Original article written by Robert Doat and Rene Evrard, Compagnie Generale des Conduits d'Eau Liege, Belgium.
Published in American Foundryman (1951)
Introduction.
On the occasion of the fourth centenary of its Vennes plant, our company has conducted research on the evolution of castings of past centuries. We summarize here, by means of silicon-carbon diagrams, the results of this research.
The structure curves adopted in some of the diagrams have been traced according to data obtrained from ancient castings which we have been able to examine under the microscope. To simplify the report, phosphourus has not been taken into consideration in preparing these diagrams, but we have shown the eutectic line corresponding to the average content of P in all of our samples of ancient castings, that is, 0.9%.
Figure 1 shows that the silicon and total carbon contents of all the ancient castings, with the two exeptions, are situated in a zone extending from 3.3 to 4.4 per cent total carbon, and from 0.4 to 1.7 pere cent silicon. This zone can be compared with the two zones of modern castings which are outlined on this diagram.
castings made by the ancient foundrymen are quite different from the ordinary modern castings, mainly due to higher carbon and lower silicon contents. The iron was made in cold blast charcoal furnaces, whereas modern irons are made in hot blast coke furnaces, and the metal then re-melted in the cupola or other furnace.
At the end of the Middle Ages blast furnaces appeared, but their weak bellows and small dimensions did not permit them to attain very high temperatures. The carbon-silicon reaction was then limited, and due to low silicon contents the metal had a strong tendency to be white.
Unfortunately, it is impossible to distinguish with certainty castings from that long-past epoch. To be able to follow the evolution of ancient castings, it is necessary to examine the firebacks and stove. plates of the 16th century. These often show dates, armorial bearings or ornamental motifs which make it possible to date them.
Figure 1. Silicon and carbon contents of 16th,17th, and 18th century castings made in cold blast charcoal furnaces range from 3.3 to 4.4 total carbon, and 0.4 to 1.7 per cent silicon. For comparison, the diagram includes the carbon-silicon zones for modern cast irons.
An examination of Fig.1 does not disclose any classification among those ancient castings. Analyses of castings of the different epochs are very much mixed up due to the appearance in this diagram of parts from various countries where the evolution of the technical processes had not been identical, parts of different thicknesses, and parts cast directly from the blast furnace or from remelted iron.
In order to understand the evolution of the ancient castings, we are going to follow it through parts of certain date, of average thickness( 0.4 to 0.8 in.) and cast in three neighbouring countries where the metallurgical processes have evolved in about the same manner-Belgium, the Grand Duchy of Luxembourg, and France.
Approximately 50 castings made in Belgium and Luxembourg from about the year 1500 to and including the 19th century and 11 castings made in France from 1584 to 1738, comprised the comparative groups. In addition, 16th century castings from Great britain, and 18 castings made in Germany from 1547 to 1738 were analyzed.
In examining the silicon and total carbon contents of the Belgian, Luxembourg and French parts of average thickness cast in the 16th century and during the first three quarters of the 17th century, it is apparent that the foundrymen of that epoch were already casting these in mottled and even slightly gray iron, but they still were working quite close to white iron. With a small decrease in blast furnace temperature or in casting section size the castings solidified white.
There was certainly opportunity to note that an increase in the amount of charcoal or prolonged holding of the metal in contact with the fuel in the furnace could make the white casting become mottled, and the mottled casting become gray, as indicated by the solid lines in arrows 1 and 2 of Fig.2.
Figure 2. Silicon-carbon diagram of the ordinary castings made in Belgium, Luxembourg, and France during the 16th and the first 75 years of the 17 centuries.
Experience also will have shown the that beyond certain limits these same means no longer increased the gray tendency of the casting, as indicated by the dotted parts of arrows 1,2 and 3. However, they must have observed that the more they increased the quantities of charcoal, the more the fluidity of the metal improved, at least up to a certain limit, which is the eutectic line.
Finally, the foundrymen of that period, who certainly were making an effort to avoid the white, hard and fragile castings, did not fail to notice that increasing the furnace blast gave them a grayer casting, since the hotter process gives more silicon. The result of this increase in silicon content is marked in Fig.2 by arrow 4. The direction of arrow 5 indicates the development of the best foundrymen, wh knew how to combine a judicious increase of fuel and blast, and who thus definitely got away from the white casting, at least for parts of average and greater thickness.
Certain foundrymen of the 17th century tried to counteract the tendency towards white iron by reducing the cooling rate of the castings. The French savant,Reamur, writes that the castings were poured in heated molds, or were removed from the mold upon solidification and placed in heated furnaces, where they cooled slowly. In any case, the line AB on Fig.2 shows the maximum silicon content used by the best Belgian, Luxembourg, and French foundrymen in gray iron castings and average thickness in the 16th century and during the first three quarters of the 17th century.
At the end of the 17th century, the European countries needed large quantities of castings on account of the wars and the great works of the reign of Louis XIV, Foundrymen made an effort to further increase the capacities of the blast furnaces, and improved the ore roasting processes. They utilized more reactive charcoals and stronger fluxes, and extended the zone of the blast furnace in which the reduction of the ore is effected, either by increasing the dimensions of the furnace, or modifying the direction of the tuyeres. Capacities of blowers were also increased. The majority of these innovations resulted in further increasing the operating temperature of the furnaces, and consequently the silicon content and the graphitization of the castings.
