Advanced Search
RSS LinkedIn Twitter

Journal Archive

Platinum Metals Rev., 2003, 47, (2), 46

Optimised Mechanical Properties of Ordered Noble Metal Alloys

I. Preliminary Strong Cold Deformation Results in Optimised Mechanical Properties

  • By B. A. Greenberg a
  • N. A. Kruglikov a
  • L. A. Rodionova a
  • A. Yu. Volkov a
  • L. G. Grokhovskaya b
  • G. M. Gushchin b
  • I. N. Sakhanskaya b
  • a
    Institute of Metal Physics, Ural Branch, Russian Academy of Sciences, 18 S. Kovalevskaya St., 620219 Ekaterinburg, Russia
  • b
    Ekaterinburg Nonferrous Metals Processing Plant, 8 Lenin St., 620014 Ekaterinburg, Russia
SHARE THIS PAGE:

Article Synopsis

The deformation behaviour of alloys that are ordered to form an L10 or L12 superstructure after heavy plastic deformation has been studied. Alloys with the LI0 superstructure, such as NiPt, FePd, CoPt and CuAu, possess an optimum combination of high strength and plasticity after thermomechanical treatment. However, such properties were not obtained in Pd3Fe, Pt3Co or Cu3Au which have the L12 superstructure. The structure of FePd was examined by transmission electron microscopy upon annealing after heavy drawing and a set of typical superstructural states was found. The conditions needed to impart high strength and plastic properties to some ordered alloys based on noble metals are discussed.

Ordered alloys based on the platinum group metals are used in contact, resistive, hard magnetic and spring materials. Their stable structures and properties impart high reliability to contact and elastic materials, and beneficial use is made of them in critical units and devices. These alloys also have a high corrosion resistance which they retain under harsh operating conditions. Introducing a heavy cold deformation into their treatment cycle improves mechanical properties and extends their fields of applications. However, it affects the relationship between the dislocation structure produced in a disordered alloy by cold deformation and the subsequent ordering process. Cahn (1, 2) developed fundamental concepts for describing the effect of a heavy preliminary deformation on ordering processes. These include:

  • whether or not plastic deformation decelerates subsequent ordering;

  • how ordering affects recrystallisation;

  • what are the conditions under which one of these processes leads the other; and

  • how the properties of the alloys change in this case, etc.

In earlier work we have shown that the rearrangement of a dislocation structure upon annealing largely depends on the ratio between the rate of the ordering process and the rate of the processes that determine the rearrangement of the dislocation (3, 4). Two limiting cases are possible:

  • if ordering is slow, the dislocations in the disordered phase have time to rearrange themselves;

  • if ordering is fast, the dislocation structure inherited from the disordered phase is built into the ordered matrix and forms a kind of framework.

In the former case, the dislocation redistribution is determined by properties of the disordered phase and does not require any special analysis. The latter case — rapid ordering and deceleration of recrystallisation — may be illustrated for an L12 superstructure by the Cu3Au alloy, which has a critical temperature of ordering, Tc = 390°C. Recrystallisation was not initially detected in Cu3Au at T < Tc, but this alloy did recrystallise in the ordered state upon long-term annealing 5. The recrystallisation lasted for 106–107 seconds at 330–380°C and for 103–104 seconds at a temperature slightly above Tc 5. Clearly, redistribution of the dislocations in the disordered phase at T < Tc would take at least 103–104 seconds. However, a high degree of long-range order in Cu3Au at 330–380°C is achieved much more rapidly — in about 102 seconds 6. Over this temperature interval the dislocations have no time to be redistributed, that is, fast ordering occurs. A similar slow down in recrystallisation upon fast ordering has also been observed in the Co3Ti alloy (7)

Unlike the L12 superstructure, the L10 superstructure recrystallises in the ordered state. The recrystallisation of a heavily deformed CuAu alloy at T < Tc was observed in 1. However, it was later found that for some alloys with the L10 superstructure, such as NiPt, FePd and CoPt, the temperature dependence of the recrystallisation rate is anomalous, that is, it decreases as the temperature approaches Tc.

In this study we aim to understand specific features of the microstructure and deformation behaviour of alloys which are ordered after a strong cold deformation, and to design a material that has both high strength and highly plastic properties. All the alloys used in mechanical testing were in wire form. Tensile test samples were 50 mm long and 0.5 mm in diameter. They were all cold deformed by drawing (c.d.); for comparison some were also quenched. Plasticity values, δ, refer to the value of the stretch before failure. The alloys studied were melted, cold deformed and mechanically tested at the Ekaterinburg Nonferrous Metals Processing Plant. The microstructures were investigated at the Institute of Metal Physics.

