An Encyclopedia of the History of Technology part 16 pps

PART ONE: MATERIALS
132
the noble metal palladium in the soluble fraction was demonstrated by Wollaston
in 1803, and he announced the discovery of another noble metal, rhodium, in
the following year. Ruthenium, the sixth and last member of the platinum group
was identified by Claus in 1844.
Wollaston’s process of platinum production was probably the first to utilize
a chemical extraction and refining process so controlled that it produced pure
fine metallic powder which was specifically intended for consolidation by the
powder metallurgical method. He soon became well known for the high
quality and ductility of the platinum he produced in this way. In the early
years of the nineteenth century, large quantities of platinum were needed for
the construction of items of chemical plant such as sulphuric acid boilers.
Between 1803 and 1820, Wollaston produced about 7000 ounces (197kg) of
ductile platinum for sulphuric acid boilers alone.
Platinum deposits were discovered in the Ural mountains in Russia in 1819
and rich alluvial platinum deposits were found in 1824. In 1825, deposits at
Nizhny Tagil north of Ekaterinburg (Sverdlovsk) were found to yield 100
ounces of gold and platinum per ton (3.6kg per tonne) of gravel, and Russian
platinum continued to satisfy the world’s demands until 1917, when political
changes encouraged the extraction of this metal from the nickel deposits at
Sudbury, Ontario (see p. 98). In 1924 platinum was discovered by Dr Hans
Merensky in a South African reef near Rustenburg in the Transvaal, now
known to be one of the world’s richest platinum deposits. Vast quantities of
copper, nickel and cobalt, and also gold and silver, are also associated with the
rich metalliferous vein.
The very large cupro-nickel ore deposits discovered at Jinchuan in Ganzhou
Province in China in 1958 also contain considerable quantities of both gold
and platinum. Rich palladium-platinum deposits have also been found in the
Mitu area of Yunan Province which is only about 80km (50 miles) away from
the nearest railway line and is therefore capable of being rapidly developed.

Powder metallurgy in electrical engineering
The first critical metallurgical requirement of the electrical engineers was a
lamp filament which was more robust and could run at higher temperatures
than the carbon filament of Edison (see Chapter 6). By the end of the
nineteenth century it was appreciated that the light emitted by an incandescent
filament varied as the twelfth power of its temperature. The most efficient
lamp, therefore, was that which could operate at the highest possible
temperature, and attention was concentrated upon the melting points of the
most refractory metals then known. These melting points reached a peak at
tungsten. Rhenium and technetium were unknown at that time and hafnium
and zirconium had never been made in pure metallic form.
NON-FERROUS METALS
133
Osmium, however, was known to have a high melting point and could
readily be produced as an exceedingly pure finely divided powder. The use of
osmium as a light filament material was first proposed in 1898 by Welsbach,
ironically the inventor of the incandescent gas mantle.
The ‘Osmi’ lamp, first produced in Berlin, although more economical in
operation than the carbon lamp was more expensive and rather delicate. It was
soon overtaken by the tantalum lamp which was introduced by Siemens and
Halske of Charlottenburg in 1903. Although the melting point of osmium,
3050°C, was marginally higher than that of tantalum at 2996°C, the latter
metal had the tremendous practical advantage of being ductile so that it could
readily be drawn into fine wire. Ductile tantalum was first produced in 1903
by W. von Bolton. The gaseous contaminants in the nominally pure tantalum
sponge were removed by melting buttons of the metal under high vacuum
conditions. The prototype vacuum furnace used by von Bolton for this work is
shown in Figure 1.12. These arc furnaces had water-cooled metal hearths, and
were so successful that scaled-up versions were designed and used by Kroll
during the Second World War for melting titanium (see p. 143). The tantalum

lamp met with great and virtually instantaneous success. Between 1905 and
1911 over 103 million tantalum lamps were sold.
Tungsten, which melted at 3422°C, was of course the ultimate lamp
filament material, although tremendous practical difficulties were initially
encountered in producing ductile tungsten wire. The first tungsten filaments,
produced in 1904 by a squirting process similar to that used for the osmium
filaments, produced a great deal more light than carbon filament lamps, but
they were brittle and expensive. The process was developed by Just and
Hanamann in 1904 at the Royal School of Technology in Vienna and lamps
were produced in Britain from 1908 in the Hammersmith Osram-GEC Lamp
Works, although the operation was never a commercial success.
Ductile tungsten filaments were first produced by W.D.Coolidge of the US
General Electric Company at Schenectady in 1909. The process he patented in
1909 is still used. Fine tungsten powder was pressed into bars about one square
inch in cross-section. These bars were then sintered by heating them electrically
in a pure hydrogen atmosphere to temperatures approaching their melting point.
They were then reduced into wire by a process of hot fabrication within a well
defined range of temperatures so that a fibrous structure was gradually
developed within the rod or wire. See Figure 1.13. The Coolidge tungsten wire
process was the first to make extensive use of the newly developed rotary
swageing machine. Swageing was continued until the bar had been reduced to
rod about 0.75mm in diameter. This was then hot drawn to wire through
diamond dies. By 1914, over 100 million lamps had been sold in the United
States alone, and manufacturing licences had been granted by the US General
Electric Company to most of the developed world. The trade name Osram, first
used by the Osram Lamp Works in Berlin, derives from the words osmium and
PART ONE: MATERIALS
134
wolfram. After 1909 it became associated with those lamps using drawn tungsten
filaments made in accordance with GE patents.

