archaeometallurgy

Monumental bronze doors of the High Middle Ages are most often discussed from an art-historical perspective. Their technical dimension has received far less systematic attention. Two closely related studies address this gap by examining the large bronze doors of Hildesheim, Mainz, and Augsburg from a materials-scientific and casting-technical point of view.

The first study takes a comparative approach to all three bronze doors (Mödlinger et al 2026). Based on alloy analyses and casting-related questions, it discusses material choice, casting strategy, and practical feasibility. The results show that the bronze doors are neither technically uniform nor best understood as isolated exceptional works. Instead, each reflects specific decisions, constraints, and priorities within its casting context.

The second study focuses on the Hildesheim bronze doors and places greater emphasis on process-oriented interpretation (Cziegler et al 2025). Casting simulations are used to explore thermal behaviour, solidification sequences, and potential risk zones during the pour. These models do not replace historical evidence. They help distinguish technically plausible scenarios from unlikely ones and allow discussion of possible workshop practices.

Medieval bronze bronze doors: Methodological approach

The methodological approach is central to both studies. The aim is not to describe medieval casting processes as if they were directly observed, nor to project modern metallurgical concepts back onto the Middle Ages. Instead, the focus lies on identifying the technical concepts, experiential knowledge, and problem-solving strategies that could realistically have been available at the time.

The analysis proceeds deliberately from a chronological bottom-up perspective. This is not meant in a hierarchical sense, but as an attempt to understand how casting practice developed through material constraints, process limitations, and workshop routines. The resulting interpretations are therefore necessarily plausible rather than definitive. They are intended to narrow the range of interpretation, not to close it.

Collaboration with Gates of Paradise Project

Both papers were developed in the context of a close collaboration with the GAPAMET project (Gates to Paradise), which focuses on the interdisciplinary study of medieval bronze bronze doors. Over the past two years, our laboratory has worked closely with the team led by Marianne Mödlinger, the project’s principal investigator. This collaboration provided the framework for many of the materials-science and casting-related questions addressed in these studies.

More information on the project:
https://www.gapamet.imareal.sbg.ac.at/en/

A related post from this collaborative context,  led to a separate paper (Asmus et al 2025) on specific workshop practices and means of replicating (wax) models.

Articles

The articles are available here:
https://doi.org/10.1007/s40962-025-01857-4
https://doi.org/10.1007/s40962-025-01820-3

References

Background

Early iron metallurgy of the first millennium BCE is often described either as technically immature or as a sharp break from Bronze Age practice. Both views are too coarse. Our new study of a bloomery steel chisel from Rocha do Vigio shows a more gradual development (Asmus et al 2026).

A composite mesoscopic image of the sampled bloomery steel chisel-tip. © 2025 Bastian Asmus

The artefact

The object was made for a bloomery iron and is dated to the ninth century BCE.  In an earlier study, we characterised the metal as bloomery steel and determined its carbon content. Quantifying carbon in hardened material is only partly reliable. For this reason, the first study focused on the body of the tool. The cutting edge was not examined at that stage (Araque Gonzalez et al 2023).

Based on a carbon content of about 0.5 wt% C and the early date of the artefact, we later investigated the tip itself. The aim was to assess whether, and to what extent, it had been thermally treated of it was  hardened at all.

Bloomery steel: Microstructure and hardness

Metallography of the cutting edge shows a homogeneous and very fine pearlitic to pearlitic-bainitic microstructure. Ferrite is present only in small amounts. Martensite is absent. This points to accelerated cooling, but not to full quenching in the modern sense.

Secondary electron image of the chisel tip close to the cutting edge, showing mostly very fine pearlite, with some upper bainite in between the feathery colonies of the fine pearlite. Image: Asmus.

Vickers microhardness measurements show a moderate hardness gradient between the softer body and the refined tip. The values are consistent with controlled thermal treatment during forging. There is no indication that maximum hardness was the goal.

Alloy chemistry

The bloomery steel is low in manganese, as expected for early iron. Its hardenability therefore differs strongly from that of modern steels. Many commonly used transformation models are based on modern reference compositions. Our results show that these models are only of limited use for early iron artefacts. More accurate transformation data for low-Mn systems are – surprisingly – still lacking.

Production context

Slags from the site confirm local primary iron production. The chisel is part of a regional metallurgical practice. It is not an imported or exceptional object.

Taken together, the evidence points to a deliberate transfer of Bronze Age thermal working strategies to iron. Early iron metallurgy in this case reflects continuity of skill rather than a technological breakthrough.

The full article is available here:

https://doi.org/10.1016/j.jmrt.2026.01.091

References

Just out now: A paper that summarises several years worth of medieval bell casting projects  (Asmus 2023). It is part with my ongoing research focus on the development of metal casting in the medieval period, roughly encompassing the period between AD 800 and 1600.

The paper is open access and can be accessed here.

Abstract

This paper presents key results from experimental work on traditional bronze casting, focusing on early to high medieval bells, roughly between the 8th and 12th century AD . It demonstrates that combining craft, historical sources, and modern science can effectively revive lost technologies. The reconstruction is based on Theophilus Presbyter’s Schedula Diversarum Artium, dated to the early 12th century, whose precise instructions were critical, though several field-scale experiments were required to refine the process.

These experiments are part of broader research into medieval large-scale casting methods in central Europe. The paper argues that more than experimental archaeology or traditional craft is needed to understand and recreate lost technologies. Researchers must invest significant time mastering materials, tools, and techniques to grasp craft processes fully. Brief experimental engagements fail to capture the depth of these traditions. This approach bridges archaeology and hands-on practice, challenging conventions in both traditional craft and mainstream academia.

Literature