Response to carbonitriding heat treatment of a PM steel, from a diffusion bonded power, sintered in different conditions | Risposta al trattamento di carbonitrurazione di un acciaio da polvere diffusion-bonded sinterizzato in condizioni diverse

The microstructural characteristics observed on repeatedly sintered samples are as follows: -from 1 to 4 sintering steps (1120 °C, 30 min.), upper bainite, featuring various morphologies, prevails, followed by polygonal ferrite, transforming austenite (to upper bainite), a low amount of residual austenite: -starting from 8 sintering steps, the ferrite islands at sample surface progressively decrease, to nearly disappear after 16 sintering steps. Upper bainite always prevails. On the outer area austenite transforms to upper bainite, in a manner which becomes more evident after 16 sintering steps; -on samples sintered 1 or 2 times, the solid solution is strongly inhomogeneous, due to the limited diffusion of Cu and Ni, confirmed by chemical concentration profiles; -the in-homogeneity decreases after 4 sintering steps, to nearly disappear after 16 steps. Correspondingly, microstructural homogeneity increases; The microstructure of samples sintered at 1250°C for one hour includes bainite, very fine pearlite and residual austenite, increasing towards the periphery and mostly localized around fine pearlite. Frequently, the residual austenite shows a beginning of bainitic transformation. The microstructure points a low cooling speed and some C-enriched (up to 0,45%, on about 0,7 mm thickness), owing to addition of CH4 to the N2/H2 furnace atmosphere. The toughness drop observed on samples sintered at 1250°C can be attributed to this carbon enrichment, and to a more homogeneous solid solution; -to compare materials at equal C content and cooling speed, some samples, already sintered at 1250°C, one hour, have been "resintered", for half-hour, at 1120°C. In a first approximation, the homogeneity level already attained after one-hour at 1250°C is not significantly modified by a subsequent soaking time of 30 minutes at 1120°C. The observed microstructures (fully comparable to those observed on samples sintered 4 to 8 times at 1120°C) are due to restoration of the correct C content (0.2% w) and to a higher cooling speed. Austenite, always present, grain-shaped, is linked to the chemical composition, less unbalanced by high Ni values. Its "cushion" effect, to damp and redistribute stresses, is restored, while toughness and suitability to deformation increase again, to levels similar to those of materials sintered from 4 to 8 times, without significant U.T.S. drop; -strength and toughness increase as the number of sintering steps at 1120°C increases. This result agrees with the microstructural observations, showing a gradual increase of acicular structures, due to a more homogeneous solid solution. After 8 hours at 1120°C anyhow, some grain-shaped residual austenite is still present; -the analysis of pore morphology puts in evidence a positive effect of sintering time on the circularity index. On the contrary, the pore distribution by area is unaffected by an increase of sintering time. The number of pores is independent of the iteration of the thermal process; -one sintering only, at 1120°C, half-hour, under endogas, already gives mechanical properties good enough for averagely stressed applications. This result fully agrees with the prevailing choices of P/M industries; -hardness and U.T.S. after sintering for one hour at 1250°C permit to assimilate this process to sinter for half-hour, at 1120°C, repeated between 4 and 8 times. The low value of strain at peak load is due to higher material brittleness. In fact the microstructure, besides bainite and very fine pearlite, presents a lot of residual austenite, prevailingly distributed at the borders of previous microconstituents and partially trans-formed. At equal area fraction, this distribution penalizes the whole plastic deformability; -only decreasing the carbon content of surface zones, from 0.45% to 0.2%, caused by a "resintering" at 1120°C, notably improves all properties, which individuate, on the whole, the materials toughness; -after 2 sintering steps at 1120°C, all the properties significantly improve. The sintering repeated more than 2 times generates cost increases not adequately justified - and balanced - by valuable property improvements; The results confirm a statement frequently repeated by some PM experts: "For any material, an optimal degree of sintering exists, but its quantitative definition is arduous". This quantitative definition, from an application point of view, may come from the degree of fulfilment of specifications involving a given thermal process. Heat treatment has been carried out in industrial equipment. The conditions were those commonly used for carbonitriding and quenchinig of