After about 1913 ductile tungsten, and then ductile molybdenum, became
the dominant materials of the electrical industry. When tungsten contacts were
introduced the reliability of motor-car ignition systems improved considerably.
The Coolidge X-ray tube, introduced in 1913, had a tungsten target and
was the first to permit the long exposures required by the new technique of X-
ray crystallography.
Figure 1.12: The vacuum arc furnace first used in 1902 for the production of
ductile tantalum. The pressed powder pellet, weighing 80–100g was melted on a
water cooled nickel hearth by a direct current arc of 50–300 amperes. During
melting, pressure in the furnace chamber increased from 5 × 10
-3
to 5 × 10
-2
mm
of Hg. This process was devised by Dr W. von Bolton of the Charlottenburg
lamp factory of Siemens and Halske. He was assisted in this development by the
engineer Otto Archibald Simpson and also by Dr M.Pirani who later invented
the thermal conductivity vacuum gauge.
Between 1903 and 1912 more than 60 million tantalum lamps were produced at
Charlottenburg. For this about one ton of vacuum melted tantalum was processed.
The furnace diagram shown here was first published by Dr Pirani in 1952.
Courtesy Vacuum.
NON-FERROUS METALS
135
Between the wars, most of the serious metallurgical research in Great Britain
was undertaken by firms such as Metropolitan Vickers, GEC, Telcon, Standard
Telephones and the major cable companies. Bell Telephones, the US General
Electric Company and Westinghouse exercised a similar influence in the United
States. The influence of Siemens and Heraeus in Germany, and the Philips
Laboratories at Eindhoven was equally profound. The healthiest and most

vigorous offspring of this marriage between metallurgy and electrical technology
was the sintered carbide industry which must, by any standards, be regarded as
one of the major metallurgical innovations of the twentieth century.
SINTERED CARBIDE CUTTING TOOLS
Tungsten wire had to be hot drawn, and the deficiencies of the steel dies which
were first used to reduce the diameter of the hot swaged tungsten rod to the
dimensions of wire which could be handled by diamond dies soon became
apparent. Tungsten carbide seemed to fulfil most of the characteristics of the
material required. The extremely hard carbide W
2
C had first been prepared by
Moissan in 1893 when he fused tungsten with carbon in his electric furnace.
Sintered carbides were first produced in 1914 by the German firm of
Voigtländer and Lohmann. These compacts, which were rather brittle, were
produced by sintering mixtures of WC and W
2
C at temperatures close to their
melting point.
As Moissan had shown, tungsten carbide could be melted and cast from an
electric arc furnace. Cast tungsten carbides dies were produced by this method
before 1914 by the Deutsche Gasglühlicht Gesellschaft, although the cast
product had a very coarse grain size and the dies were again very brittle. By
1922 Schröter of the Osram Lamp Works in Berlin had shown that tungsten
carbide, sintered in the presence of a liquid cement could be very tough. The
three metals, iron, cobalt and nickel all provided a satisfactory molten cement,
although cobalt provided the best combination of hardness and toughness.
The new sintered alloy was rapidly adopted in Germany for the
manufacture of wire drawing dies and between 1923 and 1924 was used and
sold by the Osram Group under the trade name Osram Hartmetall for wire
drawing and also for cutting tools. Friedrich Krupp AG of Essen were granted

