Thanks, Paul, and I completely agree with you that we can learn much about how blades were made long ago by looking at pre-industrial crafts that survive to this day.
I think the important part of hand-finishing is that it does not smear the surface, you can have relatively open or closed grain depending on the way you go about it. If I can organize my thoughts on the subject, I'll send you a pm.
I just ran across a doctoral thesis that looked at the metal in a Merovingian blade (as well as a katana, keris, wootz bar and a couple sabres
) -
Here's the part that ties into this thread -
7.3.1 Bajuwarian Spatha
The fragment of a Bajuwarian spatha stems from the collection of the Landesamt fur Denkmalpege (Munich). The sword was found with others at an ancient burial site close to Aschheim north east of Munich and is dated from the archaeological context to the 6th Century AD, i.e., to the Merowingian period. A part of the fragment was cleaned by the author to get down to the bare metal and then etched for 10 minutes in concentrated sulphuric acid (Figure 7.3). The etching pattern indicated that the sword is built in a sandwich structure, with four bars on each side of a bridge, which is connecting the edges in the centre of the sword. The four bars on each side of the bridge are twisted pair-wise on a length of several centimetres (compare Sachse (1994, p.24)).
Despite the patterned surface, which would indicate two phases, the CEMS spectrum of this piece (Figure 7.9) shows only _-iron with a magnetic field of 33:0T and a gaussian broadening in the outer line of 0:02mm=s. The broadening is about three times smaller than the usual broadenings obtained in backscattering, which is mainly due to the absence of saturation effects in the CEMS method. However it still indicates that this is very pure iron that is practically free of carbon.
In electron micrographs one could clearly see ferritic areas with grain boundaries only and areas with very small inclusions of less than a micrometer linear size. Electron microprobe analysis showed the inclusions to contain reasonable amounts of phosphorus, probably in the form of Fe4P. The amount of these phosphorus containing inclusions is, however, by far too small to be detected in the Mossbauer spectrum. In the electron microscope the inclusions made up about 1/25 of the area of those regions, which showed such inclusions at all. Micro-hardness was measured at about 118(9) for large ferrite grains, 150(6) for ferrite with grain boundaries and 152(8) for ferrite with inclusions containing phosphorus. The overall hardness of the sword would therefore mainly be determined by the size of the ferrite grains, which is also infuenced by the presence of the inclusions. The visibility of the pattern on the surface corresponds to different etching behaviour of the inclusion-free and inclusion-containing areas. Probably steels with different phosphorous concentrations were welded together to obtain the pattern.
From
"Non-Destructive Mossbauer Spectroscopy in Archaeometallurgy"
by Andreas Kyek, a google search should find it as a pdf file...