Patent Grant
7553382
The present invention relates to metallic glasses and more particularly to iron based alloys and iron based Cr—Mo—W containing glasses and more particularly to the addition of Niobium to these alloys.
Conventional steel technology is based on manipulating a solid-state transformation called a eutectoid transformation. In this process, steel alloys are heated into a single phase region (austenite) and then cooled or quenched at various cooling rates to form multiphase structures (i.e. ferrite and cementite). Depending on how the steel is cooled, a wide variety of microstructures (ie. pearlite, bainite and martensite) can be obtained with a wide range of properties.
Another approach to steel technology is called glass devitrification, producing steels with bulk nanoscale microstructures. The supersaturated solid solution precursor material is a super cooled liquid, called a metallic glass. Upon superheating, the metallic glass precursor transforms into multiple solid phases through devitrification. The devitrified steels form specific characteristic nanoscale microstructures, analogous to those formed in conventional steel technology.
It has been known for at least 30 years, since the discovery of metallic glasses that iron based alloys could be made to be metallic glasses. However, with few exceptions, these iron based glassy alloys have had very poor glass forming ability and the amorphous state could only be produced at very high cooling rates (>106 K/s). Thus, these alloys can only be processed by techniques which give very rapid cooling such as drop impact or melt-spinning techniques.
While conventional steels have critical cooling rates for forming metallic glasses in the range of 109 K/s, special iron based metallic glass forming alloys have been developed having a critical cooling rate orders of magnitude lower than conventional steels. Some special alloys have been developed that may produce metallic glasses at cooling rates in the range of 104 to 105 K/s. Furthermore, some bulk glass forming alloys have critical cooling rates in the range of 100 to 102 K/s, however these alloys generally may employ rare or toxic alloying elements to increase glass forming ability, such as the addition of beryllium, which is highly toxic, or gallium, which is expensive. The development of glass forming alloys which are low cost and environmentally friendly has proven much more difficult.
In addition to the difficultly in developing cost effective and environmentally friendly alloys, the very high cooling rate required to produce metallic glass has limited the manufacturing techniques that are available for producing articles from metallic glass. The limited manufacturing techniques available have in turn limited the products that may be formed from metal glasses, and the applications in which metal glasses may be used. Conventional techniques for processing steels from a molten state generally provide cooling rates on the order of 10−2 to 100 K/s. Special alloys that are more susceptible to forming metallic glasses, i.e., having reduced critical cooling rates on the order of 104 to 105 K/s, cannot be processed using conventional techniques with such slow cooling rates and still produce metallic glasses. Even bulk glass forming alloys having critical cooling rates in the range of 100 to 102 K/s, are limited in the available processing techniques, and have the additional processing disadvantage in that they cannot be processed in air but only under very high vacuum.
In a summary exemplary embodiment, the present invention relates to an iron based glass alloy composition comprising about 40-65 atomic % iron; about 5-55 atomic % of at least one metal selected from the group consisting of Ti, Zr, Hf, V, Ta, Cr, Mo, W, Mn, Ni or mixtures thereof, and about 0.01-20 atomic % of Niobium.
In another summary exemplary embodiment, the present invention relates to a method for increasing the hardness of an iron alloy composition comprising supplying an iron based glass alloy having a hardness, adding Niobium to the iron based glass alloy, and increasing the hardness by adding the Niobium to the iron based glass alloy.
In another summary exemplary embodiment, the present invention relates to a method for increasing the glass stabilization of an iron based alloy composition comprising supplying an iron based glass alloy having a crystallization temperature of less than 675° C., adding Niobium to the iron based glass alloy, and increasing the crystallization temperature above 675° C. by adding Niobium to the iron based glass alloy.
The present invention relates to the addition of niobium to iron based glass forming alloys and iron based Cr—Mo—W containing glasses. More particularly, the present invention is related to changing the nature of crystallization resulting in glass formation that may remain stable at much higher temperatures, increase glass forming ability and increase devitrified hardness of the nanocomposite structure. Additionally, without being bound to any particular theory, it is believed that the supersaturation effect from the niobium addition, may result in the ejection of the niobium from the solidifying solid which may additionally slow down crystallization, possibly resulting in reduced as-crystallized grain/phase sizes.