PM steels obtained from diffusion-bonded powders (+ 0.2% C). Carbonitriding has been carried out for 2 hours, at 880 °C, (C potential 0.65 ÷ 0.7%), followed by oil quenching from 830 °C. The oil temperature was nearly 60 °C. Stress relieving in air, at 170 °C, for one hour, has been the last step. On the various samples, properly machined, the following tests and investigations have been carried out: dimensions- HV5 hardness on the surface, HV0,05 microhardness, at 0.2 rain from the outer surface and at the core; microhardness profile in the cross section, microstructure investigation (light microscopy, SEM, EDS microanalysis)- strength on curved specimens, by TPB (three-point bending)- strain versus load diagram; fractography of failure surfaces, by SEM; estimate of distribution between brittle and tough areas; preliminary investigation on fracture toughness. Hardness increases when the number of sintering steps increases. The increase is strong passing from 1 to 2 steps, and then becomes more gradual. The hardness of the material previously sintered for one hour at 1250 °C is comparable to that of the material sintered for 4 hours al 1120 °C. Microhardness distributions are similar, but the weights of quantiles at the extremities fade when the number of sintering steps or the temperature increase. The deviations of the extreme deciles of microhardness are high at the core and modest at 0,2 mm from the surface. This difference can be attributed to the homogenizing action of carbon. Core measurements never evidenced the high peaks, again presumably, as a consequence of relatively lower carbon content. On samples sintered twice, the microhardness distribution exhibits the lowest drops near the surface. On all materials, C-enrichment was observed even at the core. The material sintered for one hour reaches the highest microhardness on the surface. Between the material sintered for 8 hours at 1120 °C and that sintered for 1 hour at 1250 °C, the differences are very modest. The 1250 material, heat-treated, shows a core microhardness intermediate between those of the materials sintered at 1120°C for half hour and for one hour. The high sintering temperature causes a decrease of microhardness at the surface and an increase at the core. On all the pictures showing microstructures, some capital letters designate the micro-constituents: AR Residual Austenite-AT = Transforming Austenite; B = Bainite; M = Martensite, PF = Fine pearlite. For short, only the results of observations made after heat treatment on S1, S2, S16 and 1250, are reported. The microstructures of the other materials present intermediate properties between those hereinafter described. S1 Material. Martensite prevails, followed by upper bainite (with different morphologies), transforming austenite (shortly, MA), and residual austenite. No clear differences appear between periphery and core. This points that C-enrichment reached the part core. The more heavily etched areas, dark under the light optical microscope, (LOM), are coarse martensite and bainite. The less etched areas, white at the LOM, at the SEM appear to be MA and fine martensite. Some residual austenite can be detected only on matrix areas surrounding a pore. Microhardness confirms the phase distribution. The microstructure inhomogeneity points the in-homogeneity of the austenitic solid solution. It is confirmed by EDS analyses. In the zones formed by residual austenite and MA the Ni content is 32% and I 1% respectively, while Cu spans from 5 to 3%. The lowest values of Ni and Cu < 1%) are detected on bainitic areas, immersed in the martensitic matrix, faraway from pores. The concentration profiles of Ni and Cu, when passing from an austenite region, near a pore, across MA, to a martensite-bainite zone, are also shown. S2 Material. The extension of less-etched areas, white at LOM, partially transformed, is notably less than S1 material. Microstructure is constituted by coarse martensite, upper bainite, very fine pearlite, partially transformed austenite and fine martensite, near a pore. The bainitic islands are rarely great, contrary to S1 material. More frequently, the matrix between pores is martensite, with rare and small islands of "stringed" bainite, The structure is less inhomogeneous than that of S1, thanks to a better diffusion of alloying elements into the austenitic matrix. EDS analyses confirm an improved homogeneity of the solid solution and a strong influence of a second sintering. After one hour at 1120 °C, the highest Ni amount in martensite areas near a pore, and in the rare zones of partially transformed austenite (that can be detected inside bainitic aggregates) is 3 ÷ 4%. Ni is practically absent in the bainitic areas. In the same zones, Cu ranges from a 2 ÷ 3% peak, on the border of a pore, to values tending to zero, inside the central bainitic regions.