In this article, we will look at the manufacturing methods that are typically used in a foundry, with comparisons of the methods used during the New Zealand gold rushes 150 years ago and now, illustrated by photos taken in March 2018 at A&G Price, Thames, New Zealand. Contrasting the technology of 'then and now' emphasizes the fact that the fields of metallurgy and engineering were really in their infancy, perhaps late infancy, during the early years of the gold rush in New Zealand. Although technology has advanced, the basic principles of sand-casting have not changed dramatically.
Nearly everything produced in a foundry in the mid-nineteenth century was made by casting metals, with some machining of the castings. Parts that were required for the gold fields included all the components of the stamper batteries, water wheels and steam engines.
Key parts of the stampers included the cylindrical cam followers that were attached to vertical shafts with the hammers or stampers attached to the bottom of the shafts. In the 1860s, some of the hammers weighed 500 to 650 lb (110 to 300 Kg) and were being dropped 6 to 10 inches (15-25cm) at a rate of more than 60 impacts per minute. Later, in the 1890s, one battery of machines in Waihi had 60 stampers, each weighing 900 pounds (200Kg) and dropping 6 inches (15cm) 92 times per minute. The noise from all these stampers was deafening.
The vertical shafts were lifted up by the cam followers sliding along the surface of the spiral-shaped cams as they rotated. It is worth noting that the cam-followers were designed to rotate as they were lifted, so that the vertical shafts rotated and the stamper would fall in a different position each time. This rotation would ensure even wear of the shoes. The stampers had removable shoes that wore down and had to be replaced periodically.
The cams are an interesting shape. They form a spiral curve designed to lift the weight of the hammers at a constant rate until the cam-follower drops off the end of the spiral, allowing the hammer to drop onto the quartz, crushing it into a fine sand.
It was important to lift the weights at a constant rate to keep a fairly even load on the power source which was a water wheel or steam engine. A battery generally had four or five stampers with the same number of cams that were all rotated into different positions on the drive shaft so that they dropped at even intervals.
The spiral of the cam is called an ‘involute curve of a circle’ or Archimedes spiral, which can be drawn by taking a piece of string, connecting a pencil onto one end, and connecting the other end to a cylindrical object such as a piece of pipe. As the pencil is moved around the cylinder, the string wraps onto the cylinder and becomes shorter so the radius of the curve gradually decreases. The amount that the string shortens with a full turn is equal to the circumference of the cylinder. This method was probably used by the men at the foundry to design their castings.
The foundries also made gear-wheels to drive the stampers, berdans and other machines. Interestingly, the teeth of the gear-wheels were also made with an involute curve on their surfaces to ensure even loads on the teeth as they slid over each other.
Berdans were another important machine used in the gold fields. These were like a dish with a shaft connected to the center so that it could be rotated. The shaft and dish were tilted slightly. The early versions had a cannon-ball placed in the dish. The tilt would cause the cannon ball to roll around the dish as it rotated, so that it remained at the lowest point. Quartz thought to contain gold was placed in the dish, and the rolling action of the cannon ball would crush it to a fine powder. Mercury was often placed in the berdan and amalgamated with the gold, a method that was also used for extracting the gold from the quartz-sand produced by the stamper batteries. In addition, the berdan could be used with hand-picked ‘specimen’ ore that had not been through the stamper battery if it had a high gold content.
When referring to the amount of gold produced by the mines, it has to be remembered that the gold bullion they produced contained about one third silver, and this affected the price they were paid for their bullion. It wasn’t until the creation of The School of Mines in Thames in 1886 that they were able to begin addressing the content of silver, copper, mercury, lead, zinc, iron, arsenic and antimony in the bullion or lost in the tailings.
Initially, water power was the cheapest form of energy, and it was harnessed using mostly overshot water wheels where the water was fed into the top of the wheel, but sometimes undershot wheels were used effectively in fast-flowing water. Overshot wheels were generally preferred when water could be supplied from a higher elevation because the water dropping down in the scoops on the front of the wheel would produce more torque and power with less water than an undershot wheel. But even the largest wheels could not produce more than 10 horsepower (HP) or 7.5 kW. Sometimes turbines were placed directly in flowing water.
