INTRODUCTION
The investment casting process is perhaps the most widely practiced mass production technique in the jewelry industry.  Requirements for alloy performance are demanding, especially when recycling of expensive materials becomes a significant issue. Casters want and need a material that will provide latitude in handling and ease the difficulty of form filling on numerous designs. Processing conditions range from precise daily routines using controlled conditions on tested designs through first time trials with torch melting on new concepts. A material must be versatile to perform in these conditions. International standard dictates that platinum must be 95% pure, therefore, the alloying choices are limited. Platinum is metallurgically incompatible with many of the additives used with gold materials.

The desirable attributes of a superior investment casting material are outlined and experimental methods to evaluate these noted. Results indicate that a material with improved hardness has superior long term wear resistance as a jewelry article and finishes significantly easier with conventional processing techniques.

HISTORICAL AND THEORETICAL CONSIDERATIONS:
Based on a study done in the late 1970's at the Johnson Matthey Research Center, a large number of alloys were explored to overcome form filling deficiencies, poor color from surface oxidation and metal mold reactions in think sections inherent to the common commercial alloy of the day, 95% Platinum-5% Copper. They divided their study into high and low temperature alloys. The high temperature group included alloy combinations of Pt with Ruthenium, Iridium, Palladium, Cobalt and Nickel. Only Cobalt showed promise from improved form filling, minimal oxidation and desirable physical properties.

Low melting point alloys based on the binary eutectics of Pt with boron or silicon produced predictable brittle behavior from extreme hardness. Gallium use was also explored in combination with gold. Final recommendations of the research were to replace the standard Pt-Copper alloy with the Pt-Cobalt material for all investment casting applications and consider the benefits of a lower melting point alloy based on Gallium additions.

The issue of platinum casting alloy design was also considered by Volpe and Lanam at the 1997 Platinum Day presentation.  They considered the blending of copper and cobalt additions to reduce the issue of magnetism from cobalt, while not sacrificing the other desirable cast properties. As expected, the cobalt copper alloy showed superior form filling characteristics, but slightly poorer surface finish compared to the ruthenium material. Some difficulties with hand welding were attributed to copper oxidation.

To date this alloy is not widely used throughout the trade.

Information in the form of phase diagrams of platinum and the various alloying additions that provide a heat treatment response is scarce and usually restricted to binary relationships. The hardening response mechanisms vary from solid state ordering in the case of copper to limited solubility resulting in two distinctly different solid solutions, in the case of gold additions. The binary relationship between gallium and platinum indicates solid solubility in the platinum rich region with the formation of brittle inter-metallics possible. Solid state solubility is temperature dependent, indicating the possibility for precipitation hardening. The platinum-indium phase relationship is very similar. The ternary relationships between combinations of platinum, gold, gallium, copper and indium are not fully documented.

The relationship with various elements used to enhance the performance of gold and silver investment casting alloys is very different when considering platinum. Reducing conditions encountered during melting operations above 1400°C cause a number of reactions that do not occur in other precious metal.

Silicon:
Used extensively in both yellow and white gold to enhance fluidity and provide an oxide free cast surface. Silicon is not compatible with platinum. These compounds concentrate at the grain boundaries greatly diminishing physical properties.

Boron:
The same issues noted above occur.

The hard and brittle eutectic melts at 789°C promoting hot shortness and reduced ductility.

Phosphorous:
Ceramics used for investment casting contain phosphates as bonding agents. Reduction of the PO4 radical may allow brittle PtP2 to form. Such compounds can accumulate at the grain boundaries severely impairing formability.

Silver:
It dissolves into platinum alloys when molten. Low melting temperature PtAg3 or Pt3Ag phases accumulate at the grain boundaries causing hot shortness during assembly or welding operations.

Zinc & Cadmium:
These elements, used in gold alloys to reduce surface tension and enhance form filling, cannot be employed with platinum because of embrittlement and vaporization issues.

