piratediver
Sr. Member
From Norm Judd. Would metallurgic testing of this spike prove country and era of origin? ....If such
tests exist - where are these?
Not necessarily in fact lots of mistakes can arise.. Here is an excerpt from my cha on this subject in Ship
fastenings.. From sewn boat to steamship... TAMU press.. There is also a cha on treenails there.. You can
find the work in quite a number of libraries..
Metallurgical analyses
Given the rapid technological advances in copper bolt manufacture during the 1780s, it is not impossible
that brass bolts of varying copper/zinc content were produced in the process of experimentation.
Hence, the percentage ratio of these two metals may not be a reliable indicator for the archaeological
dating of these fastenings.
Certainly where fastenings are found out of context and in an environment where contamination is
possible, they are not a reliable tool in site dating or identification.
. . . . in the very ability of metallurgists to extremely accurately measure the constituents of
archaeological metals, we are necessarily lead to the need to consider the effects of a phenomenon
known as the Œselective corrosion¹ of zinc in seawater. This is commonly called Œde-zincification¹, and
it is a process that leaves a rough porous surface on what externally appears to be an almost pure
copper fastening, but is actually a copper/zinc alloy. In his study of the man-of-war of the period, Peter
Goodwin notes that while Muntz Metal was used as a substitute for bronze fastenings this process
rendered them Œdevoid of zinc and honeycombed, rendering it weak and useless¹.
A stark example of this appears with an analysis of sheathing found during Shirley Strachan¹s excavation
at the wreck of the part iron-fastened, copper-sheathed Sydney Cove (1794-1797). There the
composition of a copper sheathing fragment was analysed at
CopperŠŠ. 95.00%
LeadŠŠŠŠ0.50%
ZincŠŠŠ.. 0.03%
TinŠŠŠŠ.. 0.20
IronŠŠŠ.. 0.02%
TOTALŠŠ 95.75
This leaves c. 4% of the original constitutent metal(s) unaccounted for and in querying the discrepancy,
Strachan was advised by her metallurgists that Œallowance needs be made for the leaching of zinc and
tin which occurs in a corrosive chloride environment of the sea¹.[iii] <#_edn3>
Yet again there is much more to this problem than meets the eye! In a recent study, corrosion scientists
Vicki Richards and Ian MacLeod have shown that there can be startling differences between the
metallurgical samples collected from an exposed aerobic area on any one particular fastening and those
from a low oxygenated environment from the same fastening. Their study showed that there are three
areas of metallurgical interest on any one shipwreck fastening.
i) the part fully exposed to the sea,
ii) the part completely buried in the parent timber and
iii) at the interface between i and ii above .
Using a fastening (HA 30056) from the German-built wooden-hulled barque Hadda (1860-1877) these
researchers found that the metal sampled from the aerobic area showed 72% copper, 27% zinc and 1%
lead, while for the sample from the anaerobic region, the results were 84%, 15% and 1% respectively. As
a result, it was concluded that the de-zincification process occurs more readily under anaerobic
conditions and is accelerated by high temperatures, high chloride contents of water and low water
speeds.[iv] <#_edn4> In the second case (HA30057), a copper impregnated wood sample from which a
copper alloy fastening had been removed was also tested. First it was divided into HA30057.TOP., i.e. a
sample recovered near the head of the fastening and HA30057.BOT [bottom], in this case a sample
collected lower on the shank of the bolt. The percentages of copper, zinc and lead in the sample
collected from the aerobic region were 71%, 29% and < 0.5% and from the deoxygenated area were
66%, 34% and < 0.5% respectively. This led them to conclude that the de-zincification corrosion
processes that occurred beneath the wood of the HA30056 brass fastening were not occurring on the
HA30057 bolt, indicating a more oxidising micro-environment beneath the wood.
All this indicates that the historian and archaeologist are dealing with complex phenomena
and‹somewhat ironically‹the metallurgical analysis of fastenings recovered from the sea can sometimes
present as many problems as they solve.