Mechanical Properties of Ordered Alloys after Preliminary Plastic Deformation

Ordered Alloys with L10 Superstructure

Figure 1 shows deformation curves for alloys: FePd (Tc = 650°C); CoPt (Tc = 805°C); NiPt (Tc = 645°C); and CuAu (Tc = 410°C) all with the L10 superstructure. All Curves 1 are of disordered alloys after heavy drawing, with minimum elongation before failure, that is, minimum plasticity, δ, occurs. These deformation curves are the shortest. If annealing and subsequent quenching are performed at T > Tc, the disordered alloys rapidly recrystallise. As a result (Curves 8) the yield stress., σ 0.2, decreases while the plasticity, 8, considerably increases. These deformation curves show the lowest yield stress.

Fig. 1

Deformation curves (stress/elongation) for alloys with L10 superstructure:

  • FePd: (1) after cold deformation by drawing (c.d.) to 90%; (2) same + annealing at 600°C for 0.25 h; (3) for 1 h; (4) for 100 h; (5) same but annealing at 500°C for 0.25 h; (6) for 20 h; (7) for 100 h; (8) quenching from 800°C (I h)

  • CoPt: (1) after c.d. to 96%; (2) same + annealing at 800°C for 0.5 h; (3) for 5 h; (4) for 24 h; (5) same but annealing at 650°C for 0.5 h; (6) for 5 h; (7) for 24 h; (8) quenching from 1000°C (0.5 h)

  • NiPt: (I) after c.d. to 75%; (2) same + annealing at 600°C for 3 h; (3) for 24 h; (4) for 80 h; (5) same but annealing at 550°C for 3 h; (6) for 12 h; (7) for 80 h; (8) quenching from 800”C (1 h)

  • CuAu: (I) after c.d. to 75%; (2) same + annealing at 370°C for 0.5 h; (3) same but annealing at 300°Cfor 0.25 h;(4) for 0.5 h; (5) same but annealing at 450°C for 1 h, quenched, then annealed at 300°C, 0.5 h; (6) same but final anneal at 250°C for 0.5 h; (7) same but final anneal at 370°C for 0.5 h; (8) quenching from 450°C (1 h)

Deformation curves (stress/elongation) for alloys with L10 superstructure: FePd: (1) after cold deformation by drawing (c.d.) to 90%; (2) same + annealing at 600°C for 0.25 h; (3) for 1 h; (4) for 100 h; (5) same but annealing at 500°C for 0.25 h; (6) for 20 h; (7) for 100 h; (8) quenching from 800°C (I h)CoPt: (1) after c.d. to 96%; (2) same + annealing at 800°C for 0.5 h; (3) for 5 h; (4) for 24 h; (5) same but annealing at 650°C for 0.5 h; (6) for 5 h; (7) for 24 h; (8) quenching from 1000°C (0.5 h)NiPt: (I) after c.d. to 75%; (2) same + annealing at 600°C for 3 h; (3) for 24 h; (4) for 80 h; (5) same but annealing at 550°C for 3 h; (6) for 12 h; (7) for 80 h; (8) quenching from 800”C (1 h)CuAu: (I) after c.d. to 75%; (2) same + annealing at 370°C for 0.5 h; (3) same but annealing at 300°Cfor 0.25 h;(4) for 0.5 h; (5) same but annealing at 450°C for 1 h, quenched, then annealed at 300°C, 0.5 h; (6) same but final anneal at 250°C for 0.5 h; (7) same but final anneal at 370°C for 0.5 h; (8) quenching from 450°C (1 h)

Optimum Properties for L10 Alloys

AlloyUltimate tensile strength σu, MPaPlasticity δ, %Curve
FePd1600253
CoPt2250252
NiPt2200404
CuAu1400254

As mentioned above, recrystallisation slows down over a specific temperature interval near Tc on annealing at T < Tc. In this temperature interval an optimum structural state can be produced in a reasonable annealing time. This state is characterised by a combination of high strength (ultimate tensile strength), σu, and adequate plasticity, δ, see the Table. The curves for these states are the longest deformation curves obtained for the ordered alloys.

If the annealing temperature is lowered recrystallisation accelerates, which suggests that ordering is responsible for the anomaly in the temperature dependence of the recrystallisation rate. In these cases, see Figure 1, the properties become inferior to those observed in the optimum state.

At the optimum temperature, the deformation characteristics of these four alloys change similarly with annealing time, t . Figure 2 shows typical behaviour of the FePd alloy. As annealing time increases, the plasticity,, at a given annealing temperature, increases, reaches a maximum (after ~ 1 hour) and then decreases. When δ is a maximum, the highest ultimate tensile strength, σu, is simultaneously reached. After ~ 1 hour the yield stress, σ0.2, is relatively high too, but is less than that of the unannealed heavily deformed disordered alloy. In fact, we should say that there is a temperature interval over which such curves are observed, rather than just an optimum temperature. Over this temperature interval the dependence of the deformation characteristics on annealing time is similar in shape to that shown in Figure 2. Also, the properties that the alloys attain over this temperature-time interval are of high value — on suitable treatment. These properties vary a little over this temperature interval.

