Machinability Additives for Improved Hard Turning of PM Steel Alloys

Transcription

1 Machinability Additives for Improved Hard Turning of PM Steel Alloys Bruce Lindsley and Chris Schade Hoeganaes Corporation Cinnaminson, NJ 08077, USA ABSTRACT The machining of ferrous PM alloys differs considerably from wrought materials. The role of porosity and heterogeneous microstructures complicates the machining process, often making it more challenging. In addition, the presence of martensite in the microstructure of more highly alloyed and/or sinterhardened PM components increases tool wear. One advantage of PM is that machinability additives can be easily admixed into the powder and therefore into the final part. Manganese sulfide is a well known additive for improving machinability. A new machining additive, designated MA, has been developed to compliment MnS in PM steels. Hard turning tests were performed to evaluate the effect of both additives on tool wear in different material systems. The MA additive was found to improve machinability beyond that of MnS in sintered compacts containing martensite. It additionally reduced rusting on the part surface. This paper discusses the improvement in machinability with these additives, with an emphasis on sinter-hardenable PM steels. INTRODUCTION Ferrous PM is generally considered a net or near net shape process. Nevertheless, many pressed and sintered parts are machined prior to final assembly, as certain features must be manufactured by secondary machining operations [1]. These machining operations include transverse holes (drilling) and undercuts (hard turning). PM parts are generally considered more difficult to machine than their wrought counterparts. The machinability of PM steels differs from that of wrought steel due to the presence of porosity and the often heterogeneous microstructure [2]. The porosity in PM steels makes lubrication more difficult and reduces heat transfer from the cutting surface. Additionally, porosity produces an interrupted cut with the tool, resulting in micro-impacts and a fatigue condition at the edge of the tool. To achieve similar strengths as wrought products, PM alloys typically contain higher levels of carbon. As the apparent hardness of a PM specimen approaches that of a wrought product, the microindentation hardness must be significantly higher to offset the effect of porosity. The higher microindentation

2 hardness in combination with the porosity is most responsible for the different machining response of PM compared to wrought steels. PM, however, has the advantage of being able to admix materials into ferrous alloys, and both powder producers and part makers have taken advantage of this ability by incorporating machinability additives, such as MnS, into the steel. These free machining additives improve machinability by assisting in chip formation, lubrication of the tool face and reduction of crater wear [2]. The presence of MnS in wrought steels is used extensively in free-machining grades. During the cutting process, MnS deforms along the shear plain, reduces tool contact time, and forms a lubricating layer on the tool [3]. Several studies have shown the benefits of MnS additions to the machinability of PM steels [1,2,4-6]. While manganese sulfide has many beneficial attributes, it does have some limitations and potentially negative effects. The use of MnS may damage the sintering furnace through the production of a sulfur containing gas [4]. High humidity environments can quickly oxidize the MnS and deteriorate the machinability enhancing properties [5]. In addition, MnS becomes less effective as the amount of alloying increases [6]. With the advent of sinter-hardening alloys, the machinability of these materials in the as-sintered condition is considerably more challenging. The hard, martensitic microstructure is more difficult to machine, and requires advanced tooling that can withstand higher forces and increased wear tendencies. In addition, traditional machining additives may not work effectively under these conditions. A new machining additive (MA) has been developed to enhance the machinability of sinter-hardened PM steels. The benefit of this additive will be discussed in sinter-hardened alloys, hybrid steels and in more traditional as-sintered iron-copper-carbon and iron-nickel-carbon steels. EXPERIMENTAL PROCEDURE The evaluation of different machining additives was conducted with a range of alloys, including prealloyed, hybrid, diffusion alloyed and premixed systems. The alloys studied were FLC (Ancorsteel % Cu + graphite), FLN4C-4005 (Ancorloy 4 + graphite), FD-0405 (Distaloy 4800A, + graphite), FN-0205 (Ancorsteel A1000B + 2% Ni + graphite), and FC-0208 (Ancorsteel A1000B + 2% Cu + graphite). The compositions are listed below in Table I. The chemical composition of Alloy 2 is equivalent to FD To each of these alloys, three conditions were tested: no additive, 0.35% MnS, and 0.3% MA. The mixes were compacted into rings measuring 25.4 mm ID, 44.5 mm OD and 28 mm high at a density of 7.0 g/cm 3. The rings were then sintered in a 90% nitrogen 10% hydrogen atmosphere at 1120 C for 15 minutes at temperature. A typical furnace cooling rate of 0.6 C/sec was used; no accelerated cooling was employed. FLC2-4808, FLN4C-4005 and FD-0405 samples were tempered at 205 C for 1 hour after sintering. The machinability testing was conducted at Lehigh University, and in addition, hard turning customer trials were run on Alloy 5. Turning samples were mounted on a lathe and cutting passes were made to the outer diameter of the sample. The depth of cut was 1.25 mm (0.05 in.) per pass and the feed was 0.3 mm (0.012 in) per revolution. A coated carbide tool (Seco-Carboloy TX-150) with a chip breaker design was initially used for the FC-0208 composition. However, the chip breaker design was not tough enough for the more heavily alloyed compositions, so the same coated carbide tool without the chip breaker (SNMA 433) was used for Alloys 1-4. In addition, a boron nitride (CBN 20) tool of similar geometry (SNGN 433S-LF) was also tested with Alloys 1-4 and a constant cutting speed of 2.5 surface meters per second (500 sfm) was used. The tool wear data has been assigned a value of zero after the first Ancorsteel and Ancorloy are registered trademarks of Hoeganaes Corporation

