AN INVESTIGATION INTO FULLY MELTING A MARAGING STEEL USING DIRECT METAL LASER SINTERING (DMLS) Mark. Stanford, Kevin. Kibble, Matthew. Lindop, Diane. Mynors and Colin. Durnall Department of Engineering and Technology University of Wolverhampton Shifnal Road, Priorslee Telford, UK TF2 9NT ABSTRACT This paper reports on the relative capability of the EOS M250 Extended platform to fully melt an 18Ni (‘300’ grade) maraging tool steel from sub-20 µm powder. The work describes the investigation of scanning routines necessary to achieve satisfactory metallurgical integrity and shape manufacturing capabilities of the process.
Solidification was observed to take place by cellular and cellular-dendritic growth mechanisms in DMLS of the maraging steel. Intercellular spacing was less than 1 (m and this contributed to the excellent strength and hardness achieved for both as-sintered and aged material. Aging at 490(C for 6 h led to an increase in hardness and strength through the precipitation of Ni3(Mo,Ti) and Fe2Mo intermetallics. Ultimate tensile strength increased from 1101 to 1875 MPa, Vickers hardness increased from 387 to 603 kg/mm2 but elongation was reduced from 11. 3 to 1. 8% by aging the steel.
These test results were within the tolerance specification for the material. The overall length of the test specimens was measured in the as-sintered and aged conditions. Uniformity was very high before and after the aging treatment, Shrinkage after aging was only 0. 061%. The results confirmed that the maraging steel could be successfully manufactured on the M250 Extended platform. Keywords: Laser Sintering, Maraging Steel, Additive Layer Manufacture 1. INTRODUCTION Direct Metal Laser Sintering (DMLS) was introduced in 1992 by EOS Laser Systems GmbH using its M250 platform and a CO2 laser.
The current M250 Extended version of the technology allows 20 µm layer manufacture and has been used predominantly for prototype mould tooling manufacture in a cupro-nickel alloy . The technology allows for complex tooling to be made with conformal cooling capabilities. Conformal cooling has been proven to reduce lead times and enhance productivity of a number of net-shape manufacturing processes. Distinct drawbacks of the process have been related to the porous nature of the cupro-nickel alloy after sintering and the lack of wear resistance when injecting filled thermoplastic polymers.
There is a need to move from prototype tooling manufacture to full production hard tooling manufacture to fully exploit the benefits of the M250 Extended additive layer manufacturing process. A fully melted durable steel is required that is hardenable and representative of traditional chromium alloyed tool steels. In this paper the capability of the M250 Extended platform to fully melt a Fe-18Ni-9Co-5Mo-Ti-Al (‘300’ grade, EOS MS-1) maraging tool steel, from sub-20 µm powder, is reported.
The work describes the investigation of scanning routines necessary to achieve satisfactory metallurgical integrity, acceptable surface finish and mechanical properties. The nominal composition of the ‘300’ steel is given in Table 1. Table 1. Nominal chemical composition of ‘300’ (EOS MS-1) maraging steel in weight % . |Fe |Ni |Co |Mo |Ti |Al | |Bal. |17-19 |8. 5-9. 5 |4. 5-5. 2 |0. 6-0. 8 |0. 05-0. 15 | |Cr |C |Mn |Si |P |S | |(0. 5 |(0. 03 |(0. 1 |(0. 1 |(0. 1 |(0. 01 | 2. BACKGROUND Rapid tooling (RT) applications using the EOS M250 Extended technology and DirectMetal 20 (EOS DM20) powder offers significant benefits in reducing tooling production lead times. DM20 is a nickel-bronze, which has been adopted successfully for prototype and pre-serial production run tooling used in the manufacture of a range of thermoplastic products. DM20 material still offers one of the fastest build rates, produces a superior surface finish and offers the greatest accuracy with respect to other RT materials currently available.
Recent developments related to the EOS M270 platform have led to the ability to fully melt steel powders. One material developed for the M270 platform is a ‘300’ grade maraging steel (EOS grade MS1). This material is capable of being sintered in layers of 40 µm with relatively low internal stress levels. However, unlike other commercially available steel powders this material can be manipulated in a soft condition as-sintered and machined conventionally. The steel can then be age- hardened to 50-55 Rc, thereby matching tool steels alloyed with relatively high levels of chromium (>5%).