But water supply was limited, and varied with the seasons, so steam-power became the preferred power source, and by 1871, two thirds of the 719 stamper batteries were using steam engines, despite the fact that they cost twice as much to run. They were also used to operate the winding engines to take miners underground and to retrieve the quartz, and to operate water-pumps and other equipment.
In the early days, steam engines were imported, but it wasn’t long before the local foundries were building boilers and all the parts needed to build steam engines to drive the stamper batteries. All of these grew in size and number as the mining of quartz increased. In fact, A&G Price had built their own steam-engine in the first stage of building a foundry in Onehunga in 1868 before they started providing machinery for the gold fields.
Eventually the enclosed turbines patented by Pelton in Nevada in 1859 were found to make much more efficient use of water under high pressure. They had a series of scoops arranged around the edge of a wheel and a high pressure jet of water was aimed directly at the scoops. It wasn’t until 1884 that Pelton wheels were demonstrated at the San Francisco Exhibition and A&G Price obtained the New Zealand manufacturing rights for them. They made large numbers of them starting in 1884 and some can be found in museums all over New Zealand.
Amazingly, A&G Price produced advertisements claiming that Pelton wheels were 80 to 90% efficient at extracting the available energy from flowing water. The numbers they produced do work out to 80 to 90% efficiency when checked, and government tests also confirmed it.
They ranged in size from models 18 inches (45cm) in diameter to the 3 large Pelton wheels they installed near Te Aroha. Two of the wheels were used to operate the 41 stamper battery and produced 60 HP (45 kW) each. The third was used to run berdans and produced 30 HP (22.5 kW). The three turbines were fed by a 20 inch diameter water pipe dropping 225 feet producing a pressure of 90 pounds per square inch (620 kPa) at the turbines.
After the the government completed their £70,000 scheme to bring an open-topped water race from the Kauaeranga Stream above Shortland following the contour lines around the hills to Grahamstown in 1878, and the development of the Pelton wheel, water became the preferred power source for stamper batteries. In 1884, there were 504 head of stampers, with only 6 batteries operated by steam and 20 batteries using shotover or turbine water wheels.
A special type of casting that was used at A&G Price was the pouring of white metal bearings. White metal is a soft alloy which moulds itself to the shape of the shaft, forming a firm fit that could be lubricated with oil. It could be made from combinations of lead, tin, antimony, copper and sometimes silver or zinc. It had a low melting point of 230 to 300C. After casting, the bearing would be fitted onto the shaft, but first the shaft was painted with a thin coating of blue paste called ‘engineers blue’. The bearing would be fitted, the shaft turned and the bearing removed for inspection. Any high spots would be marked blue. The engineer would then use a scraper tool like a knife to manually shave off these high spots. Finally the machine would be assembled. If the bearing had any remaining high spots there would be increased friction at that spot and the white metal would soften and run into surrounding hollows. After that there would be no high spots. This is why new cars had to be run slowly for the first few thousand miles while the white metal was ‘run in’. These days, precision machining of the bearings means that cars no longer need to be run in.
At the beginning of the 20th century, A&G Price began building steam locomotives. Because steam engines were only about 7 to 11 percent efficient at converting heat energy into mechanical energy, every effort was made to improve efficiency. Most of the inefficiency was in the boilers where much of the heat disappeared through the chimney. Both boiler designs and engine designs were improved over the years.
The term compound steam engine refers to a piston that could be pushed both forwards and backwards in its oscillating motion by steam pressure. Simple engines applied pressure only during movement in one direction.
