A Comparison of Techniques for Processing Powder Metal Injection Molded 17-4 PH Materials

Satyajit Banerjee

DSH Technologies, LLC and

Claus J. Joens Elnik Systems Division of PVA MIMtech, LLC 107 Commerce Road, Cedar Grove, NJ 07009

Paper No.: 2008-01-0305 Abstract While 17-4 PH stainless steel is a well established alloy in the powder metal injection molding industry, various techniques are used by different companies to produce the parts from this alloy. 17-4 PH is produced by sintering under partial pressures of nitrogen, sintering under hydrogen and sintering under vacuum. In this work parts made from the two different BASF feedstocks, made from a master alloy and gas atomized powder, were sintered for the same time and temperature under nitrogen, vacuum and hydrogen. The physical properties, microstructures and chemical compositions have been compared for the different sintering atmospheres and the causes for the differences discussed. Introduction 17-4 PH is the most widely used MIM stainless steel material. There are two basic ways the powders are formulated; either as a pre-alloyed atomized powder (the principal means of atomizing being gas or water) or by mixing carbonyl iron powder to a master alloy. Both these techniques result in acceptable products, depending on the sintering process employed. There are many different ways by which the 17-4 PH materials are sintered. The technology used by the founding fathers of MIM, Ray Wiech, Peter Roth, Karl Zueger and Ray Millett in the early 1980s, of pre-sintering in hydrogen to 1150oC, followed by removal to a vacuum furnace and sintering under vacuum at above 1300oC, is still being practiced. Some manufacturers use a partial pressure of nitrogen, or argon or vacuum while others use hydrogen under atmospheric pressures but at a lower temperature and longer time or hydrogen partial pressures at a high temperature in a refractory metal furnace. Zhang and German [1] had sintered 17-4 in vacuum, hydrogen-nitrogen mixtures and pure hydrogen and suggested 1300oC was the optimum sintering temperature. They found that vacuum and hydrogen-nitrogen atmospheres resulted in poor densities and lower corrosion resistance and the best corrosion resistance was obtained by sintering in hydrogen. This large variation in the sintering practice does cause a variation in the physical properties obtained which hopefully fall within the MIMA/EPMA specification limits. Neither of these specifications spells out how the alloy should be processed. It is the intent of this work to show the effects of different processing atmospheres on the end properties of MIM 17-4 PH materials.

Experiments The feedstocks used for these experiments were BASF 17-4 G (designated as G), which is made from a gas atomized powder and BASF 17-4 PH A (designated as A), which is a mixture of a master alloy blended with carbonyl iron powder. Tensile bars were molded by Arburg at their Lossburg Center in Germany, using an ISO 2740 [2] standard Type B MIM tensile bar mold supplied by IFAM, Bremen, Germany. Catalytic debinding was carried out in an Elnik Systems CD 3045 using the standard procedures for the 17-4 G and 17-4 A BASF materials and secondary debinding and sintering was done in an Elnik Systems MIM 3045 furnace under different atmospheres as discussed later. The 17-4 A material was sintered in 3 different atmospheres; hydrogen, vacuum and nitrogen, while the 17-4 G material was sintered in hydrogen, vacuum argon, and nitrogen. In case of the vacuum sinter runs, the parts were secondary debound and pre-sintered in hydrogen to 900oC after which it was sintered in vacuum. In case of the argon and nitrogen atmospheres, the debinding was done in the same gases. The 17-4 G material was sintered using a partial pressure of 400 mbar, while the 17-4 A material was sintered using a partial pressure of 800 mbar, the sintering time and temperature in both cases being 75 minutes and 1370oC respectively. One set each of the 17-4 PH A and the 17-4 PH G were processed through the secondary debinding stage and pre-sintered at 900oC in hydrogen to show the parts are effectively debound to an acceptable level by the processes used. BASF recommends that their 17-4 PH materials be sintered in hydrogen only. Hence the hydrogen runs are considered to be the base line for these experiments. The tensile bars were tested by an outside laboratory. Optical microstructures were first observed in house and the parts were etched in Fry’s reagent. The mounts were re-polished and the pictures taken by an outside laboratory, where the SEM and energy dispersive x-ray analyses were also conducted. The fracture surfaces were created by gripping the parts in a vice and breaking them by a bending action. The fracture surfaces were fresh and not treated in any manner. Carbon tests were performed using a LECO 200C carbon tester. The densities were measured by a helium pycnometer and the corrosion tests comprised of observations of the parts after immersion in 2% common salt solution at 60oC during a two week period. The designation 1 shows the least amount of corrosion and 5 the most. Results The physical properties obtained for all the sintering experiments are tabulated in Table 1.

