XRF examination of an early Viking Age sword (Petersen type B)

Lee · 17th October 2025, 04:37 PM

Also from my NEBG Ashokan Sword Seminar presentation is this examination of a Petersen type B Viking Age sword (8th to early 9th Century.)

A while after I acquired this sword, a few decades ago, it began shedding iron flakes such that I needed to find a way to conserve it. Jeff Pringle directed me to an academic paper (I do wish I could find it again) with directions that I followed. For several weeks I soaked the sword in alkalinized distilled water, regularly checked and replaced when pH fell. This was to neutralize and remove anions such as chlorine. Once the pH stopped dropping, I then transferred the sword into successive baths of isopropyl alcohol to dehydrate it. Such flakes as separated by lightly wiping with a gloved finger were sacrificed during the dehydration process which ended in acetone. I then coated the sword in Paraloid B72 to seal it.

Below is an overall photograph and a close-up of the sword after this conservation process followed by a map of where XRF measurements were made. The numbers on the map reflect the "test label" that the Olympus Vanta analyzer will include when its interpretations are downloaded. A column was added for 'region' and here the functional region of the sword was added and then used to sort the results by regions that are color coded in the next image. A "fudge factor" was applied to each result to reflect the light elements that the analyzer cannot identify and then averages were calculated by element for each functional location.

Lee · 17th October 2025, 05:00 PM

OK, this is clearly amateur science stuff with all of the limitations that implies, but even so some conclusions may be drawn.

Phosphorus is ten or more times greater in the readings from the central flats of the blade than on the bevels or edges (red arrows in the previous post.) Bands of pattern-welding are expected in a Petersen type B sword and phosphoric iron appears lighter in color and is known to have been used in the layered, twisted rods employed in pattern-welding to increase visual contrast.

Calcium (blue arrows above), on the other hand, is significantly greater at the edges and bevels again strongly suggesting that a different alloy was used in these positions versus the central flats of the blades. Calcium carbonate may be used as a flux in bloomery smelting and perhaps this might explain the difference observed.

Sulfur shows a much lesser difference.

Math could very likely bring out other relevant findings, as could the skill to interpret the actual spectra reported by an instrument calibrated for geological exploration rather than archaeology. But these are the low hanging fruits the amateur can reach.

17th October 2025, 04:37 PM	#1
Lee EAAF Staff Join Date: Nov 2004 Location: Upstate New York, USA Posts: 966	XRF examination of an early Viking Age sword (Petersen type B) Also from my NEBG Ashokan Sword Seminar presentation is this examination of a Petersen type B Viking Age sword (8th to early 9th Century.) A while after I acquired this sword, a few decades ago, it began shedding iron flakes such that I needed to find a way to conserve it. Jeff Pringle directed me to an academic paper (I do wish I could find it again) with directions that I followed. For several weeks I soaked the sword in alkalinized distilled water, regularly checked and replaced when pH fell. This was to neutralize and remove anions such as chlorine. Once the pH stopped dropping, I then transferred the sword into successive baths of isopropyl alcohol to dehydrate it. Such flakes as separated by lightly wiping with a gloved finger were sacrificed during the dehydration process which ended in acetone. I then coated the sword in Paraloid B72 to seal it. Below is an overall photograph and a close-up of the sword after this conservation process followed by a map of where XRF measurements were made. The numbers on the map reflect the "test label" that the Olympus Vanta analyzer will include when its interpretations are downloaded. A column was added for 'region' and here the functional region of the sword was added and then used to sort the results by regions that are color coded in the next image. A "fudge factor" was applied to each result to reflect the light elements that the analyzer cannot identify and then averages were calculated by element for each functional location. Attached Images

17th October 2025, 05:00 PM	#2
Lee EAAF Staff Join Date: Nov 2004 Location: Upstate New York, USA Posts: 966	OK, this is clearly amateur science stuff with all of the limitations that implies, but even so some conclusions may be drawn. Phosphorus is ten or more times greater in the readings from the central flats of the blade than on the bevels or edges (red arrows in the previous post.) Bands of pattern-welding are expected in a Petersen type B sword and phosphoric iron appears lighter in color and is known to have been used in the layered, twisted rods employed in pattern-welding to increase visual contrast. Calcium (blue arrows above), on the other hand, is significantly greater at the edges and bevels again strongly suggesting that a different alloy was used in these positions versus the central flats of the blades. Calcium carbonate may be used as a flux in bloomery smelting and perhaps this might explain the difference observed. Sulfur shows a much lesser difference. Math could very likely bring out other relevant findings, as could the skill to interpret the actual spectra reported by an instrument calibrated for geological exploration rather than archaeology. But these are the low hanging fruits the amateur can reach.