Many of the additives and principles that apply to other precious metal alloys do not produce positive results with platinum because of various metallurgical limitations.

PRACTICAL CONSIDERATIONS AND EXPERIMENTAL METHODS:
Desirable attributes of a platinum investment casting alloy are summarized as follows:

• 1. Good form filling capabilities or fluidity to fill thin sections
• 2. Low tendency to form an oxide during melting operations in neutral or oxidizing conditions required for platinum
• 3. No tendency towards forming brittle compounds with ceramic materials encountered in conventional melting operations.
• 4. A bright white color that does not require rhodium plating for enhancement or wear resistance
• 5. No excessive loss that hinders scrap recycling through re-melting, even at
• 1700-1950°C
• 6. Enhanced wear performance during service to provide a high quality finish throughout a conventional lifespan as a jewelry article.

At this point follows a detailed description of the testing parameters that can be seen in the original paper.

PROPERTIES COMPARISON TO THE PLATINUM HARD CASTING ALLOY:
A variety of physical properties are summarized with comparison to various alloys in use throughout the industry. 

Alloying additions required to produce significant hardening also reduce the melting range of 95% Pt materials substantially (125°C). The melting range is broad at 1000°C compared to the narrow 10-20°C typical of conventional platinum alloys. Color coordinates from the CIELAB system indicate a quality platinum shade. The hard casting alloy has superior hardness, tensile strength and yield strength compared to conventional alloys.

The hardness range has been specifically selected to match typical white gold materials in the as cast state. The additional 80 Vickers hardness points over the 5% iridium material represent a significant advantage in scratch resistance and strength as a jewelry article. The heat treatable alloy is too hard for conventional shop handling, but too soft for tension setting applications. The hard casting alloy represents a compromise in hardness selected to provide better wear resistance and faster finishing, while not being too difficult to handle in the shop environment.  The increased strength and lower density combine to provide design options for weight reduction that cannot be accomplished with the soft, low strength 95% platinum alloys. All of the alloys tested exhibited ductility. The hard casting alloys exhibits ductility comparable to conventional alloys such as 90%Pt- 10% Ir or 95%Pt-5%Co. These values indicate the alloy is suitable for stretching, hammer sizing, setting and other jewelry applications.

ECYCLING PERFORMANCE:
We tested the consistency and stability of physical properties under conditions of recycling. Complete reuse of a batch of metal 3 times represents an extreme abuse to indicate long term performance.  Results demonstrate that critical physical properties such hardness and ductility as indicated by percentage elongation during tensile testing do not deteriorate during abusive recycling. The material does not form brittle compounds with the ceramics used in conventional platinum processing. The retention of hardness and assay results indicate that the alloying additions are not adversely affected by extensive re-melting without the benefit of fresh material additions. Tests done on similar developmental alloys indicated the same performance through 5 cycles of 100% reuse in 2 independent tests.

FLUIDITY TESTING:
Results indicate that the 95% platinum hard casting alloy has fluidity and form filling ability comparable to the standard 90%Pt- 10%Ir material. Both are inferior to platinum cobalt. All lines on the phonograph models of each alloy filled completely.

This means details as small as 0.002" were reproduced. The sharpness of grooves failed to match the results found when zinc containing gold alloys wet gypsum investment.

Casting experiments were successful, achieving complete fills of a wide variety of jewelry articles without the use of excessive super-heat. The hard casting alloy has form filling characteristics suitable for jewelry manufacturing.

SURFACE OXIDATION:
The additives utilized to achieve enhanced hardness and strength are more prone to oxidation than PGM group metals that do not enhance hardness. A gross, thick oxide layer does not form. The 0.002" deep grooves of the phonograph model reproduced through all testing. Routine surface grinding during finishing operations easily removes the frosted surface. No network of silicates or other detrimental compounds were observed. Intentional oxidation of a finished surface at 700°C for 1 hour with protective atmosphere of flux produced no change in the surface color or finish. A thin gray adherent layer forms on the surface.