3 Table I. Nominal compositions (in wt%) of the alloys studied, balance Fe. Alloy # Designation Ni Mo Mn Cu Gr 1 FLC FLN4C trace FD trace FN trace FC trace cuts to eliminate the effects of the break-in period. Thereafter, tool wear was measured after every fifth cut on the flank of the tool. The effect of machining additives on surface rusting was studied on samples of FC-0208 and FD-0405 that contained no additive, 0.35%MnS or 0.3%MA. Several standard TRS samples were placed in a controlled environment after sintering and exposed to an average relative humidity of 87% and temperature of 24 C (75 F) for one month. Gloves were used to avoid fingerprints on the samples prior to the corrosion testing. The surface fraction covered by rust was measured for each condition using quantitative image analysis. RESULTS Microstructure The microstructures for Alloys 1 through 5 are shown in Figure 1. The cooling rate of 0.6 C/sec produced a nearly fully martensitic microstructure with some bainite in Alloy 1 (Figure 1a). The composition of this alloy is such that slow cooling rates can be used to produce martensitic microstructures. Alloys 2 and 3 have the same nominal composition, but a significant microstructural difference can be seen between the two alloys. Both alloys contain martensite and Ni-Cu rich light etching areas around the porosity in the structure. However, the centers of the base iron particles have different microstructures. Alloy 2, which is made with a prealloyed Mo base, consists of a divorced pearlite / bainite microstructure. Mo reduces the eutectoid carbon content in steels [7], and at 0.5% Mo, the eutectoid composition is nominally 0.6% C; hence Alloy 2 has no hypoeutectoid ferrite. Alloy 3 is made with a commercially pure iron base and consists of ferrite and lamellar pearlite. Much of the pearlite is unresolved at these magnifications. The Mo in the diffusion alloyed Alloy 3 does not fully diffuse to the center of these particles. Alloy 4 (FN-0205) is ferritic and pearlitic with nickel-rich regions and alloy 5 is fully pearlitic. The alloys chosen produced a wide range of microstructures at this cooling rate. The sintered carbon and average apparent hardness is given in Table II. The sinter-hardening Alloy 1 produced a high hardness corresponding to the martensitic microstructure in the sintered and tempered condition. The difference in hardness between Alloys 2 and 3 is a result of the slight different carbon content and the difference in Mo distribution. It has been shown that a difference in 3 HRA between FLN4C-4005 and FD-0405 at the same sintered carbon content can be expected due to this effect of Mo [8]. The FN-0205 is the softest material of the group, given its ferritic and pearlitic microstructure.

4 (a) (b) (c) (d) (e) Figure 1. Microstructures of the alloys at 0.6 C/sec (a) Alloy 1 (FLC2-4808), (b) Alloy 2 (FLN4C- 4005), (c) Alloy 3 (FD-0405), (d) Alloy 4 (FN-0205) and (e) Alloy 5 (FC-0208).