Fig. 2

The ultimate tensile strength, σu, the yield stress, σ, and the elongation before failure (plasticity), δ, of the FePd alloy, versus the annealing time at 600°C

The ultimate tensile strength, σu, the yield stress, σ, and the elongation before failure (plasticity), δ, of the FePd alloy, versus  the annealing time at 600°C

For the CoPt alloy, the dependence of the deformation characteristics on annealing time, after cold drawing to different degrees of deformation, is shown in Figure 3 (8). Variations in the yield stress (σ0.2) and plasticity (δ) are shown for samples annealed at 650°C (Figures 3(a) and 3(b)) and at 750°C (Figures 3(a’) and 3(b’)). All the dotted lines (Curves 1) correspond to annealing the quenched alloy. In Figure 3(b) δ ≅ 4%, that is, the plasticity drops below that of the ordered alloy which did not receive preliminary plastic deformation (Curve 1). Annealing at 750°C gives much higher plasticity (Figure 3(b’)). In this case the 8 value exhibits a nonmonotonic dependence and, depending on the degree of preliminary cold deformation, can reach 32%.

Fig. 3

Yield stress, σ0.2, and plasticity, δ, versus the annealing time, for the CoPt alloy at annealing temperatures of 650°C for (a) and (b), and 750°C for (a’) and (b’).

  • Curves 1 – quenching from 1000°C;

  • Curves 2 - cold deformation to 20%;

  • Curves 3 - cold deformation to 60%;

  • Curves 4 - cold deformation to 90%;

  • Curves 5 - cold deformation to 96% (8)

Yield stress, σ0.2, and plasticity, δ, versus  the annealing time, for the CoPt alloy at annealing temperatures of 650°C for (a) and (b), and 750°C for (a’) and (b’).Curves 1 – quenching from 1000°C;Curves 2 - cold deformation to 20%;Curves 3 - cold deformation to 60%;Curves 4 - cold deformation to 90%;Curves 5 - cold deformation to 96% (8)

The CoPt alloy has been shown to acquire an optimum structural state after annealing at a temperature from 700 to 800°C after a preliminary deformation to 60–90% 8. At the lower annealing temperature (650°C), we failed to obtain values for σ0.2 and δ as high as those obtained under the optimum treatment conditions.

Figure 4 shows changes in the mechanical properties of the CoPt alloy after an additional anneal at 600°C, following a first anneal at the optimum temperature (790°C). Annealing times at 790°C were as follows: 5 minutes (Curves 1 and 1’); 45 minutes (Curves 2 and 2’); and 4 hours (Curves 3 and 3’) 9. The results confirmed the existence of an optimum temperature and time interval for treatment: annealing at 790°C for 45 minutes (Curves 2 and 2’), which result in a set of high strength and high plastic properties. For a shorter annealing time, plasticity decreases a little but the strength characteristics are enhanced. Increasing the annealing treatment time at the optimum temperature is inefficient (Curves 3 and 3’) and impairs the mechanical properties. The plateaux which appear in the –0.2(t ) and δ(t ) dependences of Curves 2 and 2’, respectively, on the additional anneal show that the optimum structural state is preserved after this double annealing treatment In fact, the high values of these properties are retained even after annealing at 600°C for 100 hours. Note that if cold deformation is followed only by annealing at 600°C, the values of the properties would decrease due to rapid recrystallisation, (as happened in Figures 3(a) and 3(b) at 650°C).

Fig. 4

Yield stress, σ0.2, (Curves 1’-3’) and plasticity, 5, (Curves 1–3) for the CoPt alloy versus the annealing time at 600°C. The alloys were preliminary cold deformed to 90% and then ordered by annealing at 790°C for: 5 min (Curves 1’, 1): 45 min (Curves 2’, 2): and 4 h (Curves 3’, 3) 9

Yield stress, σ0.2, (Curves 1’-3’) and plasticity, 5, (Curves 1–3) for the CoPt alloy versus  the annealing time at 600°C. The alloys were preliminary cold deformed to 90% and then ordered by annealing at 790°C for: 5 min (Curves 1’, 1): 45 min (Curves 2’, 2): and 4 h (Curves 3’, 3) 9

These experiments indicated that when thermal treatment is performed at optimum conditions, the dislocations, inherited from the preliminary deformation, become immobile due to the rapid ordering processes. Consequently, recrystallisation is retarded, that is, retarded recrystallisation enables the L10 ordered alloys to retain their strength properties and their increases in plasticity (but for L12 alloys the plastic properties remain the same as for the deformed alloy — see below).