The present invention ultimately is an alloy design approach that may be utilized to modify and improve existing iron based glass alloys and their resulting properties and may preferably be related to three distinct properties. First, the present invention may be related to changing the nature of crystallization, allowing multiple crystallization events and glass formation which may remain stable at much higher temperatures. Second, the present invention may allow an increase in the glass forming ability. Third, consistent with the present invention, the niobium addition may allow an increase in devitrified hardness of the nanocomposite structure. These effects may not only occur in the alloy design stage but may also occur in industrial gas atomization processing of feedstock and in PTAW welding of hardfacing weld overlays.
Furthermore, the improvements may generally be applicable to a range of industrial processing methods including PTAW, welding, spray forming, MIG (GMAW) welding, laser welding, sand and investment casting and metallic sheet forming by various continuous casting techniques.
A consideration in developing nanocrystalline or even amorphous welds, is the development of alloys with low critical cooling rates for metallic glass formation in a range where the average cooling rate occurs during solidification. This may allow high undercooling to occur during solidification, which may result either in the prevention of nucleation resulting in glass formation or in nucleation being prevented so that it occurs at low temperatures where the driving force of crystallization is very high and the diffusivities are minimal. Undercooling during solidification may also result in very high nucleation frequencies with limited time for growth resulting in the achievement of nanocrystalline scaled microstructures in one step during solidification.
In developing advanced nanostructure welds, the nanocrystalline grain size is preferably maintained in the as-welded condition by preventing or minimizing grain growth. Also preferably, is the reduction of the as-crystallization grain size by slowing down the crystallization growth front which can be achieved by alloying with elements which have high solubility in the liquid/glass but limited solubility in the solid. Thus, during crystallization, the supersaturated state of the alloying elements may result in an ejection of solute in front of the growing crystallization front which may result in a dramatic refinement of the as-crystallized/as solidified phase size. This can be done in multiple stages to slow down growth throughout the solidification regime.
Consistent with the present invention, the nanocrystalline materials may be iron based glass forming alloys, and iron based Cr—Mo—W containing glasses. It will be appreciated that the present invention may suitably employ other alloys based on iron, or other metals, that are susceptible to forming metallic glass materials. Accordingly, an exemplary alloy may include a steel composition, comprising at least 40 at % iron and at least one element selected from the group consisting of Ti, Zr, Hf, V, Nb, Ta, Cr, Mo, W, Al, Mn, or Ni; and at least one element selected from the group consisting of B, C, N, O, P, Si and S.
Niobium may be added to these iron based alloys between 0.01-25 at % relative to the alloys and all incremental values in between, i.e. 0.01-15 at %, 1-10 at %, 5-8 at %, etc. More preferably, the niobium present in the alloy is 0.01-6 at % relative to the alloys.
Two metal alloys consistent with the present invention were prepared by making additions of Nb at a content of 0.01-6 at % relative to the two different alloys, Alloy 1 and Alloy 2. C and Ni were also included in some of the Nb modified alloys. The composition of these alloys is given in Table 1, below.
The densities of the alloys are listed in Table 2 and were measured using the Archimedes method. A person of ordinary skill in the art would recognize that the Archimedes method utilizes the principal that the apparent weight of an object immersed in a liquid decreases by an amount equal to the weight of the volume of the liquid that it displaces.
Each alloy described in Table 1 was melt-spun at wheel tangential velocities equivalent to 15 m/s and 5 m/s. For each sample of melt-spun ribbon material for each alloy, differential thermal analysis (DTA) and differential scanning calorimetry (DSC) was performed at heating rates of 10° C./minute. A person of ordinary skill in the art would recognize DTA involves measuring the temperature difference that develops between a sample and an inert reference material while both sample and reference are subjected to the same temperature profile. A person of ordinary skill in the art would recognize DSC as a method of measuring the difference in the amount of energy necessary to heat a sample and a reference at the same rate. In Table 3, the onset and peak temperatures are listed for each crystallization exotherm.
With respect to Alloy 1, as can be seen from Table 3, the addition of the Nb causes glass devitrification in three or four stages, evidenced by the multiple crystallization events. The stability of the first crystallization event increases as well, except for the Nb/Ni modified alloys. Furthermore, multiple glass crystallization peaks are observed in all cases where Nb has been added to Alloy 1.