a manufacturing licence for this exciting new sintered product in 1925 and
introduced the first successful sintered cutting tools to the world in 1927. Widia
(wie Diamant =like diamond) consisted of tungsten carbide powder sintered
together with a cement consisting of about 6 per cent cobalt. It was exhibited
in 1927 at the Leipzig Fair, where its ability to machine cast iron at unheard-of
speeds was demonstrated. A.C.Wickman of Coventry acquired the sole rights
to import Widia and sell it in the United Kingdom. In association with
Wickman, Krupps started to manufacture tungsten carbide in Britain in 1931.
PART ONE: MATERIALS
136
Figure 1.13
NON-FERROUS METALS
137
Krupp, originally the major shareholder in the Tool Manufacturing Company,
eventually became the sole owner. When the Second World War started in
1939, A.C.Wickman took over control of the factory and changed its name to
Hard Metal Tools Ltd. Manufacturing licences to produce sintered carbides
had by the mid-1930s been granted by Krupp to British firms such as BTH
Ltd, Metro Cutanit, Firth Brown Tools and Murex Ltd.
The American General Electric Company acquired the sole American rights
to the Widia process in 1928 and issued many manufacturing licences. From
this operation emerged nearly all the well-known proprietary grades of carbide.
In order to circumvent the German patents, firms such as Fansteel attempted in
1932 to introduce tantalum carbide tools which were sintered with nickel
rather than cobalt. Such materials were inferior in toughness to the cobalt
bonded composites and were never very successful.
Other powder metallurgical innovations
Other products of the powder metallurgy industry, which began to develop very
rapidly after 1910, are too numerous to describe exhaustively. Sintered porous
self-lubricated bronze bearings were introduced by the American General Electric

Company in 1913, and this was followed by sintered metallic filters, first
produced by the Bound Brook Oilless Bearing Company in 1923. Powder
metallurgy was also used to produce a variety of magnetic alloys, since it was
found that fine powders of metals such as iron, cobalt and nickel could be
cheaply produced in a high state of purity by chemical methods. Such powders
could be fully or partly consolidated into any desired shape by powder
metallurgical methods without the contamination inevitably associated with
conventional melting and alloying procedures. Permalloy, a very soft ironnickel
magnetic powder, was developed in 1928 by the Bell Telephone Laboratories. In
the 1940s when the domain theory of magnetization suggested that very fine
powders could be used to produce permanent magnets of very high coercive
force some very elegant powder metallurgical techniques were developed.
Figure 1.13: The ductile tungsten wire, first produced by Coolidge of the General
Electric Company of Schenectady in 1909 was obtained from bars of tungsten
sintered in apparatus similar to that illustrated. After a preliminary low
temperature sintering process, pressed bars of tungsten powder about in square
crosssection were gripped between water-cooled copper electrodes. The lower
electrode floated on a bath of mercury to allow for the considerable shrinkage
which occurred as the bar sintered. The apparatus was then capped by the water
cooled copper bell shown, and all air purged from the vessel by a flowing current
of hydrogen. Sintering was accomplished by passing a heavy alternating current
through the bar so that its temperature was raised to about 3300°C. Bars thus
sintered were then hot worked by rotary swageing and finally drawn to wire in
diamond dies.
PART ONE: MATERIALS
138
Powder metallurgy also made it possible to produce, merely by mixing the
appropriate powders, a whole range of composite materials which could have been
manufactured in no other way. In this way a whole generation of new electrical
contact materials was developed between the wars. These included, for example,

composite contacts which incorporated insoluble mixtures of silver and nickel,
silver and graphite, silver-tungsten, copper tungsten and, probably the most
important, silver-cadmium oxide. The last formulation is still widely employed
because it combines the low contact resistance of silver with the ability of cadmium
oxide to quench any arcs formed when the contacts open under load.
The compaction of powdered metals
As in the time of Wollaston, metal powders are still compacted in steel dies
since this is generally the cheapest and most convenient method of producing
large quantities of components on a routine basis. The technical limitations
imposed by die wall friction are generally accepted and various expedients
have been devised to obtain more uniformly pressed components and to
reduce the incidence of interior cracking caused by concentrated internal
stresses in the compact.
Considerable difficulties were encountered, however, in the early years of
the powder metallurgy industry when the production of larger and more
complex components was attempted. It was then appreciated that such
products would most effectively be compacted under a pure isostatic pressure
but it was not until 1930 that F.Skaupy devised a method of hydrostatic
pressing which was simple, cheap and industrially acceptable.
Figure 1.14 illustrates a typical arrangement which was used for pressing
relatively thin-walled tungsten carbide tubes, an operation which would not
have been feasible in a conventional steel die. The carbide powders were
contained in a sealed rubber bag, which was then subjected to hydrostatic
pressure from a fluid pumped into the pressure vessel shown. The shape and
configuration of the powder compact was ensured by the use of polished steel
tubes which were so arranged that they allowed the working fluid to act on the
compact from all directions. Compacts with high pressed densities and low
internal stresses were thereby obtained. With arrangements of this type,
compaction pressures of the order of 9250 bar (60 tons per square inch) could be
safely utilized. Pressures in a steel die were usually limited to about 2300 bar (15