WEAR RESISTANCE TEST RESULTS:
Hardness is defined as resistance to indentation. This is a good expression for wear performance in a jewelry article. We studied how alterations to alloy chemistry alter hardness. Manipulation of the gallium to iridium ratio revealed the relationship Vickers hardness varies in a roughly linear relationship that provides a basis for selecting a hardness range that will provide resistance to indentation and scratching, while not being too high to affect other properties such as ductility. Values from the work of Rushforth et. al are displayed for comparison.  Our abrasion and indentation tests attempted to simulate consumer wear exposure.  We used equipment common to the industry to allow for repetition of the method by other individuals. Experimental parameters were as follows:

1. 1. A common vibratory finisher deployed aggressive black plastic cutting media
2. 2. Investment cast washers were machined clean and square. Height was cut after width to ensure the cutting burr could be removed cleanly. 1 surface was lapped to a 1 micron finish
3. 3. Pieces were 5.5mm high, 3.Smm wide and weighed 25g for tests 1&2. They were machined to 4.0mm high and 2.5mm wide for tests 3&4.
4. 4. 1 washer of each alloy was used simultaneously in each test.
5. 5. The washers were subjected to vibratory abrasion and examined at 10 minute intervals.
6. 6. Polish loss time was defined as the inability to read the date reflected from a 5 cent piece onto the surface of the test specimen. When the date could not be read, the elapsed time was recorded. 
7. 7. Weight loss was recorded and volume loss computed after 18 hours exposure.
8. 8. The loss of volume was computed from starting mass, density and weight loss values.
9. 9. Tests 1&2 were done in February 2000 by G. Normandeau while 3&4 were done in May by D. Ueno.

Results indicate that hardness and wear resistance correlate closely. We found variation in the time of polish loss time depending on the illumination and technique deployed. Test 1 was done by visual examination under regular plant lighting. We deployed a bright fiber optic light source with fixed orientation between the illumination, abraded surface and 5 cent piece for tests 2, 3, & 4. This roughly doubled the time to polish loss. Results were unitized to compare time to the 950Pt50Ir material in each test. This minimized the effect of illumination and operator judgment. In all 4 tests the 950Pt 50Ir lost its surface reflectivity first. The platinum hard casting alloy lasted from 2.6 to 3.6 times longer, 3.1 times longer on average, before loosing its surface reflectivity. Both the 900Ptl00Ir material and 950Pt50Co lasted 2.l times longer before polish loss. The 950Pt50Ru lasted 2.36 times longer on average than the soft 950Pt50Ir. Weight loss in milligrams was very small, even after 18 hours of testing. This caused the volume loss numbers to be small. Careful attention to surface preparation was required to yield accurate loss values.

Specimens used in tests 1&2 had their edges softened by hand buffing. The smaller specimens used in tests 3&4 received careful machining and lapping to ensure their edges were sharp and close to 90 degrees. This explains the 40% or more increase in volume loss between the two test series. The effect is especially significant on the 950Pt50Ir. The weight loss numbers do not reflect the 50% change in mass between test 1&2 samples versus 3&4. Volume loss in mg/ cm3 more accurately reflects wear and compensates for density differences. The hard casting alloy lasts 30% longer in wear tests than the best conventional alloy. It lasts 210% longer than the worst conventional material. Wear test accuracy is affected by machining methods, surface preparation techniques, abrasive media performance plus interpretation of reflectivity loss.  Weighing for determination of density and calculation of volume loss can be done extremely accurately with a Mettler A6245 precision density balance. More objective results could be obtained by using a surface roughness meter to correlate visual estimation. Results indicate that wear and hardness can be closely related. Careful selection of alloy additions can greatly improve wear resistance while maintaining the 95% platinum content required for unqualified hallmarking around the world.