Figure 3. Castings made at the end of the 17th century and through the 18th century had increased silicon over the preceding period (fig.2), and essentially the same carbon content. The line"AB" denotes the maximum silicon content obtained in the preceding period.
Figure 3 shows the carbon and silicon analyses of average thickness castings at the end of the 17th century through the 18th century. The silicon content has increased notably, whereas the carbon content has remained essentially the same. The line AB (approximate maximum silicon content of castings of the preceeding period) is considerably exceeded.
At the beginning of the 18th century the foundrymen making parts of average thickness were protected from the danger of the white castings by the increased silicon content of the iron. But they encounterd another danger-that of too gray or even black castings because, with the silicon contents then attained, the castings were hyper-eutectic and of poor quality for parts subject to strains. Throughout the 18th century they tried to find a remedy for this situation.
Had Excessive Graphite.
In 1722, Reamur considered the gray castings as "impure" and advised against their use. No doubt he had observed the effects of excess graphite in the too gray or black castings. He recommended that white iron only be used, and the castings improved by malleabilization. Four years later, having pursued his research, he decide that gray iron could be used in the foundry and that, when it was too gray, it could be improved by remelting in the crucible or in various cupolas small and large. He gives descriptions of cupolas which operated with charcoal, bringing the black iron back to the eutectic. In teh 18th century, the authors'plant was oxidizing the excess graphite by again passing the iron through the reverberatory furnace. Other Belgian foundrymen loaded the blast furnace with iron scrap as well as ore, and thus reduced the graphitization of the iron.
In the second half of the 18th century, Grignon, a French foundryman and manufacturer of cannons who utilized high blast furnaces and ores containing graphitizing agents, invented, to combat the graphite, a sort of charcoal cupola. In this he remelted his iron, first breaking it into small pieces. He thus produced an "iron Regulus" which was a sort of cast steel.
Finally, in the last years of the 18th century, the Englishman, Wilkinson, in order to avoid excess graphite in the metal, invented a blast furnace of slight height, but provided with several tuyeres. In this furnace the metal remained foe a shorter toime and therefore had less tendency tp graphitize. These last two inventions did not become popular.
The process of casting metal into molds became much more important upon the development of coke and the hot blast, which made higher melting temperatures possible. The silicon contents of the ironrose suddenly from 1.7 per cent to 2 or 3 per cent or more.
When this occurred cupola melting became general. Certain analyses in the various diagrams indicate a slower development than that described so far. For example, some irons of the 18th century contain less silicon than certain 16th century irons. In those days, as at present, there were foundrymen who did not keep abreast of the times. But there is another reason for this dispersion of the analyses in the diagrams-they include irons made for casting and irons which were intended to be refined by the "Walloon" method. This refining process consisted of exposing the end of a piece of pig to the oxidizing action of the flame produced by a blower tuyere in a charcoal smelting hearth; the pig melted drop by drop and the iron, low in silicon and carbon, formed a spongy mass at the bottom of the hearth, which was then hammered to weld the grains of metal and expel the slag.
For several of the ancient dated castings it has been possible to determine definitely that they were made from iron destined for further treatment. These are the ones from Luxembourg, which produced only such iron, but where the furnace masters at times cast an occasional part with their customary metal. Silicon and carbon contents (Fig.4) of these castings made from Luxembourg smelted irons show that the process had not developed in the same manner as the irons made primarily for castings into molds.
Revert to white iron.
Fig.4-Diagram showing carbon and silicon of 16th, 17th and 18th century castings made in Luxembourg. The 18th century castings, of smelted iron, are closer to trhe white iron zone than those of the 16th century.
In fact, Fig.4 shows that the Luxemburgian castings of the 18th century (represented by triangles) form a closer group closer to the white castings than those of the 16th century (represented by circles). This shows a retreat towards the white castings (see arrow). Moreover, the smelted iron castings of the 17th century (represented by squares) are divided equally between two tendencies, which confirms that it was in the 17th century that the Luxembourg smelting furnace masters returned to the metal of their ancestors.
This evolution, the reverse of that of the foundrymen, is easily explained. From the 14th century to the 17th century, foundrymen and smelters utilized the same metal.
But the foundrymen were interested in producing gray iron castings, and the smelters were interested in having the least possible carbon and silicon to oxidize in the pig iron. In the 17th century, when they observed that the smelting of gray iron was longer
and more costly than that of the white, they separated from the foundrymen and turned back to the white iron, primarily by cooler furnace operation.
Those who had not developed as the best foundrymen were not necessarily behind the times-they may have been smelting furnace masters who knew their business well and were interested in developing in the contrary direction.
Summary
In summary, The Belgian, Luxembourgian, and French foundrymen who had developed to the greatest extent were casting parts of average thickness in white iron in the 14th and 15th centuries, in mottled or slightly gray iron in the 16th century, and in gray or even excessively graphitic iron at the end of the 17th century and throughout the 18th century. In the 18th century they were re-melting at times the blast furnace iron in order to improve it. At the beginning of the 19trh century, the appearance of coke and hot air brought about a great excess of graphite in the bl;ast furnace iron, which made the second melting of the metal indispensable.