5 Table II. Sintered carbon (wt%) and average apparent hardness. Alloy # Designation Sintered C Hardness (HRA) Hardness 1 FLC HRC 2 FLN4C HRB 3 FD HRB 4 FN HRB 5 FC HRB Hard-Turning Machining of sinter-hardened materials is a difficult challenge for PM part makers and it was discovered that utilizing MA additive is quite advantageous in these materials. When the martensitic Alloy 1 was cut with the coated carbide tool (Figure 2), both MnS and MA reduced tool wear; the MA additive was, however, much more effective. The tool wear rate of the MA containing alloy is similar to the wear rates found with the much softer FN-0205 alloy (Figure 5) under the same cutting conditions. Considering that the no additive FLC alloy resulted in accelerated wear toward failure after 35 cuts, the MA additive greatly reduced the wear rate in this alloy system. Interestingly, under these same cutting conditions, when Alloy 1 was cut with the boron nitride tool, the tool wear rate was low for all three conditions. No benefit of either machining additive was observed. It is thought that the CBN tool is hard and strong enough to effectively cut the martensitic microstructure without additives. In addition, martensitic microstructures are relatively hard and have low ductility, and therefore the benefit of chip breaking that machining additives produce is not required. Figure 2. Tool flank wear with number of cuts for different additive conditions. Alloy 1 (FLC2-4808) machined with coated carbide and boron nitride (CBN) tools. The surface finish is also an important variable in machining. As shown in the Table III, the surface finish of the workpiece machined with the carbide tool improves with machining additives. Again, the MA additive exhibits improved machinability over the MnS additive. The surface finish is less rough

6 with the MA containing mix. With the CBN tool, the surface finish was worse with machining additives compared with the no additive mix. The MnS mix was considerably worse. Overall, the surface finish is superior with the CBN tool compared with the carbide tool under these conditions. The combination of the carbide tool with the MA-containing mix also produced adequate surface results, and when combined with the good wear results, may provide a cost-effective alternative to the more costly CBN tooling. Table III. FLC hard turning surface finish (in micrometers) Tool \ Additive No Additive 0.35% MnS 0.3% MA Carbide CBN Figure 3. Tool flank wear with number of cuts for different additive conditions. Alloy 2 (FLN4C-4005) machined with coated carbide and boron nitride (CBN) tools. Figure 4. Tool flank wear with number of cuts for different additive conditions. Alloy 3 (FD-0405) machined with coated carbide and boron nitride (CBN) tools.

7 Tool wear for Alloys 2 and 3 is shown in Figures 3 and 4, respectively. Given the same nominal compositions, it was expected that these materials would behave in a similar manner. With no machining additives present (open, black circles in Figures 3 and 4), the machinability of FLN4C-4005 and FD-0405 cut with a carbide tool was considerably different. Tool wear after 50 cuts of the FLN4C-4405 was 70 µm whereas the tool wear with FD-0405 was 225 µm under the same conditions. It was unexpected that tool wear would increase as the hardness decreased as the alloy changed from FLN4C-4405 to FD With the CBN tool, the wear was similar for the two alloys, and unexpectedly, was much higher than the CBN tool wear with the FLC (Figure 2). Longer chips were formed in Alloys 2 and 3 compared with Alloy 1, perhaps resulting in increased wear of the CBN tool. Within the nominal composition of Alloys 2 and 3, the best machinability was found with the carbide insert and FLN4C The different tool life between Alloys 2 and 3 is thought to be due to the Mo distribution in the material and its effect on the resulting microstructure. The introduction of MA into either alloy reduced tool wear for both cutting tools (solid, blue squares in Figures 3 and 4). The MA provided consistently lower tool wear in all cases. The combination of the CBN tool with alloy FD-0405 resulted in the lowest tool wear. It is interesting to note that for these heterogeneous microstructures, MA provided the best improvement with the CBN tool, whereas when the microstructure was martensitic (FLC2-4808), MA gave the biggest improvement with the carbide tool. It is evident that machinability is dependent upon many variables and microstructure is an important variable when choosing cutting conditions and tool materials. The effect of manganese sulfide on tool life in these alloys was mixed (solid, red diamonds in Figures 3 and 4). No beneficial effect was observed in the FLN4C-4405 alloy for either tool material. In fact, tool wear was slightly higher with MnS and the carbide tool. Although the effect of MA was significant, in the FD-0405 diffusion alloy, the greatest reduction in tool wear was realized with the MnS additive. The CBN tool wear was similar for the MA and MnS additions, but the MnS lowered carbide tool wear considerably more than the MA. This difference in MnS behavior between the two alloys is surprising and cannot be fully explained. Manganese sulfide may be less effective in harder materials [6], so the increase in hardness of the FLN4C-4405 may reduce the effectiveness of the additive. The dramatic change in behavior with MnS, especially with the CBN tool, is puzzling. It is possible that the condition of the MnS changed, either before admixing or after sintering. The inherent oxidation problem with MnS may be responsible for the different behaviors observed. Finally, surface finish was largely unchanged between the two alloys or with machining additives (Table IV). Some roughening was found with the addition of machining additives and the carbide tool. The surface roughness was not related to tool wear with these alloys. Table IV. FLN4C-4405 and FD-0405 hard turning surface finish (in micrometers) Alloy Tool \ Additive No Additive 0.35% MnS 0.3% MA FLN4C-4405 Carbide CBN FD-0405 Carbide CBN The tool wear for FN-0205 is shown in Figure 5 and little wear was observed with either additive. The carbide tool wear was similar for MnS and MA, while some additional benefit of MnS was observed with the CBN tool. The surface finish was good in all cases. Again, it is interesting that the carbide tool wear with MA in the FN-0205 is similar to that of FLC One would not expect the tool wear to be similar with a hardness change of 57 HRB to 30 HRC for FN-0205 and FLC2-4808, respectively. This highlights the benefit of MA additive with sinter-hardened materials.

