AN INTRODUCTION TO DATA ON Introduction A large variety of carat gold alloys are available to the jewellery manu- facturer. Many of these alloys are based on traditional compositions and are still being used extensively be- cause their behaviour throughout the entire manufacturing cycle is well understood. Others are more recent formulations having improved prop- erties and behaviour in specific man- tufacturing processes. Developed and sold by suppliers of gold alloys and semi-finished products, the composi- tion_of these alloys is usually kept confidential Some alloys are clearly better suit- ed to particular manufacturing opera- tions than others, and significant sav- ings in production time and money can often be made by choosing the right alloy from the outset. Unfortu- nately, there does not exist an ade- quately standardized basis for com- paring the different carat gold alloys short of carrying out full production trials. The gold jewellery industry has therefore a very real need for more comprehensive basic data on the al- loys available GOLD ALLOYS BY PARN TAIMSALU ‘The manufacturing jeweller as well as the individual craftsman is con- cerned basically with the following characteristics of carat gold alloys: Physical and mechanical proper- ties © Appearance Resistance to tarnishing and cor- rosion © Durability and resistance to scratch- ing. (Of particular interest are the physi- cal and mechanical properties, since they help the manufacturer to decide whether a given alloy lends itself easily or only with difficulty to the various production operations. The first series of data sheets will therefore mainly concentrate on these proper- ties, and on the way in which they are influenced by cold working and heat treatment. (Gu 750-Cu 160-49 9 BN (Au 750-Cu 205-Ag 45)General description Under this headinga brief description isgiven of the type of alloy covered by the data sheet. Composition ‘The composition of the alloy is given in parts per thousand (‘o). The exact gold content used in practice often depends on the hallmarking regula- tions in force in each country. In many countries, for example, it is usual to make up alloys with 1-3%o more gold than required by law, to ensure that any possible segregation or processing losses do not lead to assay shortfalls. Physical properties Colour, density and melting range are purely physical properties of an alloy and generally unaffected by working or thermal treatment. Density, how- ever, may vary slightly depending upon the state of the alloy. Poorly cast items with extensive microporosity will have significantly lower densities. The colour of a gold alloy is very difficult to describe and descriptions of colours are helpful only if they refer to a set of samples made up to serve as standards for comparison. In response to this need for standards, the Swiss organization “Normes Indus- trielles de Horlogerie Suisse” estab- lished, in 1966, a series of standard 18 carat coloured gold alloy composi- tions which were designated by the code numbers 2N, 3N, 4N, and SN. These standards are now also accept- ed by the French industry, and, with certain modifications by the German industry (see AURUM, Pilot Issue, page 7). The first set of data sheets deals with these specific alloys, which are illustrated on the opposite page. The melting range of an alloy is simply the interval between solidus and liquidus temperatures. At tem- peratures within this range, the metal is neither completely solid nor com- pletely liquid. Mechanical properties Among the most important mechani- cal properties of interest to jewellery manufacturers are hardness, tensile strength, vield strength and ductility. In cantrast to most physical proper ties, the mechanical properties of al- loys are extremely sensitive to the state of the metal and are conse- quently summarized in a table for four conditions:cold worked to 20 per cent cold worked to 75 per cent annealed hardened by heat treatment. It is important to note, however, that small variations both in the way the alloys are prepared, and in the testing methods themselves, may lead nificant differences between the measured values. These differences in general could amount to as much as 10 per cent, or even more in the case of alloys in the age hardened state. Hardness ‘The hardness of an alloy can often provide an approximate indication of its working properties, and because it can be measured relatively easily, itis frequently quoted in the documenta- tion published by alloy manufactur- ers. However, several different meth- ods are used to measure hardness and itis not always easy to compare values obtained by different methods. Com- parison tables are only reliable if es- tablished for alloys within the same group. Inthe past, the Brinell hardiness test- ing method was used extensively, and older books and publications usually quote hardness values obtained using this method. In the U.S.A. the test ‘most frequently used and quoted is the Rockwell one. In Europe how- ever, the Vickers hardness testis the ‘most popular one for gold alloys and consequently the IGC Data Sheets will quote Vickers hardness values. Yield strength and tensile strength Yield strength is the force per square millimetre of cross-sectional area at 4 which the material exhibits a speci- fied amount of permanent deforma- tion. It provides an indication of the forces required to deform or work the alloy mechanically. Tensile strength, on the other hand, indicates the forces that cause the’ metal to break or crack. The yield strength, elastic limit, proportional limit and proof stres are all virtually identical mechanical properties of an alloy, and the term yield strength is used in the data sheets for simplicity. The units in which yield strength and tensile strength are quoted are deca-newtons per square millimetre (daN/mme), These units are used in preference to N/mm? because they give values almost identical to the more familiar older units, kg/mm: Elongation and ductility Yield strength, tensile strength and per cent elongation can all be mea sured on the same test specimen us- ing a tensile testing machine (see illustration), For this reason, the latter parameter has come to be used as a0 Wine2 fo uve forthe 18 carat 3N ihe tensie test measure of the ductility of the alloy. When an alloy is destined for deep drawing, the Erichsen cupping test is often used to provide additional infor- mation about ductility (see illustra- tion). Processing guidelines Under this heading, brief comments and recommendations are given re- garding the handling and behaviour of the alloy in production. Cold working Cold working has adramatic effect on the mechanical properties of most gold alloys. Graphs are therefore shown in the data sheets to illustrate the variation with degree of cold working of hardness, yield strength, tensile strength and elongation. Annealing Graphs are also used to illustrate how the key mechanical properties of a cold worked sample vary with anneal- ing temperatures. They are particu- larly helpful in establishing an opt mum annealing temperature. It will be noticed that a relatively low an- nealing temperature is recommended for the 18 carat coloured gold alloys. This is because annealing at higher temperatures is more difficult to con trol and can cause excessive grain growth, especially ifthe metal has not been sufficiently cold worked pre- viously. Microstructure A series of five microphotographs are included in the data sheets to illus- trate the structure and grain size of the alloys. The microphotograph of the “as cast” sample shows the structure of an investment casting, and the others indicate the effects of both under annealing and over annealing on grain size Hardening Many gold alloys can be hardened by a simple heat treatment at a relative- ly low temperature, The hardening treatment can be carried out on fin- ished items of jewellery and is par- ticularly effective ifthe alloy contains a significant amount of copper. Yield strength and tensile strength are also increased, and the table of mechani- cal properties given in the data sheet includesacolumn showing the values, which can be obtained by hardening. 