Metallographic Analyses of 19th-Century Steel Semi-Finished Products from Slovenia

Abstract

This pioneering study examines metallographic characteristics of 19th-century steel semi-finished products, kept by the National Museum of Slovenia. These artefacts, manufactured in pre-industrial ironworks in present-day Slovenia, reflect the craftsmanship and technological practices of their time. Metallographic analyses revealed significant microstructural variations within individual samples, attributed to differences in carbon content, cooling rates, and forging techniques. All samples contain non-metallic inclusions composed of Si, Mn, and other oxide-forming elements. The results indicate that the semi-finished products were often manufactured by combining steels with varying carbon contents and were sometimes hardened. Additionally, this study highlights correlations between the metallurgical properties of the analysed materials and their historical classification as “iron” or “steel”.

1. Introduction

1.1. Historical Background

Industrialisation began in the United Kingdom during the late 18th century and spread at differing speeds through the various parts of the continent in the following decades. In the Austrian monarchy, which at the time also included the majority of the Slovenian territory, the industrialisation process was enforced only as late as the 1830s. Critical inventions in European iron-making at the time were using coal or coke as fuel instead of charcoal, the puddling process, and the blasting of hot air into furnaces. The puddling process was developed through the second half of the 18th century in the United Kingdom and was finally patented in 1783 and 1784 by Henry Cort. Pig iron was heated and stirred in a reverberatory puddling furnace, causing its decarburisation and conversion to wrought iron, also called puddled iron. In this process, the amount of yielded material significantly increased compared to the process conducted in finery forges [1,2,3].

The puddling process was introduced in the Slovenian territory only in 1835 in Prevalje by August von Rosthorn from Vienna. His father, Matthew, originated from Lancashire, United Kingdom, and settled in the Austrian monarchy by the invitation of its ruler, Maria Theresa. In Prevalje, August von Rosthorn established a puddling furnace and a rolling mill, making it one of the first steelworks in the monarchy to produce puddled steel. However, the majority of iron-making centres in the Slovenian territory retained steel production in a pre-industrial manner, and the industrialisation process started to prevail in Slovenian iron-making only in the late 19th century [2,4,5].

The above-presented context also reflects the analysed steel semi-finished products from the collection of metalwork in the National Museum of Slovenia. It contains up to 100 various steel semi-finished products, which predominantly date from the first half of the 19th century. Recent studies of the collection have revealed detailed information on the production date and geographic origin of some of the semi-finished products, as well as the time of their acquisition by the museum.

Twelve semi-finished products (

Table 1. List of metallurgically analysed semi-finished products. (?)—Sample Nos. 2, 8, and 9 were presumably manufactured in Carniola.

Sample No.	Grade of Semi-Product	Original Name of the Grade	Origin	Manufacture Date	Length [cm]	Width [mm]	Inv. No.
1	Rod iron	Feines Zaineisen	Tržič	1821	64	9.4	N 39295
2	Steel	x	Carniola (?)	1st half of the 19th c.	47	15	N 38902
3	Brescian steel	Brescianstahl, gehärtet Nr. 0	Tržič	1821	43.5	8.8	N 38924
4	Strip iron	Wannen- oder Kübelesien	Tržič	1821	64.2	30	N 38897
5	Rod iron	Zaineisen Mittleres	Tržič	1821	66.2	10	N 38900
6	Rod iron	Groben Zaineisen für Schlosser	Tržič	1821	68.3	18	N 38899
7	Rod iron for artistic forging	Gittereisen, 8eckiges (Ottangoli)	Tržič	1821	67.3	17	N 38898
8	Rod iron	x	Carniola (?)	1st half of the 19th c.	65	10	N 38901
9	Rod iron	x	Carniola (?)	1st half of the 19th c.	13	82	N 39598
10	Steel	No. 0	Sava	1834	27.2	85	N 39599
11	Steel	No. 00	Sava	1834	25.3	72	N 38597
12	Iron plate	x	?	1st half of the 19th c.	33	550	N 38587

) were chosen to conduct metallographic analyses; provenance can be established for eight. Six semi-finished products were manufactured in 1821 in Joseph von Dietrich’s (with the title of baron from 1824) forges and ironworks in Tržič. They were formally donated to the Provincial Museum of Carniola, the direct predecessor of the National Museum of Slovenia, on the 1st of May 1821 but were transported to the museum premises only at the end of 1826. They were part of a large collection of 488 objects made in the Tržič ironworks and forges. Other than semi-finished products, the collection contained many files, scythes, and other tools [6].