The mechanical properties required of dies for plastics injection moulding and metal die-casting are a high hardness and a high tensile strength. For high volume production runs tool steels are preferred, with a hardness exceeding 500 HV (>50 Rc). If conformal water-cooled channels are also incorporated into the dies then some measure of corrosion resistance will also be required of the tooling. Typical tool steels contain relatively high carbon and high chromium contents.
Although chromium offers some corrosion resistance the presence of a relatively high carbon will result in carbides being present in the microstructure that can promote corrosion. Fe-Cr based tool steels are austenitised, quenched and tempered to produce a tempered martensite matrix containing Fe-Cr carbides. In quenching these alloys significant distortion can occur, as there is a volume expansion associated with the transformation to high carbon martensite. The transformation can even lead to quench cracking which may only become evident during subsequent service and result in premature tool failure.
Maraging steels offer an attractive alternative to the medium to high carbon chromium tool steels as they do not suffer the same distortion problems as these steels. This is because they have very low carbon and high nickel contents and thus are not susceptible to quench cracking. The high nickel content (and cobalt, if present) and absence of carbides also provides good corrosion resistance. Maraging steels have a high tensile strength, >1800 MPa, and a relatively high hardness, >50 Rc, through a combination of precipitation hardening and for the 300 grade a high nickel-cobalt martensite matrix.
Aging above 450(C produces rapid hardening due to the precipitation of Ni3(Mo,Ti) and Fe2Mo phases . Aging between 500(C and the austenite start temperature (As) produces austenite precipitation by a diffusion controlled reaction . Aging above 500(C leads to overaging, dissolution of Ni3(Mo,Ti) and austenite reversion. The major benefit of using this layer based sintered material is that complex conformal cooling channels can be incorporated throughout the tool section to offer downstream moulding and casting operational time reduction.
To date this maraging steel has not been ‘sintered’ using an EOS M250 platform and with far more of these platforms available across Europe there will be an industry wide benefit and increase in capacity for the production of maraging steels if these machines are able to manufacture tooling in these materials. The work carried out, and reported here, involved the development of a set of manufacturing criteria that enabled the EOS MS1 steel to be succesfully produced on an EOS M250 Ext DMLS machine. 3. EXPERIMENTAL An EOS M250 Ext DMLS machine, with machine control software PSW 3. 2, was used to build the specimens.
This version of the software is also used with the EOS M270 platform and allowed the same rotating raster scan exposure methodology to be used, for the build, as that proven to work on the M270. As there was no published data regarding the DMLS of M300 maraging steel (EOS Grade MS1) with a 200 W CO2 laser, the scan parameters and stepover distance for DirectSteel H20 (EOS DSH20) were used as a starting point. H20 is a well established material for building parts on the M250 platform. Processing parameters used were an infill scan speed of 130 mm/s at a laser power of 195 W with a step-over distance of 200 (m.
A post-contour scan was used with a speed of 250 mm/s at a laser power of 195 W. The building process exposed the build in 40 (m layers. Initially the primary criterion considered during the build process was that of the surface finish obtained from the laser sintered upper surface. Parameters were changed during the build until the build offered the least resistance to recoating. A build of test specimens, 30 mm high, 5 mm thick and 150 mm long was made to evaluate material properties and dimensional accuracy. Specimens for metallographic examination were mounted in conducting bakelite and polished using standard techniques.
