Mechanical Behaviour of Copper Conductors
in Relation to Fire Investigation
Dr. Bernard Béland
Béland, Bernard. Mechanical behaviour
of copper conductors in relation to fire investigation.
Fire and Arson Investigator. Vol 47 No 4 (June 1997). p 8-9.
[interFIRE VR note: Figures referred to in the text have
not been reproduced in this reprint.]
After a fire, all kinds of broken copper wires are found in the debris.
At the point of severance, the ends could show numerous forms such as plastic
deformation, fragile fracture, pointed ends, beads and numerous other characteristics.
The break could be due to the heat of the fire, a very high current or a
combination of these, as well as mechanical fractures, either at fire temperature
or at near room temperature, with or without current. In this article, our
interest is limited to mechanical breaks. Melting because of high currents,
arcing or from the heat of the fire is excluded.
Numerous No. 14 copper wires were broken under a mechanical force while
they were subjected to different temperatures. After the fracture, the ends
were examined under a microscope. Electricity was used to heat the wire.
Using heat transfer theory, the operating temperature can be found if the
current is known. The current was adjusted for the wire to reach a given
temperature. Then, the current was switched off and the wires were broken
immediately under tension. The heating time was only a matter of a few minutes.
Obviously, the heat could have been provided by any other means with the
same results. Electricity was just a convenient way to provide the heat.
Some wires were also heated with a torch.
When a wire is stretched at room temperature, the section slightly decreases
while the length increases. Eventually, there is a localized pinching in
one area and the wire breaks at that point. Figure 1a shows such a fracture.
This is a ductile break and most soft materials break that way. That includes
soft iron, copper and aluminum. At the severance point, there is a small
crater shown by the dotted line.
Figures 1a to 1e show the breaks obtained at different temperatures.
As can be seen in figure 1, from room temperature to about 500°C (930°F),
the break is similar, although the necking increases in length and the size
of the wire at the severance point is smaller and smaller as the temperature
is increased. As the break happens at higher and higher temperature, this
effect is more pronounced. Eventually, at around 1000°C (1830°F),
the wire behaves like taffy. The necking is quite long and, at the break,
the diameter is very small. It looks like a pointed needle. However, if
one looks closely at the end, there is a small rugged surface at the break
that can be seen with a microscope. This is shown in Figure 1d.
At still higher temperatures, the breaks were fragile as shown in Figure
1e. A close examination of the break surfaces shows some surface melting.
The temperature at which fragile breaks happen is around 1040°C (1900°F)
or higher. Copper melts at 1083°C (1981°F). All of the above breaks
were produced while the metal was still at the given temperature. It is
very difficult to measure the temperature of a small wire such as that of
No. 14 AWG. It is recalled that the indicated temperatures given above were
not measured but, rather, were computed from heat transfer theory for a
given current. The given temperatures are accurate within about 5 percent.
The shape of the break should be used only as a general guide. It should
be clear that the type of break depends not only on the temperature but
also on the time that the wire was at that temperature.
Looking at the type of break, one may have some indication of the temperature
of the wire at the time of the break. However, one should be aware that
the behavior of metals at high temperatures is complex. For example, a copper
wire could be heated to a level close to the melting temperature. Then oxidation
could happen between the copper grains. An alloy of copper and copper oxide
could then be formed between the grains. This wire could also produce a
fragile break, even if broken after the wire has been cooled to room temperature.
This phenomenon has often been observed by fire investigators. That break
resembles that of Figure 1e. This phenomenon depends on the temperature
reached and the time as well as the type of fire gases around the wire.
Fragile wires are often observed after a fire. This is due to the oxidation
between the copper grains. It just implies that the wire was subjected to
high temperatures for some time under an oxidizing atmosphere.
When a cold copper conductor is stretched and broken while carrying a
current, a parting arc is established at the separating point. It is sometimes
thought that the arc will melt the ends of the broken wire. This was tested.
A 100-A current was applied to a No. 14 AWG copper conductor. Then the conductor
was stretched and broken quickly (within a second). In that time, the conductor
temperature rise is about 12°C and the wire can still be considered
at room temperature. The break is obviously like that of Figure 1a since
the conductor is cold. That test was repeated numerous times with and without
the plastic insulation. The ends of the breaks were examined under a microscope.