The analysed items were all manufactured in a pre-industrial manner. Iron ore was melted in blast furnaces, and the extracted pig iron was later processed in finery forges to produce wrought iron or steel. Charcoal was still used for fuel, and ironworks were powered by water. The semi-finished product of “Brescian steel” was manufactured in a steel-making plant, while other Tržič semi-finished products were manufactured in ironworks and forges [6].

Apart from semi-finished products, files and various tools were also manufactured in Dietrich’s ironworks in Tržič. They used predominantly Carinthian pig iron, bought from the Treibach and Eberstein blast furnaces; only a part of pig iron was used from the blast furnace in Sava near Jesenice. In the Tržič finery forge, called Wallaschhammer (the name originates from the German word for the “Italian-type” trip-hammer), there were two finery hearths, and each was equipped with a trip-hammer weighing 600 pounds (336 kilos). Charcoal was produced from wood cut down in the vicinity of Tržič. Refined wrought iron was further processed in the forge, where semi-finished products of rod and strip iron and tool-making iron were manufactured [6,7,8].

The Tržič steel-making plant (Stahlhammer) had a finery hearth for raw steel manufacturing and two finery hearths equipped with Brescian-type trip-hammers for raw steel refining. The final products were various grades of “crate-steel” (Kistenstahl). The name is based on the manner of transport, as they were transported in crates. They were exported to the Mediterranean seaports and further on to various continents, such as India and North and South America [6,8].

In the blast furnace at Sava near Jesenice (Figure 1), roasted iron ore from the Karavanks mountain range was smelted. The majority of produced pig iron was exported, while the remainder was refined and processed in finery forges located in the vicinity of the blast furnace into various types of semi-finished products. The blast furnace that operated when the analysed items had been manufactured was presumably the same as the one described by Belsazar Hacquet circa 1780. Its height was around 6 m [9].

1.2. Research Aims

The principal aim of this investigation was to determine the differences between the semi-finished steel products, in particular with regard to chemical composition, especially the carbon content, microstructure, type, and morphology of non-metallic inclusions and hardness, in order to be able to draw conclusions about the manufacturing process and heat treatment. Because some semi-finished products have original inscriptions that mark their grades (see Table 1), a very important research aim is also to explain how these different grades are reflected by the various properties of their materials. Do these properties correspond at least approximately to the grades of the articles? We also find the comparison between similar semi-finished product grades interesting.

To identify hitherto unattributed grades of semi-finished products, the results of their analyses will be compared to analysed items with the same appearance or shape (e.g., cross-section) belonging to a known grade. For instance, the semi-finished product Inv. No. N 38902 (Sample No. 2) has a square cross-section. Three other “steel” semi-finished products also exhibit the same type of cross-section, so it can be predicted that this object is also made of “steel”.

Although a total of 12 samples were analysed, the results for only 4 of them are discussed for the purposes of this paper. The chosen samples were selected primarily based on the group’s characteristics and the object’s provenance. The presented study could serve as an interesting starting point for further analyses of other similar museum artefacts or otherwise preserved objects. This study also offers a glimpse into the pre-industrial iron-making technology of the early 19th century.

2. Materials and Methods

2.1. Materials

Objects from Tržič ironworks (Figure 2) have original inscriptions in the German language that mark their grades. One of them is a semi-finished product of “Brescian steel” (Brescianstahl) (Inv. No. N 38924, Sample No. 3). That grade of steel was introduced to Slovenian territory in the first half of the 16th century, upon the arrival of Lombard ironworking masters. They predominantly originated from the Bergamo and Brescia area, after which that grade of steel also bears its name [10]. Three semi-finished products represent various grades of rod iron (Inv. Nos. N 39295, N 38900, and N 38899; Sample Nos. 1, 5, and 6). Each of these grades served its purpose. The object with Inv. No. N 39295 (Sample No. 1) is a semi-finished product of a “fine rod iron” (Feines Zainesen), while the object with Inv. No. N 38900 (Sample No. 5) is a semi-finished product of a “medium grade of rod iron” (Zaineisen mittleres). Object Inv. No. N 38899 (Sample No. 6) was intended to be processed in lock-smithies (Groben Zaineisen für Schlosser). Item Inv. No. N 38897 (Sample No. 4) takes the form of a strip or band and is a semi-finished product of a strip iron. The inscription (Wannen oder Kübeleisen) reveals it was used to manufacture hardware for wooden containers, such as buckets and baskets [11]. Item Inv. No. N 38898 (Sample No. 7) has an octagonal cross-section, and the inscription (Gittereisen) reveals it was used for the manufacture of grilles and several other uses in artistic forging.