0.5 mm shiek gold allay unt cracks began to ‘2opeat. The depth ofthe resuling “cup” men is quoted in the data sheet. A graph of hardness against tempera- ture of heat treatment is included to show that even very low temperatures an cause hardening to take place. This is why it is necessary to quench most coloured gold alloys. . The author would like to thank Métaux Précicux SA for kindly supplying the illustrations and for the Joan of colour samples of the four 18 carat yellow gold alloys.General description Composition Physical properties Mechanical properties Processing guidelines GOLD ALLOY DATA Au 917 — Ag 55 — Cu 28 A standard 22 carat yellow alloy. Used extensively for jewellery fabrication. Component Au a % 917 # 28 Colour Density Melting range Yellow 2c *a) 17.9%@) 995-1020 ‘This alloy lends itself well to all the techniques of cold working (bending, stamping, cutting, ete.). Best results are obtained by using fine-grained material and by re- specting the cold working and annealing recommendations given in this data sheet.Cold working Effect of cold working ‘on mechanical properties Before annealing, the alloy should be cold worked to at least 50 per cent and preferably to 75 per cent. When wire drawing, even higher cold work- ing rates are possible. Low cold working rates necessitate relatively high annealing temperatures and excessive grain growth may then take place. ‘The graphs illustrate how hardness, tensile strength, yield strength and elongation vary with the degree of cold working of the alloy. la & zg 22 8 ‘VICKERS HARONESS Sails lee COLO WORK % aw @ TENSILE STRENGTH, YIELOSTRENGTH | Nimm? aa en 5 Ef lo IR = Is Is a lo Iz ie 3 Is Bie ovo works ‘The increase in hardness with increasing The increase in tensile strength (A) and The decrease in elongation (ductility) amount of cold working. Microstructure ‘The “as cast” structure of the alloy. Magnification: 200x. yield strength (B) with increasing amount with increasing amount of cold working of cold working. ‘The structure after cold working to 75 pper cent reduction. Magnification: 200x.Annealing The recommended annealing treatment for material previously cold worked to 75 per cent reduction is 30 minutes at 600°C. Goldsmiths working with similar alloys may torch anneal small arti- cles at approximately 650°C (medium red in colour when observed away from direct sunlight or artificial lighting) for about 60 seconds. Effect of annealing temperature The graphs illustrate how hardness, tensile strength, yield strength and ‘on mechanical properties elongation of a 75 per cent cold worked sample vary with the tempera- ture of heat treatment (30 minutes at each temperature). 200. "= 600 0 5 | ee oy 160] E 2 a i 0 10 | S 30 = j = 5 0 4 = 20 4 5 ae 2 29 f £ 3 Eo Zo Bo Se a EN aE EEE ANNEALING TEMPERATUREC ANNEALING TEUPERATURE ANNEALING TENPERATURESC ‘The change in hardness with temperature The change in tensile strength (A) and The change in elongation (ductility) of heat treatment, yield strength (B} with temperature of with temperature of heat treatment. heat treaiment. Effect of annealing temperature on microstructure Ghee i . Structure of 75 per cent cold worked Structure of 75 per cent cold worked Structure of 75 per cent cold worked sample after ennealing at 300°C for 30 sample after annealing at 600°C for 30 sample after over-annealing at 700°C minutes. minutes. for 30 minutes. Note that excessive grain Magnification: 200%. Magnification : 200%, grovith has taken place. Magnification: 200%.Hardening Soldering Pickling Notes Source of data 22 carat gold alloys cannot be hardened by heat treatment to any significant degree, This alloy can be soldered easily using solder alloys of the same carat and colour and either a torch or a furnace. Hot 10 per cent sulphuric acid. 512), Standard specifications: NILNS-03-50 (July 1961), Switzerland ap ets CETEHOR-U7-70 (Mordh 1966). France Din 8338 1966) Germany “1 Reni wrest dpentng othe Yate of telly, ke cold worked, mead o iy nmetd fer 20 mints 600°C ‘After cold working to 75 percent rection Vickers hardnes tester food The Wire samples 100 my fone, nomineliy2 mm diameter Test according fo DIN S0102 specifeaion, 6 et according 10 raion, Germany Diameter of ball: 15 mm Lise Sample: annealed chet 0.5mm thick The information in this data sheet was generously provided by Johnson Matthey Metals Limited, 43, Hatton Garden, London ECIN SEE, United Kingdom.BASIC METALLURGY FOR Nucleation and dentritie growth In the previous article, we saw that pure metals solidity at a characteristic fixed temperature whereas alloys solidify over a range of temperatures. There are only two exceptions to this rule : Alloys having a eutectic composition and some intermediate phases of fixed composition usually referred to as intermetallic compounds, both which also solidify at a fixed tem- perature. We also saw that even in the case of pure metals, solidifi- cation is not instantaneous. but occurs over a period of time. This is because when solidification starts, it does so from a isolated points in the melt usually from impurity particles floating in the liquid metal or from the mould or crucible wall — in much the sime way as ice begins to form on a pond. This process is, known as nucleation. At the temperature of the melting point, a few atoms come together to ‘form a tiny seed erystal around an impurity parti cle or at the mould wall. A seed crystal may also be formed occasionally due to thermal fluc- tuations in the melt, As more and few more metal atoms are deposited Figure 1- Sketch of a dendrite araw ing in three preferred directions fram original se al f in the (after I Higgins, “En yy" The. English 961) ering Metal GOLDSMITHS SOLIDIFICATION OF METALS. BY MARK F, GRIMWADE on it, this seed or nucleus begins to grow as a crystal having its own geometrical lattice arrangement. You can imagine this nucleus as being a tiny cube floating in the melt and growing preferentially in the directions of the cube edges by the formation of so-called pri- mary spikes extending from all six faces of the cube into the liquid metal. Similarly, secondary and tertiary spikes branch out sideways from the primary arms and even- tually a treedike metal crystal 1 called a dendrite is produced (Figure 1). This is the most com- mon process by which metals solidify. Grain structure In practice, a large number of nuclei are formed when liquid metal cools to its melting point All of these nuclei are capable of growth, but the direction in which dendrites will grow from them depends on the initial orientation of each nucleus. Consequently. "when dendrites growing outwards from different nuclei come into contact with each other, a bound- ary is formed at the contact interface. Metallurgists call this a grain boundary. At this stage, some liquid metal will still be trapped between the dendrite arms, but as, solidification proceeds, the arms zrow thicker by metal deposition until all the space is filled with solid. This sequence of nucleation, dendritic growth and solidification to form a polyerystalline structure is shown in Figure 