tons per square inch) to avoid the formation of internal cracks in the compact.
In the years immediately after the Second World War it was established by
organizations such as the Nobel Division of ICI that large metallic compacts
could be effectively compacted by surrounding them with a uniform coating of
high explosive which was then detonated within a large bath of water. This
technique was among those investigated for the consolidation of large ingots of
NON-FERROUS METALS
139
titanium from the sponge then being produced by the Metals Division of ICI.
The explosive approach, though versatile, is expensive and not well suited for
routine manufacture, although its value in the research and development area
has been amply demonstrated. Isostatic pressing is now widely employed for
the manufacture of smaller components: vast quantities of pressings such as the
alumina insulators of sparking plugs are economically produced in this way.
In 1958 the Battelle Memorial Institute of Columbus, Ohio, began to
modify and improve the isostatic pressing process so that it could be used to
consolidate metal powders at high temperatures. See Figure 1.15. Laboratory
gas pressing equipment was developed which allowed metals, alloys and
ceramics to be consolidated in metal capsules at temperatures up to 2225°C
and pressures up to 3900 bar (25 tons per square inch). The commercial units
Figure 1.14: The first satisfactory method of consolidating metal powders under a
hydrostatic pressure devised by F.Skaupy in 1930.
PART ONE: MATERIALS
140
available in 1964 operated at similar temperatures, although routine operating
pressures were limited to about 1100 bar (7 tons per square inch). Pressing was
accomplished in a cold wall autoclave, with the furnace used for heating the
compact enclosed within this chamber and thermally insulated from it so that
the chamber walls remained cold. The unit was pressurized with an inert gas
such as argon. HIP (hot isostatic pressing) is now extensively used to compact

either one very large component or a multiplicity of relatively small specimens.
Units with an ability to consolidate components up to 22 inches in diameter
and 108 inches long (56cm×274cm) were being used in the USA in 1964 and
Figure 1.15: The technique of hot isostatic pressing, introduced in 1960 by the
Battelle Memorial Institute is shown in the second diagram. In this approach the
metal or alloy powders are poured into shaped sheet metal cans which are
subsequently evacuated and sealed by welding. The can and its contents are then
subjected to the combined effects of heat and an external pressure of an inert gas
in a cold-walled autoclave. The hot isostatic pressing ‘HIP’ process is now
widely employed, particularly for consolidation of complex shapes from
superalloy powders.
NON-FERROUS METALS
141
much larger facilities are now providing routine service throughout the world.
‘Hipping’, as a metallurgical technique, is no longer confined to powder
metallurgy. Many critical and expensive castings such as those used for gas
turbine blading are routinely processed in this way to seal up and eliminate any
blowholes or other microscopic defects which might possibly be present.
Higher operating pressures are now available; the economics of the process are
largely dictated by the fatigue life of the pressure vessel.
TITANIUM AND THE NEWER METALS
Titanium was first produced in metallic form in 1896 by Henri Moissan, who
obtained it by electric furnace melting. His product, being heavily contaminated
by oxygen, nitrogen and carbon, was very brittle. In the early years of the
twentieth century ferro-titanium alloys were widely used for deoxidizing and
scavenging steel. Metal containing less than 1 per cent of impurities had been
made, however, and such material had a specific gravity of only 4.8g/cm
3
. The
metal had a silvery-white colour and was hard and brittle when cold, although

some material was forgeable at red heat. Widely varying melting points were
initially reported, although work by C.W.Waidner and G.K.Burgess at the US
Bureau of Standards indicated that the true melting point was probably between
1795° and 1810°C. The extraordinarily high affinity of titanium for nitrogen was
noted in 1908 and this led to a number of proposals that the metal could be used
for the atmospheric fixation of nitrogen. It was claimed that the titanium nitrides
reacted with steam or acids to produce ammonia.
Pure ductile titanium was first produced at the Schenectady Laboratories of
the General Electric Company in 1910 by M.A.Hunter, who reduced titanium
tetrachloride with sodium in an evacuated steel bomb. This approach was used
for a considerable time for producing small quantities of relatively pure
titanium for experimental purposes.
Titanium of very high purity which was completely ductile even at room
temperature was first made in 1925 by Van Arkell and De Boer who prepared
pure titanium iodide from sodium reduced titanium and subsequently
decomposed this volatile halide on a heated tungsten filament. It had a density
of 4.5g/cm
3
, and an elastic modulus which was comparable to that of steel. For
the first time titanium was seen as a promising new airframe material which
was free from many of the disadvantages inherent to beryllium.
W.J.Kroll began to work on titanium in his private research laboratory in
Luxembourg in the decade before the Second World War, using an approach
which was identical in principle with that employed by Wöhler in his work on
aluminium in 1827 (see p. 102). He found that the metal obtained by reducing
titanium tetrachloride with calcium was softer and more ductile than that
obtained from a sodium reduction, and applied for a German patent for this