The terms “iron” and “steel” for the grades of semi-finished products were used to label the difference in their characteristics, particularly in hardness. “Iron” denoted softer materials, while “steel” was used to describe harder materials. Specific grades of semi-finished products are also differentiated by their cross-section.

Two objects (Inv. Nos. N 39599 and N 38597, Sample Nos. 10 and 11, Figure 3) were manufactured in the Viktor Ruard ironworks in Sava near Jesenice. They were donated to the museum in early 1835 as a part of the collection of 13 semi-finished products from this enterprise [12]. These two objects are steel semi-finished products labelled with 0 and 00. These kinds of marks were used to label different quality grades of “steel” semi-finished products, spanning from 0000 up to the Roman numeral V [13].

The provenance of other analysed semi-finished products could not be established at this time, but they were likely manufactured and donated to the museum in the same time frame, during the first half of the 19th century. This is indicated by the similarity of their appearance, as mentioned earlier, as well as the collecting policy of the Provincial Museum of Carniola in its first decades. Sample products of Carniolan craft and industrial enterprises were among the most desired acquisitions, with iron-making ranking as one of the most important non-agrarian activities at the time [14].

As already presented above, the results of analyses of only four samples are described in more detail. After the completed metallographic analyses, we divided the samples into approximately three groups based on their characteristics, such as microstructure and carbon content (see

Table 2. Chemical composition in wt%.

Sample No.	C	S	N	Fe
2	0.83	0.00	0.00	rest
3	1.13	0.00	0.01	rest
7	0.10	0.01	0.00	rest
11	1.36	0.00	0.00	rest

). One of the groups contains samples of the hypoeutectoid steels, the other consists of near-eutectoid steels, and the third group is made of hypereutectoid steels. One of the samples was chosen based on the provenance of the object. Therefore, we chose objects with Sample Nos. 2, 3, 7, and 11.

2.2. Investigation Methods

For the chemical analysis, a piece of steel was cut from each forging at the end of the object. All elements except carbon, sulphur, and nitrogen were analysed using an OES ARL MA-310 optical emission spectrometer (Thermo Scientific, Waltham, MA, USA). Carbon, sulphur, and nitrogen were analysed using the combustion method with a LECO CS-600 device (Leco, St. Joseph, MI, USA).

For the metallographic investigations, a piece of steel was cut out of the forgings and examined metallographically. After cutting, the samples were prepared metallographically. The metallographic preparation included grinding with silicon carbide papers of grades #220, #320, #500, #800, and #1200, followed by polishing with a diamond paste with particles of 3 µm and then 1 µm in size. To visualise the microstructure, the samples were etched with 2% Nital (2% HNO₃ in ethanol). The samples were examined longitudinally and transversely to the forging direction. Macro-images were taken with an Olympus stereomicroscope. A BX61 light microscope (Olympus, Shinjuku City, Japan) was used for microstructural characterisation, while a FEG SEM Quattro S field-emission scanning electron microscope (Thermo Fisher Scientific, Waltham, MA, USA) with an energy dispersive X-ray spectrometer (EDXS) SSD Ultim^® Max 65 mm² (Oxford Instruments, Abingdon, UK) was used for non-metallic inclusion analysis. The accelerating electron voltage was 15 kV. A backscattered electron detector (BE) was used for imaging. For scanning electron microscopy, the samples were polished. Vickers hardness (HV 1) was also measured on the metallographically prepared samples using a Tukon 2100B device (Instron, Norfolk, MA, USA).