Ordered Alloys with L12 Superstructure

For comparison, data on the mechanical properties of alloys with the L12 superstructure are presented. Figure 5(a) shows deformation curves of Cu3Au alloy. Even after 120 hours annealing (at 350°C) (Curve 3) the yield stress has decreased only slightly compared to that of the cold- deformed alloy (Curve 1). Similar behaviour is exhibited by alloys Pd3Fe (Tc = 650°C) Figure 5(b), and Pt3Co (Tc = 750°C) Figure 6, ordered after cold deformation. Curves 1 and 2 in Figure 6 correspond to annealing at lower and higher temperatures, respectively, than the ordering point, Tc. After annealing at 600°C (that is, below Tc) the σ0.2 and σu values of the Pt3Co alloy remain virtually unchanged, while the δ value reaches a plateau in 3 to 5 hours and then remains unchanged with annealing time increasing up to 300 hours (10).

Fig. 5

Deformation curves for alloys:

  • Cu3Au: (1) after c.d. to 75%;

  • same + annealing at 350° C for C 1 h

  • same but annealing at 350°C for 120 h

  • same but annealing at 300”C for 1 h

  • same but annealing at 300°C for 120 h

  • same but annealing at 375°C 1 h

  • quenching from 650°C (1 h)

  • quenching from 650° C (1 h) + annealing at 300°C for 120 h (data from R. A. Sasinova)

  • Pd3Fe: (1) after c.d. to 75%;

  • same + annealing at 500° C for 1 h

  • same but annealing at 450°C for 1 h

  • same but annealing at 450°C for 72 h

  • same but annealing at 500°C for 72 h

  • quenching from 800°C (1 h) + annealing at 450°C for 72 h

  • quenching from 800°C (1 h)

Deformation curves for alloys: Cu3Au: (1) after c.d. to 75%;same + annealing at 350° C for C 1 hsame but annealing at 350°C for 120 hsame but annealing at 300”C for 1 hsame but annealing at 300°C for 120 hsame but annealing at 375°C 1 hquenching from 650°C (1 h)quenching from 650° C (1 h) + annealing at 300°C for 120 h (data from R. A. Sasinova)Pd3Fe: (1) after c.d. to 75%;same + annealing at 500° C for 1 hsame but annealing at 450°C for 1 hsame but annealing at 450°C for 72 hsame but annealing at 500°C for 72 hquenching from 800°C (1 h) + annealing at 450°C for 72 hquenching from 800°C (1 h)

Fig. 6

Mechanical properties for Pt3Co alloy versus annealing time at 600°C (Curves 1) and 780°C (Curves 2) after cold deformation by drawing to 60%:

  • the ultimate strength, σu

  • the yield stress, σ0.2

  • the plasticity, δ

Mechanical properties for Pt3Co alloy versus  annealing time at 600°C (Curves 1) and 780°C (Curves 2) after cold deformation by drawing to 60%: the ultimate strength, σuthe yield stress, σ0.2the plasticity, δ

Isothermal hardness curves for Cu3Au (5) and Ni3Fe (1) alloys annealed after a strong deformation prove reliably that the dislocation framework inherited by the L12 superstructure is much more rigid than that inherited by the L10 superstructure. If a decrease in hardness after heat treatment occurs, it corresponds to either annealing at a temperature above Tc or to annealing an alloy that cannot be ordered because of its nonstoichiometry. Note that hardness of the Cu3Au alloy decreases only after long-term annealing due to the start of recrystallisation. The Ni3Fe alloy does not soften at all after annealing, because this alloy does not recrystallise in the ordered state.

A comparison of the deformation behaviours of alloys with L12 and L10 superstructures shows clearly that these alloys differ largely with respect to the temperature interval over which recrystallisation decelerates and the yield stress changes just a little. When alloys with L10 superstructure are treated in this temperature interval, their plasticity increases considerably (see Figures 1 to 4) and reaches a maximum in the optimum structural state. This behaviour is not characteristic of alloys with L12 superstructure. To determine how to produce optimum structural states in alloys with L10 superstructure, microstructure development in the FePd alloy, annealed after heavy deformation, was analysed.

Development of the FePd Alloy Microstructure

An equiatomic FePd alloy (34.4 wt% Fe) was melted in an induction furnace under an argon atmosphere from 99.8% palladium and 99.92% carbonyl iron that had been sintered in hydrogen. The alloys were cold-drawn to a 0.5 mm diameter wire with intermediate annealings at 900°C. The total deformation was 75%.

The microstructure of the alloy was examined in a JEM-200CX transmission electron microscope (TEM). To prepare thin foils, the wire was mechanically thinned and then polished in an electrolyte comprising hydrochloric and acetic acids in a volume ratio of 1:4 at a direct current density of ~ 40 mA mm-2 (voltage = 200-230 V). At the final stage, the foils were chemically polished in aqua regia heated to a temperature of 60–70°C.