2. A good representation of a polycrystalline structure is obtained when you shake a soap solution to produce a three-dimensional network of soap film boundaries analogous to the grain boundaries in a metal. Metallurgists examine the grain structure of metals by taking a cross-section of the cast metal, grinding and polishing the surface produced to remove the worked layer left by the cutting operation, and then etching this surfuce in a suitable chemical reagent which 2 12 preferentially attacks the metal at the grain boundaries. Examination of the etched surface under a reflecting light microscope will reveal “the grain structure (Figure 3). In pure metals, there is, normally no evidence of dendritic growth and only a mass of crystals, or grains, with boundaries between them is seen in the microscope. However, if the residual liquid metal is decanted during the so- lidification process or if a void remains in the centre of the casting, then the dendrite arms may become visible (Figure 4). Grain size In general, the smaller the grain size of the metal, the better are its mechanical properties such as strength, ductility and toughness. ‘One reason for thisis that insoluble impurities such as oxides present in the melt tend to be trapped in the last liquid to solidify and often form a continuous film of brittle material on the dendrite arms and at the grain boundaries. If present in sufficient amounts, such films usually weaken the casting and render it brittle and unwork- able. Consequently, it is desirable to cast in such a way that small grains are obtained. Because each nucleus develops into a grain, the final grain size is controlled by the number of nuclei. With a slow cooling rate from the molten state only a few nuclei are produced and consequently these can grow unhindered to give a relatively large grain size. With a fast cooling rate, however, many more nuclei are produced. This is because not all nuclei form at the same time. If the temperature drops rapidly (even to below the usual_ melting point) additional nuclei form before the original ones have had any real chance to grow. As a result, the growing nuclei soon bump into their neigh- bours and therefore the resulting grain size is relatively small, a Solidification of alloys In contrast to pure metals, alloys usually solidify over a range of temperatures. The dendrites in cast Figure 2— Stages in? the solicification of a pure metal ‘A: Nucleation in the melt B : Dendritic growth € : Formation of grain boundaries 1D : Grain structure of solidified metal (After Rollason, “Metallurgy for Engi- neers”, Amold & Co., London).alloys, however, may be observed fairly easily under a microscope because either 1)A difference in alloy composi- tion exists between the centre and extremities of the dendrite arms, ive. there is a gradual change in composition from the central nucleus to the last part of the dendrite to solidify; this is known as coring, oF 2)Two or more different crystal phases are present, and in that case one phase has often grown as den- ites which are then more easily tinguishable from the other phase An example of a typical cored structure is shown in Figure 5. Cast alloys often have a cored structure. The reason why coring occurs may be seen by considering, for example, the solidification sequence of the gold-50 per cent silver alloy using the gold-silver phase diagram shown in Figure 6. ‘As the temperature of the melt falls to the liquidus line at point B (approximately 101°C) soliditi- cation begins. The first solid to form has the composition shown by point X situated on the solidus line, which is gold-36 per cent silver. Therefore, to maintain the overall composition of 50 per cent silver in the system, the remaining liquid is gradually getting richer in silver. When the temperature drops further, to 1010°C, for example. the solid now being deposited on the growing dendrites has the composition shown by point Y and the liquid from which it is being formed has the composition shown by point Z. Finally, when the temperature has dropped to ap- proximately 1000°C at point D, the last liquid, rich in silver, solidi- fies between the dendrites. A cored structure is therefore obtained in which there is a composition gradient across the dendrite arms They are richer in gold at the centres and richer in silver at the edges. This simplified explanation of coring can also be applied to other alloys by using the relevant phase sram and vertical composition line If the cored structure is undesir- able, it may be eliminated by a long heat treatment at a relatively high temperature. This permits the metal_atoms to change places by diffusion until the composition of the alloy is uniform throughout the casting. Such a heat treatment is known as homogenization Solidification in a mould When molten metal is poured into a mould it begins to solidify in- wards from the mould walls (Figure 7). If a cold metal mould is used, as is the case in ingot sting, the rate of heat removal is rapid and initially a layer of tiny Figure 3— Photomicrograph of the grain structure of pure gold in the as cast condition. (100) (Courtesy Dr. Ch, Raub) Figure 4 — Photomicrograph showing dendrites in a shrinkage cavity in a cast ‘copper billet (Courtesy Mark F. Grimwade). Figure 5 — Photomicrograph showing the cored structure of an 18 carat ye low gold alloy in the as cast condition (100) (Courtesy Dr. Ch, Raub.9. SHAWN PER CENT SILVER PER CENT GOLD 14chill crystals is formed on the mould wall. Long fingerlike grains (columnar grains) then begin to grow from the walls towards the centre of the ingot. This is shown in Figure 8. ‘The extent of this columnar growth is dependent on the temperature of the molten metal as it is poured into the mould. A low pouring temperature will allow nucleation to take place in the centre before the columnar grains have had time to grow into this region (Figure 8). When a plaster mould is used, as in lost-wax investment casting, the cooling rate is slower and equiaxed grains are obtained throughout the casting (Figure 9). ‘The grain structure, grain size and ultimately the success of the casting depends very much on pouring temperature and mould temperature. If the pouring tem- perature is too low the melt begins to solidify before the mould has been properly filled, resulting in imperfect castings with a poor surface. On the other hand, if the pouring temperature is too high, excessive oxidation and absorption of gases into the melt can occur, and coarse grains are obtained These effects may give trouble in any subsequent working operations or give brittle castings. For carat gold alloys, recommended pouring the temper- Figure 6 — Phase diagram of the gold: silver system showing the solidification sequence of the alloy Au-50% Ag, Under practical cooling conditions, the last liquid to solidify between the dendrites is rich in silver, and coring occurs. atures in ingot casting are usually of the order of 75 to 100°C above the liquidus temperature of the alloy. Shrinkage Metals and alloys undergo consid- erable shrinkage when they solidify and this is shown in Figure 10. Gold, for example, contracts 5.1 per cent by volume on solidi cation . The practical consequence of this may be seen in Figure 8. As the melt solidifies from the mould wall towards the centre, a funnel- Figure 7 — Solidification proceeding in wards from the mould wall Figure 8 — Tho grain structure of ingots cast into metal moulds at two different Pouring temperatures, A: Relatively high pouring temperature B : Lower pouring temperature, like depression known asa primary pipe is produced at the top of the