The ends often showed no evidence of arc melting. In other instances some
very minimal melting could be seen under a microscope. In all cases, most
of the broken surfaces were still evident without any melting. The amount
of melting obviously depends on the time on the cycle that the break happens.
This result may be surprising since one may expect a significant arc
to take place with 100-A. However, this is not the case. The parting surfaces
separate at a very high speed. From wave theory and mechanics of materials,
it can be shown that each surface, at the instant of separation, moves in
opposite directions at a speed given by:
v=square rootY/p (1)
in which v is the speed
Y is the Young's modulus, and
p is the density of the material.
When this equation is applied to copper, aluminum and steel, it is found
that the speed of each surface is 12,000, 17,500 and 17,000 feet per second,
respectively. The relative speed of the two surfaces is obviously doubled.
Therefore, the parting speed for copper is 24,000 feet per second.
That speed is independent of the conductor size. At 120-V, it is assumed
that the arc will stop when the distance between the parting surfaces is
0.1 inch. Then, the arc stops in about 0.3 microsecond. That time is so
small that the space between the parting surfaces do not have time to heat
up. The space being relatively could, the arc is likely to be extinguished
in an even shorter time and a smaller length than assumed. The total energy
spent in the arc is, at the most, equals to 120 x 100 x 0.0000003 = 0.0036
Joule. This is the energy spent in a 100-W light bulb lit for 0.000036 second.
For comparison purposes, breakers and fuses at the 15-A level take about
0.0-1 to 0.1 second to open with 100-A.
The above analysis is valid at room temperature. As the temperature is
higher and higher, the Young's modulus is reduced and so is the separating
speed. However, that speed is not appreciably reduced until the temperature
of 500°C (930°F) is reached. Eventually, when the temperature is
such as to render the wire very plastic and behaving like that shown in
Figure 1d, the separating speed will be much smaller than calculated above.
The arc will then last for a longer time. In tests that were run, when the
wires were broken with 100-A and at close to the melting temperature of
copper, then surface melting on the break surfaces was evident.
The above analysis is strictly valid for purely resistive circuits. If
the circuit is inductive, then the time to open the circuit will be increased.
The analysis is much more complex. Still, the opening time will be quite
small since the separating speed is high.
In this article, some phenomena that occur at the separating surfaces
of a copper wire that is broken were studied briefly. Often, the shape of
the ends of the conductors could indicate what happened. However, one must
be careful in the interpretation of the breaks. It was shown, both from
theory and tests, that a wire that is broken while carrying a current does
not show much meltingif anyat the separating surfaces.
About the Author
For the past 30 years, Dr. Béland has studied fires under laboratory
conditions and also at fire scenes. Many of the fires, including full size
fires in buildings, were started intentionally to study their behavior.
Dr. Béland specializes in the study of ignition, thermal transfer
and electrical causes. Many systems and devices were used in his experiments
to study their outcome and to determine what types of damage could be associated
with the causes that resulted in the fire. He has also experimented with
numerous systems to determine under which conditions they could constitute
Dr. Béland's research has resulted in the printing of over 100
technical articles in specialized journals such as: L'Ingenieur, Fire Technology,
Journal of Forensic Sciences, Fire & Arson Investigator, Power Apparatus
and Systems, Electrical Business, Proceedings of the Institute of Electrical
Engineers and others.
Dr. Béland has investigated over 900 fires and electrical failures
in which a total of 300 lives have been lost. He is retained by equipment
manufacturers, power companies and research centers. Dr. Béland has
done consulting work and lecturing in eight Canadian Provinces, 37 states
throughout the U.S., four European countries and New Zealand. He has served
as an expert witness in approximately 100 cases for numerous jurisdictions.
Dr. Béland has taught at numerous universities in Canada. He recently
retired from the Universite de Sherbrooke as a Professor in the Department
of Electrical Engineering. Dr. Béland is currently a private consultant
in his own firm.
Reprinted with permission.