8 Figure 5. Tool flank wear with number of cuts for different additive conditions. Alloy 4 (FN-0205) machined with coated carbide and boron nitride (CBN) tools. The effect of MA additive on machinability in a non-sinterhardening system (FC-0208) was also investigated, Figure 6. The effect of turning speed was studied in this alloy system. At slower speeds, the MA additive resulted in significantly less wear than the MnS additive. The tool with the no machining additive sample catastrophically failed after 10 cuts at this speed. Recall that a different tool geometry was used with the FC The chip breaker design proved to be insufficiently robust to cut this material. Also note the higher tool wear in Figure 6 compared to Figures 2-5. The tool wear of the MA containing sample at 0.6 m/s was similar to that of the FN-0205 at 2.5 m/s, while the tool wear of the MnS sample after 30 cuts is higher than tool wear axis of Figures 2-5. However, at higher speeds, the behavior of the MnS and MA reversed, with the MnS resulting in virtually no wear. It is thought that at this higher speed and higher energy condition, the MnS becomes lubricious and is able to coat the tool face. The cutting speed is obviously an important variable on the effectiveness of each additive. Customer trials on alloy FC-0208 with the three additive conditions showed similar tool wear results. Hard turning tests were conducted over a range of speeds and a smaller depth of cut (DOC) was used. It was again found that the cutting speed influenced the effectiveness of each machining additive. Under the different test conditions used in the customer trials, most notably the smaller DOC, the MA additive outperformed both the MnS and no additive samples at 2 m/s (400 sfm) and 3 m/s (600 sfm). The results of these trials at 3 m/s (600 sfm) are shown in Figure 7. At a speed of 4 m/s (800 sfm), the MnS additive was superior to the MA additive. The cutting speed at which the transition from superior MA performance to superior MnS performance occurs changes with different cutting conditions. The cutting conditions for FC-0208 should be optimized dependent upon machining additive.

9 Figure 6. Tool wear during machining of Alloy 5 (FC-0208) at two surface speeds and the three additive conditions Tool Wear (x 0.001") FC-0208 Baseline FC MnS FC MA CaAl2O3 Additive Number of Passes 600 SFM 0.010"/rev 0.010" DOC Carbide KC9110 Insert Figure 7. Customer evaluation of machining additives in FC Flank tool wear was measured on the coated carbide insert. Overall, the MA additive provided consistently lower tool wear for all conditions compared to the samples with no additive and is an effective machining additive. Manganese sulfide was also found to be effective in select alloys and conditions. Given all of the possible machining conditions present in the