ingot. When casting, therefore, sufficient provision must be made for a reservoir of molten metal above the casting so that shrinkage voids are confined to this region. Sometimes the top of the casting solidifies leaving_molten metal trapped inside. This will also shrink while freezing and a cavity known as a secondary pipe is left inside the casting. In a similar way, molten metal CHILL CAYSTALS| 7 a ig 5 3 - TE i Fs avin ie i MOULD WALL 1516 TEMPERATURE is trapped between dendrite arms as they grow during solidification and it is important to be able to feed molten metal back into this, region to prevent formation of any shrinkage voids. This becomes increasingly difficult with alloys that solidify over a wide temp ature range and also with certain shapes of castings. Failure to provide adequate feeding of molten metal during solidification leads to a defect known as interdendritic porosity or shrinkage porosity. This reduces considerably the strength of the casting (Figure 11). Figure 9 — Grain structure of an ingot cast in a plaster or investment mould, Figure 10 — Curves showing the de- crease in volume of a given weight of ‘metal as it solidifies A: Contraction of molten metal B : Contraction of pure metal on soli INCREASING COLD WORK, €.G. REDUCTION IN THICKNESS BY ROLLING SUPDIRECTION @ @ @OO0COO8 ©8000 HOCH HHHOOHOOD @OO8HHOHHHHOHOHOOO ©8008 0O8OOOOHO8OO @@0e20200000 @@ OQ snes. 18 Figure 2 — The effect of cold working ‘on the microstructure of a single phase metal Figure 3 — Simplified sketch of slip in a crystal lattic,DISLOCATION a350nge : ss ils 3 STRESS SERRE T- DISLOCATION TRAVELLING IN THIS DIRECTION DISLOCATION SLIP PLANE DISLOCATION aeeeie ae HEED ES to be increased, ie., the metal gets stronger, and hence the term work hardening. Eventually the number of dislo- cations (the dislocation density), increases to about 10!/em?. At this stage a vast “traffic jam” of dislocations occurs and movement ceases. The metal now has zero ductility and any attempt to continue deformation by _ in- creasing the stress level causes it to fracture, i.e., a crack appears. ‘A major cause of fracture in metalsis overworking, Other causes are “hot shortness” and “cold shortness” due to the formation of embrittling or low melting point phases. Obviously during the manufac- ture of wrought products, whether they are the _semi-fabricated products such as sheet, tube, wire, etc., or the more intricately shaped final products of the gold- smith, it is necessary to avoid overworking the metal and to restore its softness and ductility by annealing. The amount of cold working permissible before anneal- ing is necessary depends on the STRENGTH PER CENT COLO WORK (STRAIN) work hardening characteristics of the metal or alloy in question. Curve I in Figure 5 shows a metal with a relatively high rate of work hardening which requires more frequent annealing than a metal with a lower work hardening rate (Curye 1D. With ductile metals like the carat golds, silver and copper, it is good practice to anneal after about 70 per cent strain as measured by percentage reduction in thickness or area during work- ing. However, this advice can only Figure 4 — Movement of a dislocation through a crystal lattice Figure 5 — The rate of work hardening is indicated by the slope of the curve. Figure 6 — Effect af increasingannealing temperature on the properties of metals. be followed strictly in the produc- tion of semi-fabricated products. For when a goldsmith bends and shapes an article of jewellery he cannot really quantify the amount of strain introduced. Experience must then decide when annealing is necessary, but overworking at any stage could lead to fracture. Annealing The effect of increasing annealing temperature on the properties of a metal that has been cold worked is, summarised in Figure 6, and the accompanying changes in. micro- structure are illustrated below the graph. The graph may be divided into three mains areas ~ recovery, reerystallization and grain growth. ‘As the temperature is raised, at 19| i = UNLvuaaN3L OOH LV ALHaE0Hdfirst certain, recovery processes take place. The high strength and low ductility resulting from cold working are virtually unchanged and there is also little change in the microstructure (as observed in the optical microscope). Changes in the dislocation substructure can be observed by electron microsco- py but this need not concem us here. Although no annealing takes place, a reduction in the level of residual or internal stresses can be achieved by heat treating in this temperature range. The subject of residual stresses which may be introduced into a metal by non- uniform working or by non-uni- form heating and cooling will be discussed in a later article, The important temperature range for full annealing is the recrystallization range in Figure 6. It can be seen that during this process, strength decreases to a much lower value and that duc- tility increases to a high level. The cold-worked fibrous structure is replaced by a new strain-free equiaxed grain structure having a relatively small grain size. Electron microscopy shows that the dislo- cation density drops to about 10°- 10” fines/em ‘As the temperature is raised yet higher, the strength of the metal continues to decrease slightly and ductility increases to a peak value after which it may fall as the soli- dus temperature is approached. At this stage, the most noticeable change is the rapid increase in grain size. It is generally advisable to avoid annealing in the grain growth range because the presence of a large grains in the metal can lead to marked surface unevenness during subsequent working. This is the “orange peel” effect, so-called because of the similarity of the resulting metal surface to the sur- face of orange peel. It arises because individual grains deform by different amounts compared to their neighbours during cold work- ing, due to their differing orienta- tions with respect to the direction of applied stress. As a result, surfaces not in contact with the tool or die have a rumpled appear- ance (Figure 7). This happens irrespective of grain size, but the t is only noticeable by the ided eye when the grain size is large. Control of grain-size during cold working and annealing opera- tions is therefore important. The variables affecting recrys- tallization behaviour are as follows: a) The amount of cold work. A minimum amount of cold work (called the critical strain) is neces- sary before recrystallization will ‘occur. Recrystallization occurs by the nucleation and growth of new strain-free grains from isolated regions within the cold-worked structure. The critical strain, which is of the order of two per cent for a pure metal but increases with increasing complexity of alloying, is the minimum. strain level necessary to initiate nuclea- tion and growth of new grains. The greater the amount of cold work the greater is the number of nucleating sites and the smaller is the subsequent grain size (Figu- te 8). This situation is analogous to nucleation and grain-growth during solidification. b) Temperature and time. The annealing temperature at which full recrystallisation will occur is dependent on i) the amount of cold work, large amounts having the effect of reducing this temperature the composition of the metal oralloy, and iii) the time at the annealing tem- Figure 7 ~ The “Orange Peel” effect. a1perature, longer times requiring lower temperatures, but tem- perature changes have a far greater effect than changes in annealing time. In practice, if annealing is done in a furnace, then annealing times are typically of the order of 30 minu- tes. With