Then there were double and triple expansion designs. Steam under high pressure from the boiler (up to 200 pounds per square inch) was used to push the piston down but when the steam escaped through the exhaust valve it was still under significant pressure. This intermediate pressure steam was used to operate another piston but since the pressure was lower, and the volume of steam greater, a larger piston was used. In a triple expansion design, the steam escaping from the intermediate piston was then applied to a third, even bigger piston.
The term ‘geared engine’ refers to railway engines that had a gearbox. Long-haul high-speed engines did not require gears because steam can provide maximum power at zero speed, continuing to provide sufficient power at high speeds. However in shunting applications using small engines, there was some advantage in gearing down the engines to provide increased power for pulling heavy loads at low speeds. These engines needed power to be provided through multiple wheels to gain sufficient traction to pull heavy loads from a stand-still. For example, a model they built in 1912 had a two-speed gearbox driving all 16 wheels, arranged in 4 bogies with 4 wheels each and driven by drive-shafts with bevelled gears to achieve a top-speed of 10 miles per hour. The first geared-engine was built for the Wairongomai tramway in the Te Aroha gold fields in 1885. It was later used at Kennedy Bay bush tramway hauling logs and then by Smyth Brothers from 1905 to 1906. (These geared engines can be confused with cog-wheel engines that use a large gear wheel or pinion mounted under the engine to mesh with a linear rack-gear mounted between the rails on very steep inclines.)
In the last 150 years, the process of sand casting metals has changed little, and the underlying method remains the same. The history of sand casting goes back to the beginning of the bronze age, over 6500 years ago.
The ancient method of lost-wax casting was used in Roman times to make bronze statues. A model of the statue was created in wax. Sand was packed tightly around it and then molten metal was poured into it. The heat of the metal melted the wax allowed it to drain away, and the void that was left in the sand was filled with molten metal.
A more modern variation of this method is to make the model, or pattern as it is called, from foam plastic, such as polystyrene foam, that can be machined with computer-controlled (CNC) machines. The pattern can be melted during the casting process, or it can be removed before the metal is poured. If it is removed, one pattern can be used to make many identical castings. However, the pattern has to be designed for removal and more complex shapes can be made if it is not removed. In some cases, an existing object can be cast by making a mould using the object itself rather than a wooden copy.
The process of sand casting is started by building a pattern which is used to make a mould. The pattern which is usually made out of wood, is placed on a flat surface and an empty box with no top or bottom is placed over it. Then fine casting-sand is packed around the pattern, and the box filled up with sand. The box is turned over so that the pattern is now exposed on the top of the packed sand. Clearly the sand has to be able to hold its shape and stick together after being packed into the box. Now graphite or some other non-stick agent is sprinkled over the surface.
Then a second box is placed over the top and latched onto the bottom box. The lower box is called the 'drag' and the upper box the 'cope'. Again, sand is packed around the pattern, and the top box filled up with sand. At this point, we have a complete mould but the pattern is still inside it. So the two boxes are taken apart and the pattern carefully removed. You can imagine that some shapes could not be easily removed, and this has to be taken into account when the pattern is made. The sides are usually made with a slight taper so that the pattern can be lifted gently out of the sand without damaging the mould.
As metals cool, they contract which can result in dimples and defects forming in the shape of the casting. Keeping sufficient pressure in the molten metal helps to reduce this problem. The moulds shown in these photos have an extension on the filler gate, like a little chimney to increase the pressure of molten metal. Since iron is eight times heavier than water, this extra height creates a significant increase in pressure. There is also a photo (Figure 13) of extra metal being poured into the riser to maintain the quantity of molten metal in the mould.
The riser has to contain a fairly large mass of molten metal so that it will not solidify before the casting. Although it has to be large, the point where it connects to the casting itself should be small so that it can be more easily removed from the casting after it hardens. There are several variations in the design of the riser but we will not dwell on them here. Suffice it to say that it is not a simple thing to design but trial and error gave the men the experience they needed in the 1860s.
