Introduction

From a metallurgy perspective If anyone had asked “How can I make a large 24ct gold jug” the reply would be “don’t try it will not hold its shape”, manufacture in a metal as soft as pure gold was technically impractical. However this is exactly what Pugh together with myself and other specialist crafts people did achieved in 2008. Pugh is a master craftsman who calls upon a network of crafts people for expertise beyond the scope of his workshop. Being part of Pugh’s network for over 15 you know requests for help are always challenging, my contribution to the creation of the Jug was the critical water tight welding of several invisible seams by laser. The craftsmanship of the Jug was documented in the 2010 Santa Fe Symposium paper, which captured how Pugh took the product of the metallurgy lab (the gold titanium AuTi alloy 1%Ti) and transformed it into his product (the Jug).

At the time of making, craftsmanship and metallurgy were treated as independent factors. The Jug started life as a live commission, in the domain of a craft workshop, and became an unintended piece of craft research in action. Yet it is difficult to understand how craftsmanship alone can fully explain how this Jug was possible. This exposition draws together several strands of craft activity and discussions and sets out to understand how the material properties of an unusual gold titanium alloy in collaboration with the skills of a master craftsman worked together to realise the Jug.

The Jug presents an elegant and irrefutable piece of evidence to disrupt established thinking, it is an enigma of metallurgical behaviour. The Jug merged old and new techniques, demanded that workshop practices be reassessed, while defying metallurgical thinking that near pure gold is too soft for load bearing applications.

Remarkably AuTi alloy was expected to be finished with a deliberate age-hardening treatment however that final stage was abandoned. John Wright (metallurgist, amateur craftsman and co-author of the conference paper this section of the exposition is based on) was aware of the jug and was very interested in knowing more about its manufacture. When he learned the Jug had not been age hardened it presented a metallurgical puzzle. Why had the final technical procedure not been adhered to? More profoundly how does the Jug even exist? Wright was convinced that there were crucial metallurgical considerations that needed to be addresses, prompting Pugh Wright and myself to investigate the interdependency of craft and alloy metallurgy to understand how this Jug was possible.

The clients request for pure gold initiated research into high carat alloys - scientific papers, discussions with bullion dealers, and global correspondence with those few individuals with practical experience.

This table is taken from Pugh’s research folder, detailing ‘Improved Strength 24ct Gold’s’ from Chris Corti’s paper ‘Strong 24 carat golds: The metallurgy of micro-alloying’

What is interesting is, Pugh’s annotations to the table his own key for hardness – silver annealed 75 Vickers Hardness (HV industrial scale for hardness), work hardened 115 HV. Pugh utilised silver as his experiential reference point to understand metallurgical terminology and relate it to workshop practice. 

Pugh prepared a design and quotation, based on his experience, the metallurgical and practical information gleaned from his literature research so far. Certain manufacturing techniques were discussed in the literature, which covered metal casting and lath turning. But there were knowledge gaps, no information on spinning, a crucial technique in the production of this jug, work shop trials were scheduled to gain first-hand experience and material knowledge, to bridge knowledge gaps and gain a confidence in working the alloy.

The design of the jug presented visually and technically critical points of construction e.g. the spinning of the body in two parts or three? the seam across the central body of the jug required an invisible water tight fit how would that be achieved when the alloy was a rich golden colour but no solders matched the colour, finish with such an unusually high content of gold how would the surface react?

Pugh and myself are crafts people, and although the behaviour of the alloy was definitely unusual and noted, the impact and interaction between the craftsmanship and the metallurgy of the alloy, was not fully recognised at the time of the Jugs making.

Observations of alloy behaviour in response to workshop practice

Spinning, demanded a processing environment of heightened sensitivity to predict the change in material properties, although there was a longer working window of material deformation allowing the metal to create the body shape there was a sudden work hardening of the alloy at the end. This work hardening was so abrupt that Coe the spinner actually cut his finger.

Joining the two part spinnings with an invisible join, soldering- even in 18ct would leave a different coloured horizontal seam across the widest point of the jug, Tungsten inert gas (tig) welding - required a large arc and the control of heat at great speed, Laser Spot Welding provided a steady, controllable process, utilising the alloy as the weld material maintaining the colour and a less risky approach. Initial laser welds exhibited a blue oxide, probably due to the Ti – but not an invisible seam, however this was easily removed, and subsequently LSW was shrouded in argon gas to prevent oxide generation.

Observations, the alloy was ‘sticky and sluggish’, more difficult to push the metal around with the laser beam, the metal occasionally ‘popped’, the natural ripple of the laser welded was more pronounced.

Solder Trials revealed its lighter colour due to the greater reflectivity and less abrasive resistance, possibly because it was less ‘tough’ and ‘sticky’ than the surrounding gold alloy.