10 PM industry, it is recommended that both additives be evaluated to determine the best machining response for each application. This assumes that all other factors outside of machinability are equivalent. Surface Corrosion The samples that were exposed to the warm, humid environment exhibited different levels of rusting. Due to the different alloy content, the FC-0208 samples had more rust than the FD-0405 samples. Within each alloy system, the MA containing samples were very similar to the samples that contained no additive. The MnS containing samples showed a marked difference from the other two sets. In both alloy systems, the surface of the MnS containing samples had considerably more rust. Typical surface conditions are shown in Figures 8 and 9. The surface fraction of rust measured by image analysis is shown in Table V. The amount of rust on the surface of the MnS samples is almost an order of magnitude higher than either the no additive or the MA containing samples for each alloy system. More than half of the 0.35% MnS FC-0208 sample surface was covered with rust, whereas the 0.35% MnS FD had many isolated islands of rust on the surface. The presence of MnS in both alloy systems greatly increased the amount of rust on the sample surface. The MA additive appeared to have no effect on sample rusting. Since both additives have been shown to improve machinability over a range of alloys and cutting conditions, the use of MA will be preferable in certain applications given the accelerated surface corrosion of test samples containing MnS. Table V. Area percent of rust after humidity testing. Area Fraction of Rust (%) Alloy No Additive 0.35% MnS 0.3% MA FC FD No Additive 0.35% MnS 0.3% MA Figure 8. Surface of alloy FC-0208 after humidity testing. No Additive 0.35% MnS 0.3% MA Figure 9. Surface of alloy FD-0405 after humidity testing.

11 CONCLUSIONS The benefits of a newly developed machining additive (MA) have been presented. The addition of the MA additive resulted in consistently lower tool wear across the board compared with mixes with no additive. This additive is particularly effective in PM alloy systems that contain mixed to fully martensitic microstructures. In FLC and FLN4C-4005 alloys, MA additive also generated less tool wear than MnS. In FD-0405 and FN-0205, MnS was either equivalent to or better than the MA. The MA additive was also effective at select speeds in the iron-copper-carbon system, and its usefulness in this alloy system will be application dependent. The MA additive has the additional benefit of being chemically inert and non-reactive under typical environmental and sintering conditions. This benefit was realized in both the consistent machinability results and in the corrosion testing. The presence of MA additive in sintered compacts did not cause the accelerated rusting found on the surface of samples containing MnS. ACKNOWLEDGEMENTS The authors would like to thank Ed Force at Lehigh University for performing the machinability experiments and Arthur Rawlings, Craig Gamble, Paul Kremus and Ron Fitzpatrick for their assistance in sample preparation and testing. REFERENCES 1. H. Sanderow, J. Spirko and R. Corrente, The machinability of P/M materials as determined by drilling tests, Advances in Powder Metallurgy & Particulate Materials, compiled by R. A. McKotch and R. Webb, Metal Powder Industries Federation, Princeton, NJ, 1997, part 15, p R. Causton and C. Schade, Machinability: a material property or process response?, Advances in Powder Metallurgy & Particulate Materials, compiled by R. Lawcock and M. Wright, Metal Powder Industries Federation, Princeton, NJ, 2003, part 7, p E. M. Trent, Metal Cutting, 2 nd edition, Butterworths, London, H. Suzuki, M. Yoshida, H. Tanaka and Y. Ikai, 250MSA resulfurized high green strength steel powder, Advances in Powder Metallurgy & Particulate Materials, compiled by R. A. McKotch and R. Webb, Metal Powder Industries Federation, Princeton, NJ, 1997, part 15, p S. Cleays and K. Chopra, Enhanced machinability and oxidation resistance with MnS containing additions to iron, Advances in Powder Metallurgy & Particulate Materials, compiled by R. A. McKotch and R. Webb, Metal Powder Industries Federation, Princeton, NJ, 1997, part 15, p B. Hu and S. Berg, Optimizing the use of manganese sulfide in P/M applications, Advances in Powder Metallurgy & Particulate Materials, compiled by H. Ferguson and D. Wychell, Metal Powder Industries Federation, Princeton, NJ, 2000, part 5, p ASM Handbook, Vol. 4, Heat Treating, ASM International, 1991, p F. Semel, Ancorloy premixes: binder treated analogs of the diffusion alloyed steels, Advances in Powder Metallurgy & Particulate Materials, compiled by C. Rose and M. Thibodeau, Metal Powder Industries Federation, Princeton, NJ, 1999, part 7, p