torch annealing, where times are obviously much shorter, slightly “higher temperatures. are required. ©) Composition. The recrystallization temperature is strongly dependent on metal or alloy composition. In the case of Pure metals a good guide is to assume a temperature which is approximately one third of the melting temperature expressed in degrees absolute (Kelvin), ie., m.p. °C +273 = mp. °K. Pure ead (Pb) will recrystallize at room temperature and can only be cold worked at sub-zero temperatures, but pure gold recrystallizes at 200 °C. The presence of impurities can raise the recrystallization temperature by an appreciable amount and ‘it follows that alloying additions will have the same effect. Typical annealing temperatures for the carat gold alloys are in the range 500-700 °C depending on their exact compo- sition. Control of the grain size and mechanical properties of an alloy is best achieved by the use of fur- naces and accurate temperature measurement methods for anneal- ing. However, it is recognised that igure 8 ~ The offect of the amount of cold work on the recrystallized grain size, 22 for the goldsmith this is rarely convenient and that torch ane nealing is and will continue to be widely used. Judgement of correct annealing temperature is made by assessing colour changes in the heated metal by eye. It is impor- tant to do thisin a darkened comer of the workshop if consistency is to be attained, for one’s judge- ment of temperature by coulour is seriously impaired by reflections of light from windows, fluorescent lighting, ete. Finally, it is recommended that coloured ‘carat gold articles are quenched in water or a pickle after annealing, Nickel-containing white gold alloys should be air- cooled as quenching introduces high residual stress levels. The Purpose of quenching the coloured gold alloys is twofold. First, any oxide scale will be removed in the process. Incidentally, the for- mation of oxide scale is kept to a minimum by avoiding excessively high annealing temperatures and long times. Secondly, further changes in the microstructure can 8 CRITICAL PERCENT cOLDWoAK STRAIN” PRIORTO ANNEALING occur on slow cooling which will harden the alloy and this will be undesirable if further cold work- ing is to be done.GOLD ALLOY DATA General Description Hardenable high carat gold alloy with very good wear properties. Composition Component Au Ti %e 990 10 Physical Properties Colour Density "Melting Range Yellow 19 g/m’ 1090 10°C ‘Chemical Properties Stable, and remains untarnished in nor- mal environments, It should be noted, however, that titanium isa reactive metal which combines with oxygen and nitro- gen when exposed to these gases at elevated temperatures. ‘The alloy must therefore be protected from contact with air when it is in a hot condition. Methods for doing this are described below where appropriate. Production The alloy must be made in a vacuum induction furnace, evacuated, degassed, and back-filled with high purity argon to a pressure of I torr. The crucible may be of alumina, zirconia or pure graphite. Processing Properties Well suited forevery kind of cold working in the soft annealed condition. In the hardened state limited working is, still possible (c.g. for changing the sizes of rings). Operations involving material removal (urning, drilling, grinding, polishing) should be carried out only on the alloy in the hardened state. Suitable for investment casting of jewel- lery if melted and cast under high purity argon, However, experience of such ca- sting is still limited, but other casting techniques are being evaluated. Au 990 - Ti 10 (990 Gold”) Annealing Annealing at S00 °C for 1 hour yields a homogeneous structure. To avoid large trains, the annealing time may be redu- ed or. preferably, the alloy subjected to 50-70% deformation (work hardening) beforehand. Annealing in air or in a protective at- mosphere containing nitrogen results in the formation of a brown tarnish layer, which can be removed by treatment with potassium pyrosulphate or mechanically. ‘Although this does not lead to significant loss of titanium, annealing under these conditions is not recommended. Annealing can be carried out under a protective atmosphere of high purity argon at reduced pressure, Preferably, however, its carried out at 800 °C ina bath of molten B,O; or “Degussa 540° salt (the latter is more suitable). Quen- ching should be rapid to avoid exposure of the alloy to air while itis hot. Age-hardening of the Annealed Alloy ‘The changes that occur with time in the hardness of the alloy (annealed at 800 °C ‘and quenched) when it is held at tempe- atures of 400, 500 and 600 °C are illustrated in Figure 1. The recommen- ded procedure is to artificially age-har- den for | hour at 500 °C, which gives a hardness of 170 HV. ‘The age-hardening can be carried out either under a protective atmosphere of high purity argon at reduced pressure or more conveniently by immersion in a molten bath of Degussa Flux h, or of Degussa 430 salt Mechanism of Age-hardening “The phase diagram of the Au-Ti system (see LL, Murray, in ‘Phase Diagrams of Binary Gold Alloys’, Fd. by H. Okamoto & TB, Massalski, ASM International, Metals Park, Ohio, 1987, 306-11) indi- cates thatthe alloy isa solid solution of TiAus in Au below about 1070 °C. The solid solution of TiAus in Au decreases below this temperature and, as judged by the behaviour of the alloy, it would appear that saturation of the solid phase with TiAus in Aw is reached around 800 °C, and that precipitation of this compound as a hardening phase sets in between this temperature and about 400 °C, Work Hardening Curve S (Figure 2) indicates that the hardness of the annealed alloy increases with deformation from an initial value of 75 HV to 125 HV at 80% deformation. In practice a deformation of atleast 50% and preferably 75% should be aimed at. Curve H Figure 2) illustrates how a higher hardiness of 180 HV can be achie- ved by age-hardening the alloy to 125 HY before cold working ‘Thecurve, marked 18 ct, fora typical 18 carat 3N alloy (Au-Ag 125-Cu 125) is shown for comparision. This alloy work hardens rapidly to a higher hardness on cold working. Therefore it requires inter- mediate annealing steps in wrought je- wellery production, whereas 990 gokl- titanium doesnot require such annealinga Hardness (HV 5) a a Hardness (HV 0.5) 8 8 - Figure 1 Hardness of 990 Au-T) as a function of time at various temperatures The effect of cold-working on the hardness of $90 Au-T' in two starting states and on an 18 ct jewellery alloy Mechanical Properties By using different combinations of work hardening and age-hardening,cither soft | ‘or hard forms ofthe alloy can be produ- ced suitable for wrought and turned jewellery respectively Data on the dependence of tensile strength, yield strength, hardness. and clongation-to-fracture on the time of hardening at 500 °C of the alloy in three starting states are presented in Figures 3, 4and 5. Note the variation of 2-8% in the elongation-to-fracture values in Figure ‘5. The reasons for this are not understood The test specimens were all made from cast alloy which had been deformed by oo 23%, given a solution treatment at 800 °C for 1 hour and quenched. Mate- rial in the state $ (oft) was prepared by repetition of the initial treatment; material inthe state C (cold worked) was prepared Hardness (HY 0.5) by 23% cold working: and material in Minutes at soore the state H (the hardest) was prepared by age-hardening for 1 hour at 500 °C, followed by cold working by 23%. Figure 4 ‘The dependence of hardness of 990 Au-T\ in three starting states, on time at 800 °C ‘ oe s