3. Results

3.1. Chemical Composition

The results of the chemical compositions of the steels studied are given in Table 2. All steels belong to the group of plain carbon steels. Steel No. 2 is a near-eutectoid steel with a slightly higher carbon content than the eutectoid composition (0.76 wt%). Steels Nos. 3 and 11 are hypereutectoid (C > 0.76 wt%), while steel No. 7 is hypoeutectoid (C < 0.76 wt%). Although the analyses were performed using an optical emission spectrometer (OES ARL MA-310) and the combustion method with a LECO CS-600, and the signal was collected from a relatively large surface area, it should be noted that the analysis for each sample was carried out at only one location. Furthermore, no depth-resolved chemical analyses were performed.

3.2. Microstructure Characterisation

The semi-finished product N 38902 (Sample No. 2) is made of steel of unknown origin from the first half of the 19th century. Figure 4a shows the macrostructure of the sample in a cross-section. It clearly shows the direction of the material flow, i.e., the direction of plastic deformation during the forging. This is also indicated by the distribution and morphology of the non-metallic inclusions in Figure 4b. Larger non-metallic inclusions are strongly deformed and elongated in the direction of plastic deformation. According to the chemical composition, the non-metallic inclusions are oxides mainly containing silicon and oxygen with other elements such as Mg, Ca, Mn, K, and Fe.

The microstructure of the steel consists of coarse pearlite (Figure 4c), from which it can be concluded that the composition of the steel is close to the eutectoid point, i.e., 0.76% C. Figure 4d shows that the core of the product consists of a pearlitic microstructure (α + Fe₃C) with secondary cementite (Fe₃C) along the crystal grain boundaries, followed by a pearlitic microstructure towards the surface and a pearlitic–ferritic microstructure with a higher ferrite content at the surface. The development of the microstructure from the core to the surface indicates a decarburisation process (i.e., depletion of carbon from the surface of steel) that takes place at higher temperatures in an atmosphere with a lower carbon potential. The predominant pearlitic microstructure is also consistent with the results of the chemical analysis (Table 2).

Semi-finished product N 38924 (Sample No. 3) is an example of Brescian steel produced in Tržič in 1821. Figure 5a shows a macro-image of the section in the longitudinal direction in relation to the direction of deformation caused by forging. Again, the non-metallic inclusions (Figure 5b) consist of oxides, mainly silicon and manganese. The inclusions are distributed in the direction of the deformation; however, they are more irregularly shaped than in Sample No. 2, where they are more oval. Figure 5c shows the microstructure of the steel in the core of the semi-finished product. A martensitic (α′) microstructure with a large proportion of retained austenite (γ_r) and pearlite, as well as secondary cementite, can be seen at the grain boundaries of the prior austenite crystal grains. This means that the lower critical cooling rate was reached when the steel was cooled from the austenite region; therefore, a non-diffusion transformation of austenite into martensite occurred. It can also be observed that the microstructural constituents differ across the cross-section, so it is assumed that the semi-finished product was forged from several steels with different carbon contents. Figure 5d shows the areas with different microstructures: the top left is the ferrite–martensite microstructure on the surface, and the bottom is a martensite microstructure with retained austenite and a martensite microstructure with secondary cementite in the core. The ferritic–martensitic microstructure at the surface indicates decarburisation at the surface and that the temperature at which the steel was cooled did not exceed the Ac₃ temperature but was between Ac₁ and Ac₃, where the steel has a two-phase, i.e., a ferritic–austenitic (α + γ) microstructure. The areas with different microstructures, where the individual layers clearly differ from each other (Figure 4d), indicate that the semi-finished product was forged from various pieces of steel with different carbon contents. Based on the microstructural analysis and the results of the chemical analysis (Table 2), it can be assumed that the product consists mainly of hypereutectoid steel with a high carbon content (C > 0.76 wt%).

