As all the prior responses on this subject have been from researchers in
Europe, I would just note that high-alumina bloomery iron slags are very
commonly encountered in the tropics. Most of the many bloomery iron
slags that I have examined from Africa fall into this category. The
source of the alumina in tropical bloomery slogs is usually laterite
ore. Laterites form as subsoil deposits in climates that have high
annual rainfall AND a long dry season; iron is dissolved by the rains in
upper soil horizons, and is precipitated lower down in the profile as
the soil dries out. Alumina can also be mobilized and precipitated in
the same way, depending upon eH/pH conditions, and there is a continuous
gradation from iron-rich deposits (laterites) to alumina-rich deposits
(bauxites).
In thin section the alumina shows up, as Ingo Keesman has noted, as
hercynite and leucite. (The crystal structure of fayalite cannot
accomodate aluminium ions to any significant extent). The essential
potassium for the leucite comes mostly, I think, from the ash of the
charcoal fuel. Hercynite often forms intricate intergrowths with
fayalite that are not obvious in reflected light (unless DIC is used to
accentuate differences in relief) because the two minerals hve almost
the same reflectivity, but are impossible to miss in transmitted light,
where hercynite is opaque, or very dark green on thin edges.
Alumina can also enter the slag by chemical erosion of the refractory
ceramics. From observation of ethnographic iron smelting at several
locations in Africa, I conclude that most of this comes from erosion of
the tuyeres rather than the furnace shaft wall. If the furnace base is
not protected in some way (as it often is in recent African practice by
a thick layer of sand or ash) the slag will aggressively attack the soil
or clay of the base, giving rise to a transition zone of high-alumina
slag between the upper slag and the unconsumed soil at the base. These
slags are easily recognized in vertical section because of their layered
structure.
Tim Young has done some very elegant modelling of the relative
contribution of ore, fuel and refractory to slag composition in western
Britain, and anyone interested in the subject should read his papers on
the subject.
Dave Killick
|