These issues also have to be kept in mind by the foundry men who pour the metal into the mould. They have to take care to keep the channels filled up and take measures to prevent the filling ducts from cooling and setting too quickly. One method is to sprinkle chemicals onto the surface of the filler and riser to insulate it, and to generate heat and slow down the cooling. These are called exothermic chemicals because they generate extra heat through chemical reactions.
Another problem that can occur is that air and other gasses can get trapped in the void, leaving defects in the casting. To prevent this tiny holes are made in the sand casting at the highest points in the pattern to allow gas to escape.
When the two boxes are put back together, there is a void in the shape of the pattern (Figures 4,5 and 6). But how do we pour the molten metal into the void? Well we should have included some ducts or holes in the sand for the metal to run into the void. There are usually at least two vertical rods placed in position before the sand is placed in the boxes. One, referred to as a gate, is for pouring the metal into the void, and the other, called a riser, allows surplus metal to rise up and maintain a constant pressure of molten metal in the void as it begins to cool.
Finally, after the metal is cold, the box is opened, the sand shaken out and we should find a nice metal object that looks just like the wooden pattern. But is it identical? Not quite. It will have shrunk by a few percent as it cooled. The amount of shrinkage depends on the kind of metal that is being cast.
To compensate for this, the pattern maker has to make the pattern slightly bigger. To do this, he would use a special ruler that has the required amount of contraction built into it. So, although it may have been marked in inches, each inch would be, for example, 5% longer than a true inch. That way, the pattern ends up being 5% bigger, and the final casting ends up being just the right size. The pattern-maker would have several rulers designed for the different amounts of contraction expected for different metals.
Kauri timber was a very fine grained timber which could be easily carved and shaped to make excellent patterns. There are many fine examples of kauri patterns in existence today, exhibiting excellent craftsmanship and that can be seen in several local museums. Some are still used by local historical groups such as The Hauraki Prospectors Association at the Goldmine Experience in Thames, to cast parts that are identical to the originals made 150 years ago. A&G Price also have a large library of patterns which they still use (see Figure 8 above).
Now that we have a mould, how do we produce the molten metal to pour into it? These days an electric induction furnace can be used to produce strong electric currents inside the metal, causing it to heat up and melt, or an ark furnace can be used to produce sparks between carbon electrodes and the pool of melting metal.
But electricity was not available in the early days of the Thames foundries, and the metal would have been melted by burning a fuel such as coke which is made by burning coal in a chamber with a restricted supply of oxygen. If iron was being made directly from iron ore (rocks with a high iron content), the ore would be loaded into a blast furnace in layers alternating with coke and other materials like limestone.
This furnace was a tall tower lined with special refractory fire bricks. A fire was started at the bottom of the tower, and air was pumped in at high pressure at the bottom. This would melt the iron which would run to the bottom of the furnace. When a plug was removed, liquid iron would pour out.
The resulting cast iron was often called pig-iron because when it was poured out into a channel, it flowed into many side channels to form ingots that looked like piglets feeding. Similar furnaces could be used to melt ingots of pig-iron imported from England or Australia.
When used for casting, the molten metal is usually produced in a crucible which is a ceramic vessel. It would be poured into a container like a bucket for transport to the moulds. During transport and pouring, the metal cools and therefore, it has to be heated to a slightly higher temperature to compensate. These days, a digital thermometer using a thermocouple is used to measure the temperature precisely.
As you can imagine, the process of casting described above produces a solid metal object. What if the customer wanted it hollow? Machining may be required but whenever possible, this is achieved by making an object called 'a core' that is placed in the sand-mould after removing the pattern. If this is made of sand or clay, it can be knocked out of the casting after it is finished, leaving a hollow inside. These cores are made in another casting process that does not involve molten metals, but these days they are made by mixing two chemicals that form a resin.
These days cores are made by mixing plastic polymers such as epoxy with sand.
The lathe was considered the primary machine tool as it can perform so many operations and originally it was used to create other machine tools. There are many machine rooms filled with lathes, milling machines, shapers, and now computer controlled work-stations that can do all kinds of machining operations in a single CNC machine.