Another possibility to be considered by the archaeologist seeking metallurgical analyses is that there can
also be Œcompositional variations¹ within any single fastener. These are possibly the result of particular
casting techniques and differential heating or cooling rates of the metal. A case in point are the
fastenings from the French-built ex slaver James Matthews (?-1841) and the American China Trader
Rapid (1804-1811) built at Braintree Massachusetts. There Ian MacLeod and Maria Pitrun¹s analysis of a
square-section spike from the Rapid (RP3074) showed that it was a leaded arsenical copper with small
amounts of tin, antimony, silver and bismuth and traces of iron and zinc. Further, their analyses indicate
that the spike was fabricated from a long rod, which was then either hot or cold-worked (probably
rolled) and then cut the required length.
A spike RP3373A from the same wreck was a mixed copper/tin/zinc alloy that they found was similar to
a Œheavy-leaded bronze¹. In not having been mechanically worked or heat treated, they deduced that it
appeared to be part of the carpenter¹s shipboard stores. Spike RP 3373B was a Œleaded-zinc bronze¹
that displayed evidence of Œimperfect casting technology¹ and Spike
5004/T11 was found in its Œas cast state¹ not having been mechanically worked or heat treated¹. Here
evidence of shrinkage and gas porosity provided to MacLeod and Pitrun some Œproof of imperfect
foundry practice¹.[v] <#_edn5> So too with their analysis of a copper spike from the James Matthews,
JM 160/T8. It showed evidence that it¹s shaping was performed by Œcold hammering¹ and that some
provided evidence of Œimproper working¹.[vi] <#_edn6>
Further to what can be learned from such analyses, the metallurgist L.E.
Samuels also examined this material and he indicated that the microstructure revealed the James
Matthews fastenings had been Œwrought from a cast ingot, and they conceivably could have been
either hot worked or cold worked with intermediate annealings¹. He deduced that another spike had
been Œfabricated from a cast bar by cold-reducing part of its length to form a shank, leaving a head in
the as-cast condition¹.[vii] <#_edn7>
Notwithstanding what can be learned in respect of the mode of working and the composition of the
metals under scrutiny, when dealing with fastenings, clearly we cannot therefore assume that the
composition of the fastening found on any wreck is Œas forged¹ or Œas driven¹, partly as a result of
Œdezincification¹ and partly as a result of casting and other differences.
Nor can we now assume that the composition of any particular fastening is uniform throughout!
Pirate Diver
tests exist - where are these?
Not necessarily in fact lots of mistakes can arise.. Here is an excerpt from my cha on this subject in Ship
fastenings.. From sewn boat to steamship... TAMU press.. There is also a cha on treenails there.. You can
find the work in quite a number of libraries..
Metallurgical analyses
Given the rapid technological advances in copper bolt manufacture during the 1780s, it is not impossible
that brass bolts of varying copper/zinc content were produced in the process of experimentation.
Hence, the percentage ratio of these two metals may not be a reliable indicator for the archaeological
dating of these fastenings.
Certainly where fastenings are found out of context and in an environment where contamination is
possible, they are not a reliable tool in site dating or identification.
. . . . in the very ability of metallurgists to extremely accurately measure the constituents of
archaeological metals, we are necessarily lead to the need to consider the effects of a phenomenon
known as the Œselective corrosion¹ of zinc in seawater. This is commonly called Œde-zincification¹, and
it is a process that leaves a rough porous surface on what externally appears to be an almost pure
copper fastening, but is actually a copper/zinc alloy. In his study of the man-of-war of the period, Peter
Goodwin notes that while Muntz Metal was used as a substitute for bronze fastenings this process
rendered them Œdevoid of zinc and honeycombed, rendering it weak and useless¹.
A stark example of this appears with an analysis of sheathing found during Shirley Strachan¹s excavation
at the wreck of the part iron-fastened, copper-sheathed Sydney Cove (1794-1797). There the
composition of a copper sheathing fragment was analysed at
CopperŠŠ. 95.00%
LeadŠŠŠŠ0.50%
ZincŠŠŠ.. 0.03%
TinŠŠŠŠ.. 0.20
IronŠŠŠ.. 0.02%
TOTALŠŠ 95.75
This leaves c. 4% of the original constitutent metal(s) unaccounted for and in querying the discrepancy,
Strachan was advised by her metallurgists that Œallowance needs be made for the leaching of zinc and
tin which occurs in a corrosive chloride environment of the sea¹.[iii] <#_edn3>
Yet again there is much more to this problem than meets the eye! In a recent study, corrosion scientists
Vicki Richards and Ian MacLeod have shown that there can be startling differences between the
metallurgical samples collected from an exposed aerobic area on any one particular fastening and those
from a low oxygenated environment from the same fastening. Their study showed that there are three
areas of metallurgical interest on any one shipwreck fastening.
i) the part fully exposed to the sea,
ii) the part completely buried in the parent timber and
iii) at the interface between i and ii above .