With respect to Alloy 2, an increase in glass stability with multiple crystallization events is observed with the addition of Nb, except for the Nb2 modified alloy at a quench rate of 5 m/s. At quench rates of 15 m/s, the alloys demonstrate three crystallization events. Furthermore, the crystallization temperature increases with the addition of Nb.
All alloy compositions were melt-spun at 15 m/s and 5 m/s and the crystallization enthalpy was measured using differential scanning calorimetry. In Table 4, the total crystallization enthalpy is shown for each alloy melt-spun at 15 m/s and 5 m/s. Assuming that the 15m/s samples are 100% glass, the percent glass found in the lower cooling rate corresponding to quenching at 5 m/s can be found by taking the ratio of crystallization enthalpies, shown in Table 4.
With respect to Alloy 1, the base alloy (Alloy 1) was found to not form a glass when processed at low cooling rates equivalent to melt-spinning at a tangential velocity of 5 m/s. However, it was found that the niobium addition greatly enhances glass forming ability in all of the modified alloys, with the exception of the Nb4C3 modified Alloy. In the best case, Nb4 Modified Alloy 1, it was found that 99.3% glass formed when processed at 5 m/s.
Similarly, in Alloy 2, the alloy was found not to form a glass when processed at low cooling rates equivalent melt-spinning at a tangential velocity of 5 m/s. However, it was found that the glass forming ability was enhanced with the niobium addition. In the best case of Nb4 Modified Alloy 2, the amount of glass at 5 m/s was found to be 82.5%.
The melting events for each alloy composition melt-spun at 15 m/s are shown in Table 5. The melting peaks represent the solidus curves since they were measured upon heating so the liquidus or final melting temperatures would be slightly higher. However, the melting peaks demonstrate how the melting temperature will vary as a function of alloy addition. The highest temperature melting peak for Alloy 1 is found to be 1164° C. The addition of niobium was found to raise the melting temperature but the change was slight, with the maximum observed at 43° C. for Nb4 Modified Alloy 1. The upper melting peak for Alloy 2 was found to be 1232° C. Generally, the addition of niobium to this alloy did not cause a significant change in melting point since all of the alloys peak melting temperatures were within 6° C.
The hardness of the Alloy 1 and 2 and the Nb modified alloys was measured on samples heat treated at 750° C. for 10 minutes and the results are given in Table 6. Hardness was measured using Vickers Hardness Testing at an applied load of 100 kg following the ASTM E384-99 standard test protocols. A person of ordinary skill in the art would recognize that in the Vickers Hardness Test, a small pyramidal diamond is pressed into the metal being tested. The Vickers Hardness number is the ratio of the load applied to the surface area of the indentation. As can be seen, all of the alloys exhibited a hardness at HV100 over 1500 kg/mm2. As shown, the hardness of Alloy 1 was found to be 1650 kg/mm2 and in all of the niobium alloys the effect of the niobium was to increase hardness, except for Nb2Ni4 Modified Alloy 1. The highest hardness was found in Nb2C3 Modified Alloy 1 and was 1912 kg/mm2. This reportedly may be the highest hardness ever found in any iron based glass nanocomposite material. The lower hardness found in Nb2Ni4 Modified Alloy 1 is believed to be offset by the nickel addition which lowered hardness.
For Alloy 2, a reduced change in hardness was observed as a result of the niobium addition. This may be due to the near perfect nanostructures which are easily obtainable by the high cooling rates in melt-spinning of Alloy 2. It is believed that for weld alloys that the niobium addition may result in high hardness because it may assist in obtaining a fine structure according to the increase in glass forming ability, glass stability, and the inhibition of grain growth by multiple crystallization paths. A case example is also shown in Case Example 3.
The yield strength of the devitrified structures can be calculated using the relationship: yield stress (σy)=⅓VH (Vickers Hardness). The resulting estimates were between 5.2 to 6.3 GPa.