Figure 3 The dependence of tensile and yield strength of 990 Au-Ti in three starting states, on timo at 600 °C S = annealed, C = cold worked, H=hardened Elongation to fracture (9%) mains at sass Figure 5 The dependence of elongation-to fracture of 990 Au-T' in three starting states, on time at 500 °C S = annealed, C = cold worked, H= hardenedMelting & Casting Jewellery can be cast using the lost-wax process, although experience of casting of the alloy is limited. Both melting and casting operations must be conducted in a closed vessel filled with high purity argon at reduced pressure (150-250 torn), the pressure being increased immediately after casting. A phosphate-bound invest- ‘ment medium should be used and a casting temperature of 1250-1300 °C Pickling Pickling with the usual acids (including hydrofluoric acid) is not possible, Appli- cation of a coating of potassium pyro- sulphate by immersion in a saturated solution of this salt, followed by heating to 500 °C, removes annealing tarnish, (ifthe material isheld at the temperature of 500 °C for 10-15 minutes, it will be hardened in the same operation), The texture of as-cast 990 Au-Ti 130) Grain Large changes in the texture ofthe alloy occur when its deformed and hardened. This is illustrated in Figures 6, 7, 8 and 9. The dramatic grain refining effect of age-hardening 70% deformed alloy for T hour at 500 °C (Figure 9) is to be noted. A striking improvement in the grain size of the as-cast alloy is also observed if small amounts of ruthenium and boron are incorporated in the alloy. Soldering & Welding Surfaces to be soldered must be cleaned and the whole surface to be heated must be covered with a flux to prevent tarni- shing. Suitable fluxes are Degussa Flux- or -h and Canning but others may be applicable, Standard 22 carat brazes can be matched to the alloy Soldering is preferable to welding since some titanium may be lost from molten areas when welding. The amount of solder used should be kept toa minimum. Texture of 990 Au-T Coin & Medallion Production ‘The solutionized (annealed) alloy (70 HY), after cold working to 120 HY, is very suitable for production of coins and medallions, since tis can be done without any annealing step between strip rolling and the blanking and striking operations The coins or medallions can be age- hardened by heating to 500°C for 1 hour. ‘Wear Resistance Abrasive wear tests on discs, wet tum- bling tests on coins and wear tests on rings indicate that the wear properties are good compared with these of similar articles made of other standard gold alloys. Refining of Serap ‘The titanium is simply removed from the alloy by melting under Degussa precious metal salt 640. Gold of 999.8 fineness is produced after 20 minutes at 1100 °C, land is suitable for production of new alloy. The titanium in the alloy does not ‘complicate refining by the Miller proces Fs ‘after 70% cold-worked followed by solutionvaing at 800 °C for 1 hour (190) Figure 8 Texture of solutionized material age-hardened at 600 °C for 1 hour (x10) Figure 9 Grain refinement following age-hardening af 70% deformed material for 1 hour at 800 °C (x1 30)ENVIRONMENTAL EFFECTS ON GOLD ALLOYS AND THEIR RESISTANCE TO TARNISHING AND CORROSION All metals are affected to a greater or lesser extent either by th MARK F. GRIMWADE, environment or by certain substances, or both, Gold alloys do not escape from this rule and the author throws light on the causes of mishaps which often leave goldsmiths Most metals react in some way or another with their surroundings. Some metals react with oxygen in air to form a stable oxide on their surface — a phenomenon which is enhanced by heat. Others dissolve in acids to form metallic salts. Some corrode rapidly in solutions of salts and thus deteriorate at the seaside, for example, while others appear to be quite unaf- fected by this environment. Many are also affected by contact with mercury, with which they form substances called amalgams. This behaviour of metals is related to the topics examined in this article — tarnishing, corrosion and_ stress corrosion cracking of gold alloys, and the effects of mercury. Without dwelling on the theory of the corrosion and chemical behaviour of metals and alloys, it is worth noting that chemical at- tack or corrosion is basically an electrochemical process. The metal being attacked by the substances in its environment behaves in a similar way to an electrode in an electrolytic cell or battery. The Electrochemical Series When a metal comes into contact with certain substances, whether potentially corrosive or not, a voltage difference is created be- tween the two. The value of this Figures Tab Effect of six days’ exposure to synthetic perspiration on the surfaces of two samples of yellow gold alloys. a) 18.carat b). 9 carat (Magnification : 50%) (Photos by Laboratoire Hanri Dubois, Swit erlan and jewellers somewhat bewildered, voltage depends on the particular metal-chemical_ system involved, but by comparing it with that for a standard electrode, it is found that metals can be assigned a unique value known as the standard electrode potential (E°). Table A shows the values of E° for some metals. Note that the ‘more positive the value, the more resistant is the metal ‘to attack. Gold is at the top of the list and is followed by platinium, palladium and silver in that order. Tt is be- 1a cause of their position in this list, which is known as the ELECTRO- CHEMICAL SERIES, that these metals are called the NOBLE METALS and are noted for their corrosion resistance. Resistance of Pure Gold to Chemical Attack Gold is often being referred to as the incorruptible metal par excel- lence because of its remarkable resistance to chemical attack in nature. Indeed, gold is totally un- affected by exposure to wet and dry atmospheres and it does not react with air or oxygen even at high temperatures. It is for this reason that gold is usually found in the free metallic state, some- times in an almost pure form but usually containing a proportion of silver. Gold ‘is also insoluble in most mineral acids including. sulphuric acid (HySO,), hydrochloric acid (HC1) and “nitric acid (HNO, ). ‘This property permits its separ- ation by dissolution of the silver during the final stages in certain refining processes or during assays by the cupellation method (fire assay). In both cases, when suf ficient silver is present in a gold alloy, it can be dissolved with hot concentrated nitric acid, lea pure gold behind asa residue. This process is known as parting. However, even pure gold can be dissolved in aqua regia (a mixture of three parts of concentrated HCI to one part of concentrated HNO) to give a solution of gold chloride: Under certain special conditions, gold may also dissolve in solutions of chlorides which contain dis- solved chlorine gas and in solutions 13of cyanides. This property is ex- ploited by the gold mining industry for extracting gold from crushed and finely ground ore by dissolving it ina dilute solution of sodium or potassium cyanide, Known as the cyanidation process, it relies on the fact that gold can be precipitated out again from such solutions by the addition of zine dust : 1 :Gold ore + Potassium cyanide —> Potassium gold cyanide + waste IL: Filter off waste IIL : Potassium gold cyanide + zine dust — Potassium zine eyan- ide + gold dust IV : Filter off the gold. The main application of eyan- ides in the jewellery industry today is as one of the essential constitu- ents of most gold plating and elec- tropolishing baths. Cyanide sol- utions are also used for etching metallographic specimens of gold and its alloys prior to microscopic