The semi-finished product N 38898 (Sample No. 7) is iron for artistic wrought ironwork, produced in Tržič in 1821. Figure 6a shows a macro-image of the longitudinal section in relation to the forging direction. The macro-image reveals the flow of plastic deformation during forging and the layers of different microstructures in the cross-section. This suggests that the semi-finished product was made of several steels with different carbon contents. Figure 6b shows non-metallic inclusions in stringers along the direction of plastic deformation. The EDXS analysis indicates these oxides contain oxide-forming elements (Mn, Si, Ca, and Mg). Figure 6c shows a predominantly ferritic microstructure with areas of a ferritic–pearlitic microstructure. However, an intermediate region of coarse-grained ferrite is surrounded by a zone of fine-grained ferritic–pearlitic microstructure. Figure 6d also shows an area with a predominantly pearlitic microstructure and some ferrite, indicating that the steel in this area contains a higher proportion of carbon. The microstructure on the surface contains a higher proportion of ferrite, which indicates that the surface has been partly decarburised. This is due to the processing of the semi-finished product in an atmosphere with a lower carbon potential and higher temperatures.

Based on the microstructural analysis and the results of the chemical analysis (Table 2), it can be assumed that the investigated sample is low-carbon steel. The result of the chemical analysis is consistent with the predominant ferritic microstructure with low carbon content. It is, therefore, obvious that areas with a predominantly pearlitic microstructure with a higher carbon content were not included in the chemical analysis.

The semi-finished product N 39597 (Sample No. 11) is made of steel produced in Jesenice in 1834. Figure 7a shows a macro-image of the sample in the longitudinal direction in relation to the direction of deformation during forging. The macro-image shows the flow of the material due to plastic deformation and the areas with different microstructures in layers. The oxide inclusions, which mainly contain Si, Mn, and Ca, are also distributed in stringers aligned with the direction of the plastic deformation of the steel (Figure 7b). Silicon dioxide (SiO₂) played an important role in historical steelmaking processes, particularly in forge-welded steels. Finely ground quartz sand was intentionally added as a flux to lower the melting point of slag and to facilitate its expulsion during forging. However, some slag remained entrapped in the steel, forming non-metallic inclusions. Due to the deformable nature of silicate-based inclusions, they were elongated in the direction of plastic deformation, resulting in the characteristic stringer morphology. These inclusions further support the interpretation that the larger steel pieces were produced using traditional forging techniques involving the combination of multiple types of steel with different chemical compositions and microstructures.

Figure 7c shows an inhomogeneous microstructure with a predominance of martensite. In the centre of Figure 7c, in addition to martensite, a secondary cementite was precipitated at the grain boundaries of the prior austenite grains. Figure 7d shows a martensitic microstructure with retained austenite on the surface, while in the core, there is a higher proportion of pearlitic microstructure with secondary cementite in addition to a small amount of martensite, which is shown in Figure 7e. This microstructure developed due to the poor hardenability of the plain carbon steel during quenching. From the martensite content on the surface, it can be concluded that the steel was quenched (hardened) in water. As a result, a higher surface hardness was achieved. From the microstructure analysis and the results of the chemical analysis (Table 2), it can be concluded that the semi-finished product was made from steel with a high carbon content.

3.3. Hardness

The hardnesses of the steel samples analysed are shown in

Table 3. The results of hardness measurements of samples (HV1).

Sample No.	Site 1	Site 2	Site 3	Average
2	311	316	303	310
3	891	901	882	891
7	147	145	138	143
11	870	886	901	886

. As the microstructure of the samples is very complex and, above all, inhomogeneous, the hardness depends heavily on the location of the measurement. This is due to the fact that the samples were made from different steels with different carbon contents. In addition, the samples were heated at different temperatures and cooled at various cooling rates. In addition, decarburisation of the steel surface also occurred during heating, which led to a lower hardness in the surface area. As a result, the microstructure was differentiated locally, which led to different hardnesses.

Sample No. 3 (891 HV1) and Sample No. 11 (886 HV), both with the highest carbon content (hypereutectoid steels), had the highest average hardness. High hardness corresponds to a high-carbon martensite with secondary cementite. Martensitic microstructure in steel indicates that the steel has been quenched in water from high temperatures to transform austenite into hard martensite (hardening). The average hardness of Sample No. 2 was 310 HV, which could indicate a microstructure of pearlite and secondary cementite, while the lowest average hardness was achieved in Sample No. 7 (143 HV1) with a ferritic–pearlitic microstructure.