Finishing Trials, The ‘sticky’ nature of the alloy caused the lemel to clog files and mark the surface (similar to platinum), thus the use of emery paper with lubricants was necessary. Polishing the alloy with a mop produced a mottled surface with alternating areas of dull and high polish. The Ti appeared to create hard areas, which were accentuated when mop polishing removed the softer surrounding gold. Interestingly at the point of contact between mop and alloy surface the metal did not flow.

The initial comparison with normal high carat golds used by craft-people highlighted the preconceived view that 1% of additional ‘exotic’ material wouldn’t make that much difference!  However the ‘not much difference’ notion was quickly dismissed once they got their hands on the alloy.

We decided to examine a cross section of the alloy an off-cut left over from the Jug manufacture. I refer to Wrights comments on the micrographs:

‘The tiny white dots grew in the various process stages, from titanium particles to form Au4Ti - titanium rich intermetallic. Why titanium rich? Each particle of titanium is surrounded by gold and a reaction starts at the interface to form Au4Ti because this gives out heat, the temperature rises, accelerates the reaction and more titanium diffuses from the centre to the spherical interface which grows outward.

They all show a typical dendritic (tree like) cast structure whereby the intermetallic particles have formed early in the cooling process and they float in a still liquid sea of gold.’

When the metallurgy findings are mapped alongside the observations of crafts people a number of connections emerge. The intermetallic compounds (the white dots) lower the energy of the system and reduce the thermal conductivity of the alloy. Laser spot welding utilised a focused laser, combined with the low thermal diffusivity (reduced the spread of heat) of the alloy those factors created a weld that did not impact the work hardening of the material around it. Furthermore the intermetallic compounds interrupt the flow of the alloy with hard lumps, this can be evidenced in the sudden work hardening during spinning, and the issues of orange peel surfaces created during polishing where the softer surrounding pure gold is easily polishing out leaving the compounds behind. The observations of crafts people supports the theory and the greater impact the intermetallic compound had than was originally thought.

Much more research is needed to fully understand all the alloys. characteristic and the role of the intermetallic. What is clear now but unknown at the time of making was that the material behaviour of the alloy were intuitively optimised in the realisation of the Jug. The success of the Jug was a true collaboration integrating design, materials together with  highly skilled and adaptable craft people.

Metallurgy Interpretation

Wright (metallurgist) needed to understand Pugh’s decision to not age harden the Jug. In his Santa Fe paper ‘Buy By Weight: Think Volume’, Wright notes it is standard practice to use percentage weight to describe alloy composition e.g.18ct is 75% gold 25% alloying metals such as copper, zinc or palladium. Bullion determines the price on the world commodity markets therefore weight is commonly used. The AuTi alloy is 99% Au 1% Ti, 100 grams of alloy is of 99 grams of pure gold and 1 gram of titanium.

The AuTi alloy is rare and not widely used (it was specifically alloyed for this Jug commission), therefore collating first-hand experience from crafts people was valuable to gain their insights on the alloys response to craft processes. Without exception everyone experienced behaviour that was unexpected (very quick work hardening, low thermal flow through the metal, orange-peel surface texture). Can 1% (weight) of Ti explain this alloy behaviour?

To the craftsperson the internal structure of any alloy is a three dimensional composition of metal atoms that adapt their structures in response to the application craft processes. For example a sheet of alloy can be work hardened through hammering, making it difficult to shape, the application of heat anneals the sheet which changing the structure and it becomes malleable and ready for reshaping. For a crafts person process and material behaviour work together. To investigate the metallurgy of this alloy a step change in thinking was required to present a new perspective to reframe the alloy.

Thinking three dimensionally is important to crafts people and to explain this alloy more fully. Volume rather than weight illustrated the differences of Au and Ti and posits a theory for the mechanism of interaction between Au and Ti within the alloy.

Firstly volume, titanium is much less dense (atomic weight 47.867) than gold (atomic weight 196.966) therefor a lot more titanium atoms (volume) are required to make up 1 gram of titanium, than needed for 1 gram of pure gold. On a volume basis, in 100 grams of AuTi alloy you have 4.1% of titanium atoms which volumetrically represents over 4 times more impact on the behaviour of the alloy than when considered 1% in weight terms.

Secondly structures, the metallurgical phase diagram for Au and Ti illustrates the interaction of Au and Ti. Titanium can form up to four different intermetallic compounds, the properties of an intermetallic can differ considerably from their individual constituent and often do not inherit their parent’s properties. Theoretically each of the titanium atoms can combine with four gold atoms forming the intermetallic compound Au4Ti. This compound is capable of reacting with over 20% of the gold and in doing so locks up an unexpectedly high number of gold atoms. From a behaviour perspective the compounds do not disappear into the pure gold matrix; they form sizeable hard clumps with a different crystal structure from the highly ductile surrounding gold which is locally purer than before, because most of the titanium is now concentrated in the intermetallic compounds. From a crafts perspective the alloy is not extensively malleable but ‘interrupted’ by the intermetallic compounds resulting in unusual material behaviour.