examination. Corrosion and Tamishing of Carat Gold Alloys Carat_gold alloys contain. base metals as alloying additions, and their resistance to. corrosion and tarnishing is therefore not as high as that of pure gold. This is particu- larly the case with those alloys that have a microstructure con- sisting of two phases, one of which contains a high proportion of the base metal. For example, many nickel-con- taining white gold alloys consist of two phases (of which one is nickel- rich with relatively little gold) if they have been allowed to cool too 14 slowly after annealing. When in t condition, even 18 carat alloys of this type are readily attacked by nitric acid. When relatively large amounts of zinc are added to coloured carat gold alloys, their resistance to cor- rosion and tamishing may improve because zinc increases the range of concentrations over which these alloys are single phase. However, the alloys may then become more prone to stress corrosion cracking — an important topic which will be discussed later. Tamishing is a constant problem with silverware and brass and is the consequence of the presence — besides oxygen and moisture — of sulphur-bearing gases (eg. hydrogen sulphide) in the atmosphere. Hence it is not surprising that low carat gold alloys should also tamish, due to their high silver and base metal contents, Research work on gold-silver- copper (Au-Ag-Cu) alloys has shown that they are immune to tarnishing if the proportion of gold atoms in relation to the total A number of atoms in the alloy is greater than 50 per cent. Figures 1a- ’b shows that alloys falling into this category are located above a line drawn from the 15.6 carat gold- silver (Au-Ag) alloy to the 18 carat gold-copper (Au-Cu) alloy. Because they lie within this zone, 18 and 22 carat alloys do not nor- mally tarnish. It should be noted, however, that the presence of zine and nickel in modern commercial 9 and 14 carat golds, as well as their structural condition (whether they have been quenched to retain a single-phase structure or were slowly cooled to give a two-phase structure) can affect their resis- tance to tarnishing. Blackening of Skin or Clothes Tarnishing is a form of superficial corrosion that can be caused by substances other than sulphur-con- taining fumes, such as soaps, cos- metics, or perspiration. Sometimes customers complain that skin and clothing have been blackened by their gold jewellery. Although the Jewel itself may not actually looktarnished, it is likely that traces of tarnish have in fact rubbed off its surface and caused the observed blackening. Also, it has been sug- gested that because the composi- tion of perspiration varies from fone individual to another, some may be more affected by this problem than others. One method that is often used to inhibit skin blackening is to plate the surface of the article with a layer of 18 or higher carat gold alloy Mechanical abrasion is another cause of blackening whereby very fine metallic particles are transfer red from the surface of the article onto the clothing. Unfortunately, it seems that synthetic fibres used in textiles enhance this effect and again the only way of minimizing the consequences is to electroplate the article with a layer of hard gold. Stress Corrosion Cracking Superficial corrosion and tarnishing are not the only forms of environ- mental attack to which gold alloys of caratages of 14 or less may be susceptible, since localized and often invisible corrosion at_posi- tions where the alloys are highly stressed may lead to what is known as stress corrosion cracking. Articles of jewellery are often stressed in certain areas as a result of working or of other factors and Figure 2 Stress corrosion crack in a wrought 9 carat gold ring shank (Magnification : 25x), Figure 3 Stress corrosion cracking of a 9 carat gold medallion after hallmarking, hence the problem of stress cor- rosion cracking in 9 and 14 carat gold jewellery is often baffling to jewellers because it may occur not only during fabrication but also during subsequent storage or even after the jewellery has been sold to a customer, In fact, the basic causes of this embarrassing phe- nomenon are relatively simple to appreciate. As the term stress corrosion cracking suggests, cracking only occurs when stress and corrosion are combined either by subjecting the gold alloy to a stress whilst in a corrosive environment or, as is often the case, by exposing it to such an environment when it is already in a stressed state. Yet in the absence of stress, the gold ar- ticle would not suffer from that particular environment Scientific investigation into the behaviour of gold alloys in the presence of different. potentially corrosive substances has demon- strated that provided the gold con- tent of an alloy is greater than about 40 atomic per cent, stress corrosion is unlikely to occur. In practice this means that all 18 and 22 carat gold alloys should be im- ‘mune to stress corrosion cracking. Chemical reagents which have been shown to cause accelerated corrosion in the presence of a stress include nitric acid, hydro- chloric acid, aqua regia, potassium dichromate (K;Cr,0;), chromium trioxide (CrO, ) and ferric chloride (FeCl,). The fact that some of these reagents are common con- stituents of pickling baths gives a major clue to the reasons behind this type of failure, which may oc- cur not only at the fabrication stage, but also during subsequent storage if the jewellery is kept in an area that is contaminated, even lightly, by fumes from pickling tanks, Failure of low carat gold jewel lery through stress corrosion crack- ing after sale toa customer is less common, but can be readily ex- plained if it is found that the cus- tomer lives or works in an environ- ment where the atmosphere is con- taminated by dust or fumes con- taining chlorides 15Internal Stress An important contribution to fail- ure by stress corrosion are stresses that are tensile in nature, These may be either external stresses ap- plied to the jewel whilst in a cor- rosive environment or residual or internal stresses introduced during an earlier fabrication step. For ex- ample, in nearly all metal working processes — whether machines or hand work are involved — the met- al is deformed in a non-uniform manner and is left in an internally stressed condition. stresses. Nickel white alloys are particularly sensitive to such treat- ments and should, therefore, be cooled at a fairly moderate rate, by air cooling on an iron plate, for example. Prevention of Stress Corrosion Cracking Potential risks of stress corrosion cracking on finished jewellery may be avoided by a full anneal fol- lowed by slow cooling to room temperature. If a final work hard- Some examples of stress cor 4a rosion cracking due to internal stresses may be seen in the photo- graphs reproduced here. Figure 2 shows the shank of a 9 carat gold ring made by cold stamping, which cracked soon after the ring had been finished. Figure 3 is a photo- graph of the back of a 9 carat gold medallion which cracked about 24 hours after hallmarking. The crack was initiated by the localized de- formation and the resulting inter nal stresses which occured when the hallmarks were punched in. Figure 4a is microphotograph of a section of a 9 carat chain which cracked suddenly during etching in a 10 per cent solution of ferric chloride. It can be seen here that the stress corrosion cracks run along the grain boundaries, One further cause of internal stress which is not always taken into account is rapid cooling from high temperatures that takes place when an article is quenched directly after annealing