4. Discussion

The analysed semi-finished products were manufactured of various grades of steel, ranging from hypoeutectoid to hypereutectoid steels. The grade of steel depended on the manner of subsequent use and the type of final products made from these semi-finished products. It has been proven that, even if the items were manufactured in a pre-industrial manner, without concurrent modern technology of the period, the master ironworkers could produce various grades of semi-finished products from the same raw material. This, thus, demonstrates the considerable know-how of the contemporary iron-making masters.

In the first half of the 19th century, the majority of high-carbon steel was produced either in the cementation process, in the crucible furnaces, or in the traditional way in the finery forges as the analysed semi-finished products from the National Museum of Slovenia show [3]. It was only in the last decades of the 19th century that the mass production of high-carbon steel was developed using the Bessemer-Thomas and Siemens-Martin processes. From that point on, steel was also used as a construction material. In the puddling process that spread across Europe during the first half of the 19th century, the quality range of the material obtained was significantly narrower. However, it enabled a much higher amount of iron production. One highly significant example is the rise of iron production in the United Kingdom between the mid- and late-18th centuries. In 1750, bar iron production reached around 18,900 tonnes, while in 1796 it reached 131,500 tonnes. The puddled iron, often also called wrought iron, was predominantly used as a construction material, and the chemical composition could vary significantly. It was of low carbon content, usually in the range between 0.05% and 0.15% but also reaching up to 0.3% C. Puddled iron also contained higher proportions of phosphorus and sulphur, at first reaching up to 0.4% and the latter reaching up to 0.07% [3,15,16,17].

This is also reflected by the metallographic analyses of the puddled iron constructional elements around the world. The microstructure of samples taken from the steel tank in the water tower in Lower Silesia, Poland, from the late 19th century, is mostly ferritic; locally, there is also some ferrite–pearlite, and the maximum carbon content is 0.07% C [18]. Similar results were obtained from the samples extracted from the railway bridge in the same region and the infrastructure of the Central Railway Station in the region’s capital, Wrocław, all dating to the second half of the 19th century [19]. Puddled iron was also used as a construction material for several bridges in Brazil and Portugal [20] and also as a construction material for the Argentinian Navy ship that sank near the coast of Argentina and was built in the UK in 1880 [21].

In contrast, some European regions retained traditional iron-production technology, similar to Slovenia. Such was the case with the wrought iron rods from the early 19th century, which originated from construction works in Turin, Italy. These rods contained around 0.25% carbon; the distribution of carbon was heterogeneous and based on the very low phosphorus and sulphur content. The rods were most probably manufactured in the indirect process using charcoal and later refined in the finery forges. The microstructure was predominantly ferrite and lamellar pearlite, and also some tertiary cementite was detected [22].

5. Conclusions

The microstructural characterisation was carried out on selected semi-finished products from the first half of the 19th century, and the hardness was measured. It was found that the microstructure in the individual samples varied greatly depending on the location in the sample, suggesting that different steels with different carbon contents were used for the production of semi-finished products by forging. In addition, the cooling rate was higher on the surface than in the core of the semi-finished product, which led to a different microstructure in the individual semi-finished products. It was also determined that the semi-finished products made of high-carbon (hypereutectoid) steels were quenched in water to obtain hard martensite in the microstructure (hardening). However, due to the low carbon potential of the atmosphere during heating, decarburisation of the steel surface occurred. All samples also contain non-metallic inclusions, which were identified as oxides and contain mainly Si and Mn, as well as some other oxide-forming elements such as Mg and Ca. The chemical composition of the inclusions varies only slightly between the different samples. The inclusions are distributed in stringers, with the direction of plastic deformation caused by forging.

The results of analyses have proved that the names for various grades of semi-finished products correspond to the metallographic characteristics of the items. The differences in microstructure, carbon content, and hardness are especially visible in comparison between “iron” and “steel” semi-finished products. The external appearance of the semi-finished products could also be correlated with a specific grade, at least in some cases. For example, a square cross-section of a semi-finished product could be a sign of a semi-finished product made of “steel”.

Due to the scarcity of published metallographic analyses of similar museum objects, analyses of construction elements from approximately the same period were used for the purposes of conducting the comparison. In the 19th century, wrought iron construction material was predominantly manufactured using the puddling process. That resulted in different microstructures of the material and slightly higher proportions of phosphorus and sulphur. An example of construction material analyses that was also produced in the traditional manner derived from northern Italy.