Using a fastening (HA 30056) from the German-built wooden-hulled barque Hadda (1860-1877) these
researchers found that the metal sampled from the aerobic area showed 72% copper, 27% zinc and 1%
lead, while for the sample from the anaerobic region, the results were 84%, 15% and 1% respectively. As
a result, it was concluded that the de-zincification process occurs more readily under anaerobic
conditions and is accelerated by high temperatures, high chloride contents of water and low water
speeds.[iv] <#_edn4> In the second case (HA30057), a copper impregnated wood sample from which a
copper alloy fastening had been removed was also tested. First it was divided into HA30057.TOP., i.e. a
sample recovered near the head of the fastening and HA30057.BOT [bottom], in this case a sample
collected lower on the shank of the bolt. The percentages of copper, zinc and lead in the sample
collected from the aerobic region were 71%, 29% and < 0.5% and from the deoxygenated area were
66%, 34% and < 0.5% respectively. This led them to conclude that the de-zincification corrosion
processes that occurred beneath the wood of the HA30056 brass fastening were not occurring on the
HA30057 bolt, indicating a more oxidising micro-environment beneath the wood.
All this indicates that the historian and archaeologist are dealing with complex phenomena
and‹somewhat ironically‹the metallurgical analysis of fastenings recovered from the sea can sometimes
present as many problems as they solve.
Another possibility to be considered by the archaeologist seeking metallurgical analyses is that there can
also be Œcompositional variations¹ within any single fastener. These are possibly the result of particular
casting techniques and differential heating or cooling rates of the metal. A case in point are the
fastenings from the French-built ex slaver James Matthews (?-1841) and the American China Trader
Rapid (1804-1811) built at Braintree Massachusetts. There Ian MacLeod and Maria Pitrun¹s analysis of a
square-section spike from the Rapid (RP3074) showed that it was a leaded arsenical copper with small
amounts of tin, antimony, silver and bismuth and traces of iron and zinc. Further, their analyses indicate
that the spike was fabricated from a long rod, which was then either hot or cold-worked (probably
rolled) and then cut the required length.
A spike RP3373A from the same wreck was a mixed copper/tin/zinc alloy that they found was similar to
a Œheavy-leaded bronze¹. In not having been mechanically worked or heat treated, they deduced that it
appeared to be part of the carpenter¹s shipboard stores. Spike RP 3373B was a Œleaded-zinc bronze¹
that displayed evidence of Œimperfect casting technology¹ and Spike
5004/T11 was found in its Œas cast state¹ not having been mechanically worked or heat treated¹. Here
evidence of shrinkage and gas porosity provided to MacLeod and Pitrun some Œproof of imperfect
foundry practice¹.[v] <#_edn5> So too with their analysis of a copper spike from the James Matthews,
JM 160/T8. It showed evidence that it¹s shaping was performed by Œcold hammering¹ and that some
provided evidence of Œimproper working¹.[vi] <#_edn6>
Further to what can be learned from such analyses, the metallurgist L.E.
Samuels also examined this material and he indicated that the microstructure revealed the James
Matthews fastenings had been Œwrought from a cast ingot, and they conceivably could have been
either hot worked or cold worked with intermediate annealings¹. He deduced that another spike had
been Œfabricated from a cast bar by cold-reducing part of its length to form a shank, leaving a head in
the as-cast condition¹.[vii] <#_edn7>
Notwithstanding what can be learned in respect of the mode of working and the composition of the
metals under scrutiny, when dealing with fastenings, clearly we cannot therefore assume that the
composition of the fastening found on any wreck is Œas forged¹ or Œas driven¹, partly as a result of
Œdezincification¹ and partly as a result of casting and other differences.
Nor can we now assume that the composition of any particular fastening is uniform throughout!
Pirate Diver