What about the quality of the metal? Most castings are actually made using alloys of several metals combined, and the exact mixture determines the properties of the metal alloy. These days a highly specialized metallurgist is employed to analyze the metals before and after pouring to ensure that the mixture is correct.
An x-ray-spectrophotometer is employed to rapidly analyze metal samples that are removed from the crucible at various stages. In this machine, a powerful beam of electrons is directed at a polished metal sample in a vacuum causing x-rays to be emitted from the sample. Figure 21 shows metal samples with two burn marks where the electron beam struck the surface. The wavelength or 'colour' of the x-rays are different for all the metal elements. The percentage of each metal in the mixture is instantly displayed on the screen. The metallurgist can then put various additives in the mix to create exactly the required alloy before it is poured into the moulds. They make all kinds of interesting metals at A&G Price from aluminium alloys to several types of stainless steel and even a bronze alloy containing silicon that looks like gold - the new Thames gold - used in movie sets.
They also have high-tech methods for detecting defects in the final product including ultrasound for finding bubbles inside the metal castings. Sprays of liquid dye can be applied to the surface and they concentrate in cracks making them visible. There is also a device that produces an electric current in the surface of the metal so that metallic particles suspended in a liquid spray are attracted to the sharp edges of any cracks, making them visible. Then corrective action such as welding can be used to correct the defects. This kind of quality control is demanded in military applications.
During the time of the goldrush, they did not have any of this technology and limited knowledge of alloys, so they would have relatively little quality control. They probably used almost exclusively cast-iron from the blast furnaces.
There are several types of cast-iron but generally it has a very high carbon content from the coke, perhaps 2-4% and several other elements including silicon. The carbon reduces the melting point from that of pure iron, from 1535 degreesC to about 1200-1375 degreesC which was a significant advantage in the early days of iron smelting. The carbon content also makes cast-iron hard but very brittle so it was inclined to break. To compensate for the brittle character and possible defects, the engineers made their castings bigger and more sturdy with bigger safety factors. But they would still have problems with castings breaking, and requiring the foundry to make more replacement parts. Cast-iron could not handle loads under tension and tended to fail, but it could handle heavy weights under compression. Engineers did not realize this, and many bridges built with cast iron collapsed around the era of the gold rush. The solution at the time was to use wrought iron. The iron is heated and hammered repeatedly to physically squeeze carbon and impurities out of the iron. This resulted a softer product that had much greater strength under tension.
As the understanding of metallurgy evolved in the 20th century, new methods were developed including the addition of nickel to make steel tougher and chromium to make it harder, or both to make stainless steel.
Steel is made from cast-iron by burning off some of the carbon. This was achieved by blowing air through the molten iron, reducing its carbon content to less than 2%. This became possible in 1856 with the patenting of the Bessemer furnace, but it appears that this type of furnace was not available in New Zealand during the early goldrush era. The resulting steel is a less brittle product which can be modified by various methods of heat treatment. Similar methods of heat treatment were used many years ago.
The properties of iron and steel can be modified by heat treatment to change the crystal structure within the metal. For example, steel can be hardened by heating to a high temperature until it is red hot (over 900C) and then dumping it into oil or water. The crystalline structure that forms at high temperature is locked in by the sudden cooling. However, the hardened metal is too brittle for most purposes so it has to be softened just a little by heating it just the right amount. This process is called hardening and tempering.
The tempering process can be controlled by observing the colour of a polished piece of metal. As it is heated in air, the surface begins to oxidize and change colour: first it changes to a straw-yellow which gradually darkens to brown and eventually to gun-metal blue. If you want to make a hard wood-chisel, you would stop heating when it is a light straw colour. Tools can also be case-hardened where the inner core of the metal is tough but soft, and only the surface is hard and brittle.
Although the buildings at A&G; Price are up to 150 years old, all these methods are used extensively in the modern foundry to produce a wide range of products for diverse companies in New Zealand and around the Pacific and Asia.