Table 1: Physical properties

Sample UTS 0.2% YS

% Elong.

% Carbon Density Corrosion

A-H2 956.3 757.0 3.8 0.007 7.66 1 A-Vac 835.0 623.3 7.2 0.004 7.62 3 A-N2 863.9 665.3 1.5 0.004 7.29 5

A-Pre-sintered 0.027

G-H2 953.5 684.0 4.2 0.023 7.65 2 G-Vac 851.5 617.8 10.8 0.003 7.65 3 G-Ar 879.1 652.2 8.2 0.010 7.58 3 G-N2 1132.1 580.5 12.1 0.007 7.68 3

G-Pre-sintered 0.033 We observe the following from this table:

• Both powder types pre-sintered to 900oC show similar carbon contents. Carbon contents of the sintered parts are all very low, well below the minimum carbon requirements.

• The two powder types sintered in hydrogen show similar properties as do the two powder types sintered in vacuum.

• The gas atomized powders sintered in vacuum and argon also show similar properties. • Both powders sintered in nitrogen show different properties. The master alloy containing

parts did not sinter to the same density as the others and showed a significantly lower density, though the other mechanical properties are comparable.

• The gas atomized powder sintered in nitrogen had the highest tensile strength and percent elongation but the lowest yield strength and corroded at a faster rate than the samples sintered in hydrogen.

• Though the parts sintered in hydrogen, vacuum and argon show similar mechanical properties, the corrosion properties are not the same. Parts sintered in hydrogen have the best corrosion properties while parts sintered in nitrogen the worst.

Microstructures

Figure 1: A-H2

Figure 2: A-Vac

Figure 3: A-N2

Figure 4: G-H2

Figure 5: G-Vac

Figure 6: G-N2

Figures 1 through 6 show the microstructures of the samples sintered in hydrogen, vacuum and nitrogen for the master alloy based feedstock (A) and the gas atomized powder (G). The microstructures for the gas atomized powder sintered in argon have not been included because these are so similar to those of the vacuum sintered parts. In all the above figures the first picture is an unetched photo taken at 100X, while the second is an etched picture taken at 100X and the third is the etched picture taken at 400X. Figures 1 and 2 show the structures for the master alloy based feedstock sintered in hydrogen and vacuum. The structures are both similar, except perhaps the amount of pores and possible drag outs. Figure 3 show the same material sintered in nitrogen where the pores and drag outs are much higher. There also seems to be an impediment to sintering and the grains are smaller. Figures 4 and 5 show similar structures for the gas atomized powders sintered in hydrogen and vacuum, except for the amounts of pores and drag outs. Figure 6 in the unetched condition is similar to figure 5 in the unetched condition, but after etching reveal a finer grain size than in case of the other gas atomized powders. Based on the microstructures obtained the samples sintered in nitrogen (A-N2 and G-N2) and the corresponding samples sintered in hydrogen (A-H2 and G-H2) were sent out for a nitrogen analysis. The results are given below in Table 2.

Table 2: Nitrogen Analysis

Sample %

Nitrogen Sample %

Nitrogen

A-H2 0.0004 A-N2 0.18 G-H2 0.0004 G-N2 0.15

The samples sintered in nitrogen show a large amount of nitrogen in the parts while those sintered in hydrogen show almost none. SEM Studies Figure 7a show the fracture surface from the master alloy based part sintered in nitrogen. The fracture surface show the metallic matrix is ductile. Numerous spherical particles are seen, most of which are silica per the energy dispersive x-ray analysis (EDXA) shown in Figure 7b.

Figure 7a: Fracture surface of the master alloy powder based material sintered in

nitrogen showing many spherical inclusions.