MANUFACTURING ISSUES:
Melting:
Here follows detailed instructions as to how to melt the hard cast platinum alloy. This includes employing cover gas. Consult the actual paper for details

Flask Temperature:
Technical advice with temperature recommendations are found in this section of the actual paper.

Feeding Gates and Sprues:
The method of feeding molten metal to a platinum casting should be a special consideration compared with items cast in gold. This applies to any platinum alloy.  Feeding a material with a broad melting range is best accomplished through thicker than average feeding gates. As always, it is important to place these as close as possible to the thickest section of the cast item

The basic principles that apply to platinum casting, will promote quality product with the hard casting alloy as well. These have been documented at previous Platinum Day Symposiums.

Assembly Issues:
The hard casting alloy can be brazed with conventional platinum solders in the 1100 to 1400°C range. It can be self welded with the clean heat of an oxy-hydrogen torch. Care must be taken to accommodate its lower melting range. Pure platinum foil cannot be used for welding operations because of the melting temperature difference.

Field Trial Results:
Small scale field trials with torch melting and casting have been successful.

Virtually everyone who handles the materials notes that finishing times are 50% less compared to other platinum alloys. A high polish is easier to achieve with fewer steps. Smearing is reduced compared to softer alloys like 950Pt 50Ir. Operations that require reasonable softness such as bead setting can be done as well. Drilling for setting small stones produced less tool wear. In general setting operations required less time and effort. Efforts focusing on casting diamonds in place with the platinum hard casting alloy have been successful. In general, the material is well received in both the casting room and bench areas.

CONCLUSIONS & AREAS FOR FURTHER STUDY
1. Previous studies indicate limited alloying alternatives to achieve the multitude of properties desired from an investment casting material.

2. Many of the additives commonly used to enhance the properties of gold materials, cannot be used with platinum alloys because of metallurgical incompatibility.

3. The newly designed hard casting alloy has a platinum white shade that matches any conventional alloy based on the CIELAB color system.

4. The platinum hard casting alloy is 27% harder than the harder conventional materials and 95% harder than the 95Pt5Ir formulation.

5. Ductility of the hard casting alloy is equivalent to other materials with 20% higher strength combined with 8% lower density.

6. Recycling performance, defined as retention of hardness, strength and ductility, with suitable form filling is good.

7. Fluidity test results rank the new hard casting alloy close to 900Pt 100Ir. Both are inferior to the 950Pt50Co material.

8. Surface oxidation is not a significant problem based on furnace exposure tests at 700 and 1100°C for 1 hour. No gross dark oxide forms. Oxidation is equivalent to the common 950Pt50Co material. The thin layer present after casting or heating for soldering can be easily removed by conventional abrasive methods.

9. Wear tests applied to 5 common platinum alloys in a series of 4 distinct trials over 3 months, indicate that the hard casting alloy lasts 30% longer than the best conventional alloy before loosing a reflective finish . Superior results, lasting 210% longer than the soft 950Pt50Ir alloy, were repeated in all trials.

10 Platinum wear trials require careful attention to detail in surface preparation and weighing. Surface roughness measurement through an objective meter would be a good supplement to visually rating reflectivity loss.

11 Manufacturing trials supplemented by metallographic examination of a range of articles confirmed internal cast integrity. No incidence of non-metallic inclusions, gross shrinkage porosity or gas porosity was found.

12. Ceramic flask temperatures and casting feeding gates can be adjusted to obtain quality results over a broad range of common jewelry items

13. Field trials have produced positive feedback from casting operations, to assembly benches to finishing areas.  Finishing times are generally reduced 50% compared to softer conventional alloys.

V8N7

Platinum Alloy Design for the Investment Casting Process
Greg Normandeau & David Ueno
Imperial Smelting & Refining Co. of Canada Ltd.

This is an abbreviated version of the original work. For full technical details, please consult the original paper.