To produce feed stock powder for plasma transfer arc welding (PTAW) trials, Alloy 1, Nb2Ni4 Modified Alloy 1 and Nb2 Modified Alloy 1 were atomized using inter gas atomization system in argon. The as-atomized powder was sieved to yield a cut which was either +50 μm to −150 μm or +75 μm to −150 μm, depending on the flowability of the powder. Differential thermal analysis was performed on each gas atomized alloy and compared to the results of melt-spinning for the alloys, illustrated in
Plasma Transferred Arc Welding (PTAW) trials were done using a Stellite Coatings Starweld PTAW system with a Model 600 torch with an integrated side-beam travel carriage. Plasma transferred arc welding would be recognized by a person of ordinary skill in the art as heating a gas to an extremely high temperature and ionizing the gas so that it becomes electrically conductive. The plasma transfers the electrical arc to the workpiece, melting the metal.
All welding was in the automatic mode using transverse oscillation and a turntable was used to produce the motion for the circular bead-on-plate tests. For all weld trials done the shielding gas that was used was argon. Transverse oscillation was used to produce a bead with a nominal width of ¾ inches and dwell was used at the edges to produce a more uniform contour. Single pass welds were made onto 6 inch by 3 inch by 1 inch bars with a 600° F. preheat as shown for the Alloy 1 PTA weld in
Hardness measurements using Rockwell were made on the ground external surface of the linear crack specimens. Since Rockwell C measurements are representative of macrohardness measurements, one may take these measurements on the external surface of the weld. Additionally Vickers hardness measurements were taken on the cross section of the welds and tabulated in the Fracture Toughness Measurements Section. Since Vickers hardness measurements are microhardness one may make the measurements on the cross section of the welds which gives the additional benefit of being able to measure the hardness from the outside surface to the dilution layer in the weld. In Table 7, the welding parameters for each sample, bead height and Rockwell hardness results are shown for the linear bead hardness test PTAW specimens.
Backscattered electron micro graphs were taken of the cross section of Alloy 1, Nb2Ni4 Modified Alloy 1 and Nb2 Modified Alloy 1, illustrated in
Nine, one hour X-ray diffraction scans of the PTAW samples were performed. The scans were performed using filtered Cu Kα radiation and incorporating silicon as a standard. The diffraction patterns were then analyzed in detail using Rietvedlt refinement of the experimental patterns. The identified phases, structures and lattice parameters Alloy 1, Nb2Ni4 Modified Alloy 1 and Nb2 Modified Alloy 1 are shown in Tables 8, 9, and 10 respectively.
Noted from the results of the x-ray diffraction data, is that niobium addition caused face centered cubic iron (i.e. austenite) to form along with the α-Fe which was found in Alloy 1. For all the samples, the main carbide phase present is a M7C3 while the main boride phase in all of the PTAW samples has been identified as a M23B6. Furthermore, limited EDS (Energy Dispersive X-Ray Spectroscopy) analysis demonstrated the carbide phase contains a considerable amount of boron and that the boride phase contains a considerable amount of carbon. Thus, all of these phases can also be considered as borocarbides. Also, note that while similar phases are found in a number of these PTAW weld alloys, the lattice parameters of the phases change as a function of alloy and weld conditions, Table 7, indicating the redistribution of alloying elements dissolved in the phases. The iron based PTAW microstructures can be generally characterized as a continuous matrix comprised of ductile α-Fe and/or γ-Fe dendrites or eutectoid laths intermixed with hard ceramic boride and carbide phases.
The fracture toughness was measured using the Palmqvist method. A person of ordinary skill in the art would recognize that the Palmqvist method involves the application of a known load to a Vickers diamond pyramid indenter that results in an impacted indentation into the surface of the specimen. The applied load must be greater than a critical threshold load in order to cause cracks in the surface at or near the corners of the indentation. It is understood that cracks are nucleated and propagated by unloading the residual stresses generated by the indentation process. The method is applicable at a range at which a linear relationship between the total crack length and the load is characterized.
The fracture toughness may be calculated using Shetty's equation, as seen in Equation 1.
Wherein ν is Poisson's ratio, taken to 0.29 for Fe, ψ is the half-angle of the indenter, in this case 68°, H is the hardness, P is the load and 4a is the total linear crack length. The average of five measurements of microharness data along the thickness of the weld was used to determine the fracture toughness reported. The crack resistance parameter, W, is the inverse slope of the linear relation between crack length and load and is represented by P/4a.