Figure 7 presents curves showing the ordering kinetics of the drawn FePd alloy after isothermal annealing. The ‘C’-curve for the FePd alloy, ordered without preliminary deformation, is also shown (the dashed line). In comparing curves that correspond to the 90% ordered phase, it is seen that the dashed line is almost fully shifted to the right, that is to a longer annealing time. For FePd, ordering is thus accelerated by deformation.

Fig. 7

Ordering kinetics of temperature/ annealing time curves of the FePd alloy ordered (to 5, 20, 50 and 90%) after drawing to 90%. The dashed line corresponds to the ordering kinetics of the FePd alloy quenched from 800°C after a preliminary deformation

Ordering kinetics of temperature/ annealing time curves of the FePd alloy ordered (to 5, 20, 50 and 90%) after drawing to 90%. The dashed line corresponds to the ordering kinetics of the FePd alloy quenched from 800°C after a preliminary deformation

Figure 8 shows the initial structure of a heavily deformed disordered FePd alloy. A fibrous structure elongated along the wire axis is clearly visible in the optical micrograph (Figure 8(a)). The transverse fibre size is 5–10 µ m. A TEM examination of the initial structure revealed an elongated cellular structure in the fibres. The cellular structure is an order of magnitude smaller than the fibrous structure. Figure 8(b) shows microstrips with a high density of dislocations.

Fig. 8

Structure of disordered FePd alloy drawn to 75%:

  • optical micrograph of the fibrous structure (longitudinal section):

  • TEM image of a microstrip structure

Structure of disordered FePd alloy drawn to 75%: optical micrograph of the fibrous structure (longitudinal section):TEM image of a microstrip structure

FePd (Tc = 650°C) samples annealed at 600°C for 1, 3, 6,14 and 24 hours after preliminary deformation were examined by TEM. No recrystallisation was detected in the samples annealed for 1, 3 and 6 hours. These samples retained their initial fibrous and cellular structure, but after annealing for 1 hour a lamellar structure was observed inside the cells. Typical micrographs are shown in Figure 9. The lamellar structure represents colonies of twin-like c-domains separated by boundaries parallel to {101} planes. The electron diffraction pattern corresponding to the dark-field image shown in Figure 9(a) contains superlattice reflections. This observation of lamellar structure directly confirms the presence of ordering. Thus, although the fibrous and cellular structures are preserved, the optimum state is ordered. A longer annealing treatment initiates recrystallisation. Isolated recrystallised grains in a sample annealed at 600°C for 14 hours are seen in Figure 9(b). Unrecrystallised regions are still retained, even after annealing at 600°C for 24 hours.

Fig. 9

Structure of the FePd alloy annealed at 600°C after drawing to 75%:

  • dark-field image (taken in the (110) reflection) of the lamellar structure after annealing for 3 hours, and

  • recrystallised fine grains after annealing for 14 hours

Structure of the FePd alloy annealed at 600°C after drawing to 75%: dark-field image (taken in the (110) reflection) of the lamellar structure after annealing for 3 hours, andrecrystallised fine grains after annealing for 14 hours

After annealing at 575°C a lamellar structure and slow recrystallisation were observed. The evolution of a drawn structure at this annealing temperature is similar to that observed at 600°C, but recrystallisation begins earlier. The cell walls, colonies of differently oriented c -domains in the cells, and very few fine grains can be seen in samples annealed at 575°C for 2 hours. However, after the sample was annealed at 575°C for 4 hours (Figure 10) a considerable part of its volume had recrystallised, although the lamellar structure was still preserved in some regions. Comparing the microstructures annealed at 600 and 575°C shows that the recrystallised volume of the sample annealed at 575°C for 4 hours is the same as that of the sample annealed at 600°C for 14 hours.

Fig. 10

Junction of a recrystallised grain and a lamellar structure in the FePd alloy annealed at 575C for 4 hours after drawing to 75%

Junction of a recrystallised grain and a lamellar structure in the FePd alloy annealed at 575C for 4 hours after drawing to 75%

Figure 10 shows parallel antiphase boundary strips terminating at the boundaries of new grains. This configuration is typical of alloys ordered after cold deformation. Cahn also noted this fact in his review (Fig. 5 in (1)). We have also observed a similar configuration in the CuAu alloy ordered after cold deformation (Fig. 4 in (11)).

A lamellar structure was not observed in FePd samples deformed and then annealed at 500°C (Figure 11). A considerable fraction of the sample volume of the FePd (shown in Figure 11) was recrystallised after annealing for 24 hours.

Fig. 11

Formation of recrystallised grains in the FePd alloy annealed at 500°C for 8 hours after drawing to 75%

Formation of recrystallised grains in the FePd alloy annealed at 500°C for 8 hours after drawing to 75%

Series of Typical Structural States

All the above thermomechanical treatments were used to realise numerous structural states that correspond to different deformation curves. Typical states, A, B, C and D, corresponding to the deformation curves given in Figure 12, are outlined. Values of the deformation characteristics and the microstructure scales are for a FePd alloy deformed by drawing to 90% (Figure 1(a)).

