VSF 2010 Report - Nabo

using an ISIS ABT-55 scanning electron microscope equipped with an Oxford Link
Analytical AN 10/55S energy dispersive spectrometer system (SEM-EDS). Measurements
were performed with a 15kV accelerating voltage with a ≈40% deadtime. All sample
preparations and analyses were performed at the University of Aberdeen.
The precision and accuracy of the instrument used in this study was tested through
repeated analysis of reference materials relevant to the analysis of slag (see Table 1). Three
reference glasses, produced by Brill (1999) from the Corning Museum of Glass were analysed
(Glass B, Glass C and Glass D). The results from recent microbeam analyses of these
standards were employed as the reference values for this study (Vicenzi 2002). The accuracy
and precision tests for the lead-rich reference glass (Glass C) were significantly higher than
those results obtained from the other two glasses, perhaps owing to an internal matrix affect.
Subsequently, it was decided to focus to pay greater attention to the results obtained from
Glass B and Glass D. The results are generally good. Minor elements can be detected and
quantified for amounts over 0.2 %. The relative quantification error for minor element oxides
(1 wt%) the relative error was
around 2-3%. The accuracy of the calibration and validity of the ZAF correction procedure for
SEM-EDS was tested with the latter mentioned reference materials. The calibration of the
scanning electron microscope was based on pure elements and simple compounds. Oxygen
was not measured in this study, but was calculated based on stoichiometry. All results are
reported as averages and expressed in weight percent of the element oxides.
Slag classification and terminology
There is no consensus on how slag should be classified and the terminology employed to
describe those classifications 15 . Due to the variety of slag and residues, reports should
consider each assemblage within a flexible framework that is site specific, as each site can
produce its own specific types (Crew 1996). Different types of slag are produced during the
iron smelting and smithing processes, which can often be differentiated by their morphology,
size, and characteristic features. Density and colour can also be used to differentiate slag
types, but these properties can vary greatly between slags from the same process. Slags from
the bloomery-smelting and smithing process are compositionally very similar, and so this is
not often used to differentiate between the two processes. Slags can also exhibit mixed
characteristics, or intermediate characteristics that make them difficult to relate to a
metallurgical process easily, and so it is worthy to conceive of a continuum of slag types
rather than a division of easily definable types.
Initial efforts to categorise metallurgical residues from Vatnsfjörður found close
parallels to those devised by McDonnell and Maclean (2010) for their archaeometallurigcal
assessment of the residues from Hofstaðir (a Viking Period Hall in north-eastern Iceland).
This report employs a similar terminology and classification system to that employed by
McDonnell and Maclean (2010). Some terms have not been employed due to their ambivalent
meaning. For purposes of clarity and comparability, the terminology and classification system
employed here will be described in detail with reference to previous slag assessments in
Iceland.
The archaeometallurgical residues can be separated into two groups, those that are
15 For a general introduction and outline of slag morphologies (classification systems) that have been utilised in
archaeological studies, the reader can be guided towards the online glossary of terms and definitions produced by the
Historical Metallurgy Society, as well as their catalogue of the National Slag Collection containing exemplary photographs.
The archaeometallurgy guidelines produced by English Heritage provides an informative and useful introduction to slag
types and descriptions.
101