Figure 7b: EDXA of the spherical particles seen in Figure 7a showing that most of these

are silica. Figure 8a show the fracture surface from the gas atomized powder sintered in vacuum. The fracture surface show the metallic matrix is mostly ductile. The spherical inclusion particles seen here are much less than those shown in Figure 7a. Most of these are chromium oxides per the energy dispersive x-ray analysis (EDXA) shown in Figure 7b.

Figure 8a: Fracture surface of gas atomized powder sintered in vacuum show a smaller

number of inclusions compared that seen in Figure 7a.

Figure 8b: EDXA of the spherical particles seen in Figure 8a showing that these are

chromium oxides.

Figure 9a shows the fracture surface of the master alloy powder based material sintered in hydrogen. The fracture surface show the metallic matrix is ductile. No oxide inclusions are seen on this fracture surface. The EDXA is that of the base 17-4 PH material.

Figure 9a: Fracture surface of the master alloy powder based material sintered in

hydrogen shows no inclusions.

Figure 9b: EDXA shows the spectrum for the 17-4 PH base metal.

Discussions The differences in the carbon content are negligible and well below the limits of 17-4 PH materials.

The density values obtained in this study show that the master alloy powder sintered in nitrogen does not reach full density. The nitrogen atmosphere reacts with the master alloy forming chromium nitrides on the master alloy surface. These chromium nitrides result in a diffusion barrier which retards the sintering process and densification. In case of all the other samples the densities reached were almost the same. The mechanical properties of the hydrogen sintered alloys are similar. The properties of the alloys sintered in vacuum and argon are similar and are close to those sintered in hydrogen. The properties of the samples sintered in nitrogen are different from the others because of the nitrogen content of these materials. It seems that all the sintered parts except that from the master alloy powder sintered in nitrogen have properties close enough in the as sintered condition so as to meet the heat treated properties of MIM 17-4 materials per the MPIF Standard 35 [3] developed by the MIMA. The corrosion tests performed however tell a different story. The corrosion test performed was one which is harsher than those suggested in the MPIF Standard 35 [3], but is closer to that performed by the medical device industry where Ringer’s Solution [4], an artificial body fluid solution, is used. In the event the material is used in contact with the human body, this test is more relevant than those suggested in the MPIF Standard 35. The tests conducted show that the samples sintered in hydrogen are more resistant to corrosion than the others, something which is neither expected nor evident from the mechanical test results. The metallographic structures are not easy to interpret. The differences between the unetched structures for the hydrogen and vacuum sintered samples are small and become complicated because of the presence of the fine (between 2 to 5 µm in diameter) spherical oxide inclusions as seen in the SEM studies. These oxides are easily dragged out during polishing and show a “pore count” greater than that expected from the density measurements. The pores and drag-outs become more difficult to differentiate after etching. This complicates the interpretation of the microstructures. The etched microstructures for the samples sintered in hydrogen, argon and vacuum show the typical as sintered microstructures of 17-4 PH materials; dendritic like delta ferrite in a martensitic matrix. In case of the samples sintered in nitrogen, chromium nitrides change the structures observed. In case of the gas atomized powders, a finer grain structure is observed, while in case of the master alloy based powder the chromium nitrides prevent densification and homogenization of the elements by diffusion from the master alloy into the carbonyl iron powder. The fracture surfaces observed by the SEM verify the presence of oxide inclusions in samples sintered in nitrogen, argon or vacuum. Silica was observed only in case of the master alloy based powder mixture and probably comes from one of the constituents. Chromium oxides were observed in all the above cases. The interface between the oxides and the metallic matrix result in potential differences that set up the galvanic currents needed to initiate corrosion. Hence though these samples meet all the mechanical requirements of the 17-4 PH alloy, they behave poorly under severe corrosive conditions and fail to hold up as one sintered in pure hydrogen or a

