The furnaces for these are very different. Iron was extracted from the ore in a blast furnace using charcoal with water powered bellows blowing air into the furnace. The end of the blast process providing molten iron (for further details see Warren Furnace handout, 2000), bronze was normally made away from the foundry and only melted in a reverberatory furnace ready for casting. The bronze source would often be either failed previous castings or even captured cannon. The reverberatory furnace used heat from a firebox alongside the furnace to melt the bronze, it was not necessary to force the feed air using bellows. Raby was casting both iron and bronze ordnance at Warren Furnace and thus this would have required both a blast furnace and a reverberatory furnace each with their own casting pits on the site.
As the operation of the blast furnace has been covered in a previous talk, we will concentrate on the reverberatory furnace and the generation of the molten bronze. The largest pieces of metal were put in first and nearest the fire, this could be scrap cannon or even captured ones. In 1770, Raby was given 50 tons of bronze by the Ordnance Board to be able to cast 10” and 13” mortars for sea service.
The fire was slowly increased and the partly fused metal pushed further into the furnace until completely melted at temperatures of just above 1000ºC. The molten metal would be stirred using dry pine poles (which carbonised) to ensure it was well mixed. The firing would have been taking place for many hours when the surface of the molten metal would be scraped using a wooden rake to drag the floating dross out of the doors onto the furnace floor. Smaller metal pieces were added at this time, this could include the waste from machining operations as well as pure copper or tin. The molten metal was stirred again to ensure the added metal was evenly distributed.
There is a high level of co-ordination required in the stages approaching the time when the furnace was tapped. The casting pit was being prepared and the risk of moisture entering the moulds increased if there was a delay before the furnace was ready to be tapped, also the quality of the casting improved if the loam channels were still hot from their baking. There was also a high level of skill and experience required to identify exactly when the furnace was ready to pour in both the iron and bronze processes.
The furnace was tapped allowing the molten metal to flow into the channels, the sprue holes at the top of the moulds were held shut until the metal level was above the hole to prevent dross entering the mould. If there were multiple moulds in the casting pit, then moulds could be filled in a sequence to ensure that they filled quickly. Once the feeding channel started to solidify, the channel was severed from the mould at the sprue hole.
About 24 hours later, the castings had cooled sufficiently for them to be removed from the pit by digging out the earth fill. This was very dirty work alongside the still hot castings. Bronze castings had to be cooled quickly to reduce the crystal size, hence the need to dig them out whilst still hot. The casting and mould was then raised from the pit using heavy lifting equipment. The weight of the mould and casting could be up to 8 tons.
The mould now had to be removed from the casting. The iron hoops and staves were removed and saved for re-use. The clay was broken off and the casting cleaned. If the gun had been cast on a core, this was now removed to leave the rough bore. Then started the laborious task of sawing off the feeding head, this would take 9 hours for a bronze 24-pounder (14 inch diameter), the feeding head for a large mortar would take even longer. Once the head was removed, the casting process was completed and the machining processes started.
The most important step was the machining of the bore, this had to be perfectly straight and round and its diameter had to be accurate to within 1/48” (0.5mm). The iron ordnance cast on a core still had to be machined to the correct calibre. Early boring methods concentrate on rotating the cutting piece, whereas later boring machines rotated the gun.
Raby had a boring machine at Warren Furnace, this is demonstrated by the guns being transported there from Gravetye with their ‘heads on’. There is also bronze and iron swarf debris at the furnace from the machining operations. It is most likely that the Warren Furnace boring machine was water powered although horse gins were used where sufficient water was not available. There are also Knight’s Carrier records of small quantities of steel being brought fromLondonto Warren Furnace, this is probably for tool making such as the boring and cutting tools.
The later boring machines rotating the gun were mounted on an extremely heavy foundation to be able to withstand the forces involved and support the gun without vibrating or moving during the machining operations. The alignment of the bore was critical and therefore the centre point of the muzzle was carefully marked and drilled a short distance. The breech of the cannon was mounted into a chuck, facilitated by an extra length in the cascable casting which had a square end. The muzzle was supported by a centre point mounted on the boring table and inserted in the muzzle centre hole. The outside of the muzzle was then machined to exactly fit a hardened bearing that would support the muzzle for the rest of the machining operations.
The bore was drilled with three separate drills before being polished in the fourth operation. For each drilling operation the bit was slowly moved forward using a windlass and a worm drive to put pressure on the cutting bit. For bronze guns, it was considered preferable by the board of ordnance to machine the outer surface of the barrel to provide a smooth surface which enabled defects in the casting to be seen more easily. This external machining could be done at the same time as the drilling operation where there were no ornaments. The remaining external areas had to be hand finished after boring.
It was also necessary to drill a vent connecting the end of the bore to the firing pan. The trunnions also had to be finished to the exact size and position required by the Board of Ordnance, correcting any slight misalignment with the bore derived from the modelling and later boring processes. The short extensions at the cascable and the muzzle used to mount the gun on the boring machine were cut off and the areas finished.
The gun would then be sent to the Board of Ordnance for proof testing where it would be thoroughly examined and fired a number of times before being inspected again. Failure at proof testing was not uncommon and thus there was a significant capital investment at risk and the gun being defaced and returned to the founder.
We have discussed here the whole process of gun manufacture. In 1761, Raby made 181 guns which were proved at Woolwich, as we have seen there is a large amount of skilled work involved in the manufacture of models and then moulds for every one of these guns, both of which are destroyed in the process. The machining operations would not look out of place in a heavy machine shop today, only the source of power has changed. It is difficult to contemplate that this was all occurring before the start of the Industrial Revolution.
Introductions to ‘The Carriers Accounts of Robert Knight’, Jeremy Hodgkinson, Bulletin of the Wealden Iron Research Group Series 1 no. 13 & 14 (1978) FHA
The Carriers Accounts of Robert Knight, FHA
The Raby Background: The Midlands, London and the Weald, Jeremy Hodgkinson within Alexander Raby, Ironmaster, Surrey Industrial History Group (2000) FHA
Wealden Iron, Ernest Straker (1969) FHA
The Art of Gunfounding, Ed. Carel de Beer (1991) FHA
The Iron Industry of the Weald, Henry Cleere & David Crossley (1995) FHA
Felbridge History Group handout ‘Warren Furnace’ SJC 01/00 (2000) FHA
Field notes, Warren Furnace, Worth, Sussex TQ348393, Bulletin of the Wealden Iron Research Group Series 2 no. 12 (1992) FHA
Course notes from Jeremy Hodgkinson’s The Iron Industry of the Weald (2001)
Gunfounding in the Weald, Jeremy Hodgkinson, Journal of the Ordnance Society Vol. 12 (2000)