Specimens were etched in Marble's Reagent. Scanning electron microscopy (SEM) was used to evaluate microstructure and the fracture surfaces of the tensile specimens. The SEM was a Zeiss EVO50 EVP, operating at 20 keV, coupled with an Oxford Instruments LINK energy dispersive spectrometer (EDS). Hardness tests (Vickers, 20 kg load) and tensile tests were carried out on samples that were as-sintered and had been aged-hardened for 6 h at 490(C followed by air cooling. The tensile tests were conducted on a Zwick-Roell 1474 machine, according to European Standard EN 10002-1:2001 Metallic materials-Tensile testing, Part 1. . RESULTS AND DISCUSSION The mechanical properties are given in Tables 2 and 3, for as-sintered and aged test specimens. Table 2. Mechanical properties of ‘300’ (MS1) maraging steel as-sintered. |n |E |Rp0. 2 |Rm |( Break |HV20 | |5 |GN/m? |N/mm? |N/mm? |% |kg/mm? | |x |170 |981 |1101 |11. 3 |387 | |s |4. 89 |5. 66 |6. 12 |0. 626 |5. 98 | |( |2. 88 |0. 58 |0. 56 |5. 52 |1. 55 | Table 3. Mechanical properties of ‘300’ (MS1) maraging steel aged 490(C/6h. n |E |Rp0. 2 |Rm |( Break |HV20 | |5 |GN/m? |N/mm? |N/mm? |% |kg/mm? | |x |181 |1794 |1875 |1. 8 |603 | |s |3. 70 |16. 24 |9. 88 |0. 498 |8. 14 | |( |2. 04 |0. 91 |0. 53 |27. 96 |1. 35 | E = Youngs Modulus Rp0. 2 = 0. 2% proof stress Rm = Ultimate tensile strength ( Break = Elongation HV20 = Vickers Hardness x = Mean s = Standard deviation ( = Coefficient of variation in %. Five specimens were tested for each condition.
Aging at 490(C for 6 h led to an increase in hardness and strength through the precipitation of Ni3(Mo,Ti) and Fe2Mo intermetallics. Ultimate tensile strength increased from 1101 to 1875 MPa, Vickers hardness increased from 387 (39-40 Rc) to 603 (55-56 Rc) kg/mm2 but elongation was reduced from 11. 3 to 1. 8% by aging the steel. The test results for both as-sintered and aged conditions were within the tolerance specification given by EOS for the manufacture of MS1 steel on the M270 platform and confirmed that the material could be successfully manufactured on the CO2 laser based M250 platform .
The test specimens also showed very little variation, as the coefficient of variation, (, was low for all measured mechanical test parameters except for the % elongation in the aged specimens, however, even here the % elongation is within specification . The overall length of the test specimens was measured in the as-sintered and aged conditions. The results are given in Table 4 where it can be seen that the uniformity is very high before and after the aging treatment, e. g. the coefficient of variation, (, is 0. 030 and 0. 019% respectively. Shrinkage after aging was only 0. 061%. Table 4.
Dimensional uniformity on the overall specimen length of ‘300’ (MS-1) maraging steel (n = 5). | |As-sintered |Aged 490(C/6h | |x, mm |190. 448 |190. 332* | |s |0. 058 |0. 036 | |( |0. 030 |0. 019 | *Shrinkage = 0. 061% x = Mean s = Standard deviation ( = Coefficient of variation in % The microstructural evaluation for the as-sintered material revealed fine precipitates in the martensite matrix, around 20-100 nm in size, see Figure 1.
This along with the Vickers hardness, of 387 kg/mm2, proves that some aging took place in the build process. This ‘natural’ aging also explains why a solution heat treatment, prior to aging, was not necessary for the achievement of a tensile strength of at least 1800 MPa. Several inclusions, some exceeding 100 (m in length, were observed, see Figures 2 and 3. EDS confirmed that these were TiO2:Al2O3 combined oxides. The long axis of the inclusion, shown in Figure 2, is aligned in the direction of the laser scan and at the layer build interface. pic] Figure 1. As-sintered. Fine precipitates can be seen in the martensite matrix, around 20-100 nm in size. [pic] Figure 2. As-sintered. The large inclusion, was identified by EDS as TiO2:Al2O3. To the mid-left of the inclusion it can be seen that the matrix has a cellular appearance. [pic] Figure 3. As-sintered. The inclusion was identified by EDS as TiO2:Al2O3. The view is from a section perpendicular to the layer build sequence. Figure 3 shows a view taken from a section perpendicular to the layer build sequence.
This image coupled with that from Figure 2 shows that the inclusions are in the form of plates and are deposited at interfacial build layers. In Figure 2, there are regions of the matrix where cellular growth can be seen. The microstructure, after aging, was similar to that as-sintered, see Figures 4 and 5. The increase in hardness and strength with aging, refer to Tables 2 and 3, can thus be attributed to an increase in precipitation density and an increase in martensite volume. The ‘new’ precipitates will be of a size difficult to resolve in the SEM, i. e. probably