wrought alloy would. The formation of chromium nitrides has the further effect of reducing the chromium available to form the protective oxide that gives the stainless steel the corrosion protection ability. This renders the parts sintered in nitrogen more prone to corrosion. Banerjee [5] had conducted experiments in hydrogen with leaks on 17-4 PH MIM alloys and the results also showed poor corrosion behavior when the hydrogen is impure. This is because thermodynamically, per the Ellingham diagrams [6], you need a pure reducing atmosphere at high temperatures to reduce SiO2 to metallic Si or Cr2O3 into metallic Cr which will dissolve into the matrix at high temperature. Should the oxygen content of the atmosphere increase above a certain level, the reduction process will stop and could revert itself depending on the oxygen partial pressure and the temperature at which this reaction is taking place. Hence purity of the hydrogen is also crucial in proper processing of 17-4 PH alloys. Though experiments with 316L or other low carbon stainless steels were not conducted in this experiment, the thermodynamics of reducing oxides of silicon and chromium, which are stable at high temperatures, remain the same. This means that to obtain the best sintered properties all low carbon stainless steels must be processed in pure hydrogen atmospheres and sintered at a temperature high enough for the hydrogen to reduce these oxides. If these alloys are processed in vacuum, or argon or nitrogen or in wet or impure hydrogen the sintered properties will fail to meet the properties of parts sintered in pure hydrogen or the wrought alloy. Hence, depending on the processing technique used the alloy can be changed and may cease to be the alloy it was intended to be. The above statement is true irrespective of the binder system used to MIM low carbon stainless steel type alloys, because the binder is a carrier to mold the part. After the binder is properly eliminated, it is the sintering process which determines the material composition. In the 1980s the only hydrogen furnaces available could either be used to 1150oC or needed an Inconel retort which limited the temperature to around 1200oC and required frequent retort replacement. Hence part manufacturers have used either vacuum or nitrogen furnaces that could reach temperatures over 1300oC to produce these materials to the desired higher density. This is no longer the case today when hydrogen furnaces specially designed for MIM which are capable of attaining temperatures over 1400oC are available. We believe that the MIM standards should be changed to process all low carbon MIM stainless steels in hydrogen above 1300oC so that parts with properties and chemical compositions closest to wrought materials may be provided to customers. Conclusions

• MIMed low carbon stainless steels such as 17-4 PH and 316L should be processed in pure hydrogen for results comparable to wrought products. Zhang and German [1] and BASF are correct in their recommendations that these alloys be sintered in hydrogen. This recommendation also holds true irrespective of the binder system used to mold the parts.

• Low carbon stainless steels being processed under nitrogen result in nitrogen adsorption into the alloy, which could result in chromium nitrides in the structure.

• Low carbon stainless steels being processed under vacuum, argon or nitrogen have oxide inclusions which result in poor corrosion resistance.

• Impure or dirty hydrogen results in parts with poor corrosion resistance [5]. • Low carbon stainless steels may cease to be the alloy they were supposed to be

depending on the technique used to process the alloy. • MIM standards should be changed to process all low carbon MIM stainless steels in

hydrogen above 1300oC so that parts with properties and chemical compositions closest to wrought materials may be provided to customers.

Acknowledgements The authors would like to thank Arnd Thom of the BASF Corporation, Ludwigshafen, Germany for supplying the feedstock used for this study and Hartmut Walcher of Arburg Corporation, Lossburg, Germany for molding the tensile bars used in this study. References 1. H. Zhang and R. M. German, “Powder Injection Molding of 17-4PH Stainless Steel”,

Powder Injection Molding symposium, 1992, Edited by Philip H. Booker, John Gaspervich and Randall M. German, MPIF, Princeton, NJ, 1992, p. 219.

2. ISO 2740:2007(E), Sintered metal materials, excluding hardmetals – Tensile test pieces, ISO, Geneva, Switzerland, 2007.

3 MPIF Standard 35, Materials Standards for Metal Injection Molded Parts, Issued 1993 Revised 2000 and 2007, MPIF, Princeton, NJ, 2007.

4. “Ringer’s Solution”, www.whonamedit.com/synd.cfm/2119.html. 5. Satyajit Banerjee, “Structure – Property Relationship of Metal Injection Molded 17-4 PH

Orthodontic Parts”, Powder Injection Molding symposium, 1992, Edited by Philip H. Booker, John Gaspervich and Randall M. German, MPIF, Princeton, NJ, 1992, p. 181.

6. Ellingham diagrams, “An Introduction to Metallurgical Thermodynamics”, David R. Gaskell, McGraw Hill, New York, 1981, p. 287.