to retain maximum, softness. The external areas cool at a faster rate than the core and the resulting uneven thermal con- traction causes localized internal 16 ened condition is preferred for added strength, the alternative is to apply a brief stress-relieving heat treatment at a relatively low tem- perature (250-300 °C). Asa result, the external stresses are eliminated whilst the work hardened state is essentially preserved ‘The Effect of Mercury on Gold Alloys The effects of mercury on many metals, including gold alloys, can- not strictly be described as corro- sion or chemical attack, However, it is appropriate to. discuss this topic here because it is notuncom- mon for a jeweller to be asked by a customer “Why has my gold ring turned white?” The cause can often be traced to the person having handled a broken mercury thermometer. ‘When put into contact, gold and mercury combine to form an amal- gam that may be pasty or liquid depending on the relative propor- tions of the two metals. Since mercury evaporates at a relatively low temperature, most jewellers know that gentle heating under a flame, followed by repolishing, generally restores the gold ring to its original condition. In fact, the author has even succeeded in treat- i gemstone-set rings by placing Figure 4a Stress conpsion erackingalong the grain boundarié® in a 9 carat gold chin link. (Magnification : 300x). Figure 4b Stress corrosion cracking ‘around grain boundaries in 9 carat gold casting. (Ntagnification : 300%).them on a thin asbestos pad heated from underneath, Because mercury vapour is highly toxic and should never be inhaled, every precaution must be taken to carry out this kind of operation in a well-venti lated area or, ideally, ina fume cup- board (a special ventilated cabinet). The fact that mercury can be evaporated by heating is also the basic principle behind two pro- h are now virtually ob- solete because of the serious health hazards involved. The first one is S - és ts6canar < < é o £ s € ¥ s WEIGHT PER CENT COPPER —> the application of gold coatings to metal substrates such as bronze o brass by fire gilding, which has fortunately been superseded by gold plating. The amalgam of gold and mercury was first spread evenly over the surface of the article to be gilded and then carefully heated to evaporate the mercury (hence the term fire gilding) to leave the gold coating behind. The other process that involved mercury in the past was the separ- ation of metallic gold from crushed CARAT él ore by amalgamation. Floating to the surface of the liquid amalgam the waste could be skimmed easily. The clean amalgam was then heated to evaporate the mercury, leaving gold and some other metals behind as a residue to be melted down for refining, . Figure 5 The three regions inthis ternary diagram indicate different degrees of resistance to tamishing and attack by acids for alloys of the Au-Ag-Cu system. The presence of zinc and nickel in such alloys affects the exact position of the zone boundaries, Zone A: Tarnish resistant, Also resis. tant to acids except aqua regia Alloys ig this region may be superficially attacked by acids which dissolve out sil vver and copper, leaving the surface richer in gold Alloys tarnish and may dis. integrate if left in strong acids, Zone B Zone C 722 CARAT, DARK YELLOW Au917—Ag32—CuS1 2218 General description A standard 22 carat dark yellow alloy. Used extensively for jewellery fabrication. Composition Component Au Ag cu % 917 32 SI | | Physical properties Colour Density Melting range Dark yellow giem3_ °C *() 178%) 964-982 Mechanical properties State of As | Cold*(3) | Cold*(s) _Annealed*(4) the alloy Cast | worked worked 600°C 20% 15 % 30 min, Hardness*(s) ee 0 120 165 70 Tensile*(6) Strength 36 50 28 daN/mm2 0.2 % Yield Strength*(6) 2 45 95 daN/mm? Elongation ete Ls 10 30 Erichsen*(7) Cupping 9.2mm Test Processing guidelines This alloy lends itself well to all the techniques of cold working (bending, stamping, cutting, etc). Best results are obtained by using fine-grained material and by re- specting the cold working and annealing recommendations given in this data sheet.Cold working Before annealing, the alloy should be cold worked to at least 50 per cent and preferably to 75 per cent. When wire drawing, even higher cold work- ing rates are possibie. Low cold working rates necessitate relatively high annealing temperatures and excessive grain growth may then take place. Effect of cold working ‘The graphs illustrate how hardness, tensile strength, yield strength and ‘on mechanical properties elongation vary with the degree of cold working of the alloy. zm. | | eo © = 2” * “ 159 z 2 0 = 2 100 _| |§ x0 2 | Z 5a a 2m = 22 a = 2 2 0 Z 10: go Zo Bo 3 leo Ie le le Ie] |# I. Ie Ie Ie Is] |2 I. Is le Is Ie cOLOWORKS COLDWORK® COLD WORK» The increase in hardness with inereasing The increase in tensile strength (A) and The decrease in elongation (ductility! amount of cold working. yield strength (B) with increasing amount with increasing amount of cold working, of cold working, Microstructure ‘The “as cast” structure of the alloy. Magnification: 200x. The structure after cold working to 75 per cent reduction, Magnification: 200%Annealing Effect of annealing temperature on mechanical properties ‘The recommended annealing treatment for material previously cold worked to 75 per cent reduction is 30 minutes at 600°C, Goldsmiths working with similar alloys may torch anneal small arti- cles at approximately 650°C (medium red in colour when observed away from direct sunlight or artificial lighting) for about 60 seconds. The graphs illustrate how hardness, tensile strength, yield strength and elongation of a 75 per cent cold worked sample vary with the tempera- ture of heat treatment (30 minutes at each temperature). ml 200 1 i "om # | | @ — 40. i 150 | z 2 = E400, Ee al 100 | | Soon) - = 5a | 8 = 2 3 am | | |3 3 = 2 too =" fo 3o sl. le is le Oo le ff le la) |2 1 le fs te ia ANNEALING TEMPERATURE °c ANNEALING TEUPERATURE %¢ ANNEALING TEMPERATURE*c ‘The change in hardness with temperature of heat treatment, Effect of annealing temperature ‘on microstructure ‘The change in tensile strength (A) and The change in elongation (ductility) yield strength (B) with temperature of with temperature of heat treatment. heat treatment. Structure of 75 per cent cold worked sample after annealing at 400°C for 30 minutes. Magnification : 200%. Structure of 75 per cent cold worked Structure of 75 per cent cold worked sample after annealing at 600°C for 30 sample after over-annealing at 700°C minutes, for 30 minutes, Note that excessive grain Magnification : 200x. growth has taken place. Magnification : 200x.Hardening Soldering Pickling Notes Source of data 22 carat gold alloys cannot be hardened by heat treatment to any significant degree. This alloy can be soldered easily using solder alloys of the same carat and colour and either a torch or a furnace. Hot “w) “ay 2 % ” The 10 per cent sulphuric acid. Stondard specications: NIHS.0850 {Jy 1961), Switsrond CETEHOR07-70 (Merch 1966), France DIN 8238 (1966), Germany Density varie slighty depending on the tate of the alloy, iz. cold worked, annealed or hhordened Previously annealed for 30 minutes at 600°C After cold working f0 73 per cent reduction Vickers hardness tester, Toad I ke Wire semples 100 mm fong. nominelly 2mm diameter Units: 1 div ~ 1.03 kg force (1.02 kp) Test according 10 DIN 30102 specification, Germany Diarncter of ball: 15 nina Semple: enneaied sheet 0.5 mm thick information in this data sheet was generously provided by Johnson Matthey Metals Limited, 43, Hatton Garden, London ECIN 8EE, United Kingdom.