Two crack length measuring conventions were chosen for evaluation. The first convention is designated as Crack Length (CL) and is the segmented length of the actual crack including curves and wiggles beginning from the indentation edge to the crack tip. The second convention is called the Linear Length (LL) and is the length of the crack from its root at the indentation boundary to the crack tip. Initial indentations were made with nominal 50 kg and 100 kg loads and based on the appearance of these indentations, a range of loads was selected.
The crack lengths for the two conventions were measured by importing the digital micrographs into a graphics program that used the bar scale of the image to calibrate the distances between pixels so that the crack lengths could accurately be measured. A spread sheet design was used to reduce the data for computing the fracture toughness. This data was plotted and a linear least squares fit was computed in order to determine the slope and the corresponding R2 value for each crack length convention and is shown in Table 11. This data, along with the hardness data, was inputted into Shetty's equation and the fracture toughness was computed and the results are shown in Table 12. It can be seen that Alloy 1 when PTA welded resulted in toughness values that were moderate. With the addition of niobium in the modified alloys, vast improvements in toughness were found in the Nb2Ni4 Modified Alloy 1 and the Nb2 Modified Alloy 1.
While not limiting the scope of this application, it is believed that the improvements in toughness found in the niobium alloys may be related to microstructural improvements which are consistent with the Crack Bridging model to describe toughness in hardfacing alloys. In Crack Bridging, the brittle matrix may be toughened through the incorporation of ductile phases which stretch, neck, and plastically deform in the presence of a propagating crack tip. Crack bridging toughening (ΔKcb) has been quantified in hardfacing materials according to the following relation; ΔKcb=Ed[χVf(σ0/Ed)a0]1/2 where Ed is the modulus of the ductile phase, χ is the work of rupture for the ductile phase, σ0 is the yield strength of the ductile phase, a0 is the radius of the ductile phase, and Vf is the volume fraction of ductile phase.
The reduction in microstructural scale as shown by the Hall-Petch relationship (σy≈kd1/2) and the increase in microhardness found from the niobium addition, is consistent with increasing yield strength. Increasing yield strength, increases the work of rupture resulting in the observed toughness increase. Increasing amounts of transition metals like niobium dissolved in the dendrite/cells would increase the modulus, thus increasing the toughness according to the Crack Bridging Model. Finally, the uniform distribution of fine (0.5 to 1 micron) M23(BC)6 and M7(BC)3 ceramic precipitates surrounded by a uniform distribution of ductile micron sized γ-Fe and α-Fe dendrites or eutectoid laths of is also expected to be especially potent for Crack Bridging.
A study was launched to verify the improvement in hardness in weld/ingot samples by adding niobium to Alloy 2. The alloy identified as Nb6 modified Alloy 2 in Table 1 was made into a 12 lb charge using commercial purity feedstock. This alloy was then atomized into powder by a close coupled inert gas atomization system using argon as the atomization gas. The resulting powder was then screened to yield a PTA weldable product which was nominally +53 to −150 μm in size. To mimic the PTA process, a 15 gram ingot of powder was arc-welded into an ingot. The hardness of the ingot was then measured using Vickers at the 300 gram load. As shown, in Table 13, the hardness of the arc-welded sample ingot was very high at 1179 kg/mm2 (11.56 GPa). Note that this hardness level corresponds to a hardness greater than the Rockwell C scale (i.e. Rc>68). Also, note that this hardness is greater than that achieved in Table 7 and that shown in
| Number | Name | Date | Kind |
|---|---|---|---|
| 4297135 | Giessen et al. | Oct 1981 | A |
| 4365994 | Ray | Dec 1982 | A |
| 4495691 | Masumoto et al. | Jan 1985 | A |
| 4842657 | Masumoto et al. | Jun 1989 | A |
| 6258185 | Branagan et al. | Jul 2001 | B1 |
| 6350323 | Inoue et al. | Feb 2002 | B1 |
| 6689234 | Branagan et al. | Feb 2004 | B2 |
| 20040250929 | Branagan | Dec 2004 | A1 |
| 20080053274 | Branagan et al. | Mar 2008 | A1 |
| Number | Date | Country | |
|---|---|---|---|
| 20060180252 A1 | Aug 2006 | US |