Fig. 12

Deformation curves for typical structural states

Deformation curves for typical structural states

  • The disordered state of the alloy subjected to a heavy cold deformation (Curve 1 in Figure 1 (a)) is characterised by a fibrous structure (Figure 8(a)) with finer cells inside the fibres, a high yield stress (1000 MPa) approaching the ultimate tensile strength, and a low plasticity (δ = 2.5%).

  • The recrystallised state of the disordered alloy (Curve 8 in Figure 1(a)) is characterised by 0.2 400 MPa, σu = 950 MPa and δ= 27%.

  • The optimum structural state of the alloy ordered after a heavy deformation (Curve 3 in Figure 1(a)) is characterised by fibrous and cellular structures and a lamellar structure inside the cells (Figure 9). The lamellae are not more than 0.05 µm thick. The fibres and the cells differ little in size to those typical of state (A). However, a recovery occurs, and the yield stress is less than the yield stress attained in state (A).

  • The recrystallised state of the ordered alloy (Curve 4 in Figure 1(a)) is characterised by the absence of lamellar structure and by σ0.2 = 400 MPa, σu = 1100 MPa and δ =12%.

Comparing the properties in the typical structural states, it can be seen that the yield stress of the deformed disordered alloy (A) is almost preserved in the optimum state (C) but the strength and plasticity in state (C) increase considerably and exceed those obtained in the other structural states. In other words, the plasticity of the alloy increases and the yield stress remains at a high level under the optimum conditions that ensure ordering after a heavy plastic deformation.

A Model of Deformation Behaviour and Causes of Improved Plasticity

The question arises as to the mechanisms by which alloys that order by formation of the L10- type structure are plasticised, as the plasticity of polycrystals in ordered state (D) is much lower than the plasticity of polycrystals in disordered state (B) (Curves 8 in Figure 1). This can be seen by comparing the two lowest curves in Figures 1 (a) to 1(d). The effect is similar in all the alloys studied. A decrease in plasticity is related to grain-boundary embrittlement resulting from a displacement of impurity atoms to grain boundaries upon ordering. It is known that grain-boundary embrittlement caused by segregation of impurity atoms also occurs in inter- metallics, such as Ni3Al (12), but in this case choosing single crystals substantially increases plasticity. However, unlike intermetallics, the ordered polycrystalline alloys used in our work also exist in a disordered state in which they can be strongly deformed.

The effect of a heavy deformation shows up first as changes in the grain-boundary structure of an alloy in a disordered state. Drawing causes the formation of a fibrous structure (Figure 8(a)), where the fibres differ by the angle of their rotation about the wire axis. The fibre boundaries have a different structure and larger total surface area than the grain boundaries. Consequently, the danger of embrittlement caused by the segregation of impurities at the fibre boundaries upon subsequent ordering is eliminated.

Mughrabi (13) proposed that the microstructure of strongly deformed materials was composite-like with one component (the cell walls) being characterised by a high dislocation density and the other component (intracellular volumes) being free of dislocations. Since the yield stress approaches the ultimate tensile strength upon stretching, the deformation curves of such microstructures are the shortest among those shown in Figure 1. Therefore, although the structure may be considered to be composite-like, the combination of high strength and adequate plasticity, which is typical of composites, does not occur in this state, and the deformation behaviour is not composite-like.

After rapid ordering, the composite-like structure in L10 and L12 alloys is preserved and recrystallisation decelerates considerably. Consequently, a high yield stress is retained after rapid ordering and the formation of both the L10 and L12 superstructures. However, the deformation curves (Figures 5 and 6) of the alloys with L12 superstructure remain relatively short, although they are a little longer than those in the initial state (after cold deformation by drawing). Nevertheless, some features of the recovery and plasticisation distinguish this superstructure state from the initial state. In any case, the deformation behaviour of alloys with the L12 superstructure is not composite-like.

Thus, if alloys ordered into L12 or L10 types after cold deformation are compared, both types preserve the main microstructural elements (fibres and cells) and the dislocations forming the cell walls become immobile, but optimum properties can only be realised in alloys with the L10 superstructure. It is reasonable to assume that the difference in deformation behaviour depends on some microstructural element which is present only in the L10-type alloys. Based on the TEM results it may be stated that the lamellar structure, which appears in L10 alloys, prevents the concentration of stresses at the cell walls, that is, this structure acts as a buffer during plastic deformation.

Objections may be raised to this explanation: the situation may be simpler and the strength of the lamellar structure itself may be the decisive factor. However, the strength of a CoPt polydomain is less than 1000 MPa, which is much smaller than the strength (> 2000 MPa) attained in the optimum state. So, an optimum structural state is reached thanks to the following three factors:

  • First, recrystallisation decelerates upon ordering and thus the initial microstructure (fibres and cells) of a heavily deformed alloy remains almost unchanged.

  • Second, the cell walls become more rigid when the dislocations lose their mobility upon ordering.

  • Third, the cell walls become stronger due to the formation of a lamellar structure.

The question arises whether an optimum structural state can be realised in other superstructures. It can, if the following conditions are simultaneously fulfilled:

  • the alloy in the disordered state is sufficiently ductile and lends itself to heavy cold deformation;

  • ordering is fast enough over some temperature interval so that the dislocations lose their mobility, and as a result, recrystallisation is suppressed and the framework becomes rigid;

  • the ordered matrix has a specific structure, which decreases the danger of stress concentration at the framework-matrix interface, and, thus ensures the high strength of the framework.

II. Anomalous Temperature Dependence of the Yield Stress in FePd Alloys

A lamellar structure, which ensures an optimum state in the FePd alloy, impedes recrystallisation and inhibits the formation of coarse grains. However, if the deformation characteristics of a polycrystal have to be measured in order to estimate the values for a single crystal, then a coarse-grain polycrystal is required. To produce coarse-grain polycrystal the following treatment was used. A disordered FePd alloy was deformed, as before, to 75% and annealed at 850°C (above Tc). The recrystallised disordered alloy was then ordered by slow cooling to 400°C at the rate of 10 deg/day. This treatment produced an ordered polycrystal with grains about 20 µm in size. The grains included several c -domains the boundaries of which could be clearly seen in an optical microscope. Thin (0.12 mm) bands of the test polycrystal were stretched in the temperature interval from room temperature to 600°C. Figure 13 shows the yields stress, σ0.2, versus the test temperature, T. It can be seen that the dependence σ0.2(T) is nonmonotonic with a maximum at Tmax =200°C, and with anomalous behaviour at T < Tmax, that is, (σ0.2 increases with temperature; at T > Tmax there is a normal trend. A plateau is observed over the interval from 300 to 400°C.

Fig. 13

The nonmonotonic temperature dependence of σ0.2 in a FePd polycrystal. The error in reading σ 0.2 is due to the small transverse dimensions of the samples: 0.12 x 3.5 mm. The test portion of the samples was 30 mm long

The nonmonotonic temperature dependence of σ0.2 in a FePd polycrystal. The error in reading σ 0.2 is due to the small transverse dimensions of the samples: 0.12 x 3.5 mm. The test portion of the samples was 30 mm long

We obtained an analogous σ0.2(T) dependence for a polycrystal of an ordered CuAu alloy with grains ~ 5 pm in size. The CuAu polycrystal was prepared by a similar treatment (14). As is known, the critical size of grains below which the σ0.2(T) anomaly is not observed is different in different alloys: ≈ 8 µm in Ni3Al (15) and ttt 50 µm in TiAl (16). Actually, as follows from the analysis performed in (17) for the interaction of dislocations with c-domain boundaries, the intensive retardation of dislocations at a boundary of this type arises from the fact that at transition from one domain to another there occurs either a change in the type of dislocations (a single dislocation —> superdislocation), or a change in the configuration of the superdislocation, or, by virtue of the tetrag- onality of the lattice, a change in the Burgers vector of dislocation. The latter case is accompanied by the generation of domain boundary dislocations (similar to grain boundary dislocations). In this case the flow stress σi, has the following form: a plateau and high-temperature drop (Figure 14). For the case of grain boundary hardening the flow stress has the analogous form. According to existing concepts the thermal work hardening effect is due to thermally activated transformations of glissile dislocations into sessile ones. In this case the flow stress σs has the temperature peak (Figure 15) (18).

Fig. 14

The temperature dependence of the yield stress of the CuAu alloy for the single domain (I) and polydomain (2)

The temperature dependence of the yield stress of the CuAu alloy for the single domain (I) and polydomain (2)

Fig. 15

Temperature dependence of the yield stress of Ni3Ga alloy for different orientations of the single crystal (18)

Temperature dependence of the yield stress of Ni3Ga alloy for different orientations of the single crystal (18)

The possibility of the superposition of domain (grain) boundary and thermal hardening of superstructures was proposed by Greenberg and colleagues (19). The superposition of stresses is described by the equation:

Possible forms of stress are presented in Figure 16.

Fig. 16

(a), (b) Possible forms of curves σj(T) - grain boundary stengthening and σs(T) - thermal strengthening (c) (e) Variations of the superposition of thermal and grain boundary strengthening

(a), (b) Possible forms of curves σj(T) - grain boundary stengthening and σs(T) - thermal strengthening (c) (e) Variations of the superposition of thermal and grain boundary strengthening

The curve σ0.2(T) in Figure 13 epresents a variant of the superposition of thermal and grain-boundary (or domain-boundary) strengthening (Figure 16(c)). What is actually meant here is the methods of superposition of the flow stress, which has a temperature maximum, and the stress required for dislocations to break through the boundaries. The latter stress is described by a curve with a plateau, which is replaced by a descending trend with increasing temperature. The ratio between the temperature at the plateau (Ti in Figure 16) and the temperature of the maximum (Ts in Figure 16) largely determines the shape of the σ0.2(T) curve resulting from the superposition.

From the data it follows that the temperature anomaly of the yield stress may be observed in a single crystal of an ordered FePd alloy. However, different situations may arise. If the FePd alloy is similar to TiAl in the sense that some interatomic bonds are covalent-like (20), the σ0.2(T) anomaly will be observed at any orientation, favourable for both superdislocations and single dislocations. Otherwise, single dislocations will not have the thermally activated mechanism of blocking, as opposed to superdislocations. This situation takes place in CuAu where blocked single dislocations were not detected (14). But then one may expect that an anomalous σ0.2(T) will be observed only at orientations when the Schmid factor of single dislocations is very small.

Conclusions

The FePd alloy was selected for study from the large group of alloys having a L10 superstructure. There is growing interest in FePd-based alloys (used previously for permanent magnetic materials) as a nanocrystalline medium for high density magnetic recording. For this reason a knowledge of the structural states of FePd is necessary, and analysis of some of them is presented here.

BACK TO TOP

References

  1. 1
    R. W. Cahn, ‘Recovery, Strain Age-Hardening, and Recrystallization in Deformed Intermetallics’, in “ High Temperature Aluminides and Intermetallics ”, eds. S. H. Whang et al., The Minerals, Metals, and Materials Society, Warrendale, PA 1990, p. 245
  2. 2
    R. W. Cahn, in “ Intermetallic Compounds JIMIS- 6 ”, ed. O. Izumi, The Japan Institute of Metals, Sendai, 1991, p. 771
  3. 3
    B. A. Greenberg and Yu. N. Gornostirev, Scr. Metall., 1985, 19, 1391
  4. 4
    B. A. Greenberg and Yu. N. Gornostirev, Scr. Metall., 1985, 19, 1397
  5. 5
    W. B. Hutchinson,, P. M. Besag and G. V. Honess, Ada Metall., 1973, 21, 1685
  6. 6
    L. R. Weisberg and S. L. Quimby, Phys. Rev., 1958, 110, 338
  7. 7
    T. Takasugi and O. Izumi, Acta Metall., 1985, 33, 49
  8. 8
    L. G. Grokhovskaya,, B. A. Greenberg,, A. E. Ermakov and B. P. Adrianovskii, Fiz. Met. Metalloved., 1988, 65, ( 5 ), 1007
  9. 9
    L. G. Grokhovskaya,, B. A. Greenberg, A. E. Ermakov et al., Fiz. Met. Metalloved., 1989, 67, ( 5 ), 983
  10. 10
    V. N. Indenbaum,, Yu. N. Gornostyrev, B. A. Greenberg et al., Fiz. Met. Metalloved., 1989, 68, ( 2 ), 382
  11. 11
    B. A. Greenberg,, G. Hug, O. V. Antonova et al., Intermetallics, 1997, 5, ( 4 ), 297
  12. 12
    K. Aoki and O. Izumi, Mater. Trans., JIM, 1978, 19, 203
  13. 13
    H. Mughrabi,, T. Ungar,, W. Kienie and M. Wilkens, Philos. Mag. A, 1986, 53, 793
  14. 14
    B. A. Greenberg,, O. V. Antonova and A. Yu. Volkov, Intermetallics, 1999, 7, ( 11 ), 1219
  15. 15
    T. P. Weihs,, V. Zinoviev,, D. V. Viens and E. M. Schulson, Acta Metall, 1987, 5, 1109
  16. 16
    S. C. Huang, Scr. Metall., 1988, 22, 1885
  17. 17
    B. A. Greenberg,, V. I. Syutkina and E. S. Yakovleva, Fiz. Tverdogo Tela, 1968, 10, 1330
  18. 18
    S. Takeuchi and E. Kuramoto, Acta Met., 1973, 21, 415
  19. 19
    B. A. Greenberg and Yu. N. Gomostirev, Scr. Metall., 1982, 16, ( 1 ), 15
  20. 20
    B. A. Greenberg,, O. V. Antonova,, A. Yu. Volkov and M. A. Ivanov, Intermetallics, 2000, 8, ( 8 ), 845

Acknowledgements

This study has been financed in terms of the programme ‘The National Technological Basis’ (subprogramme Technology of New Materials’) and VI Competition among young scientists projects of the RAS (Grant No. 69).

The Authors

B. A. Greenberg is a Professor and Head of Laboratory at the Institute of Metal Physics, Ural Branch, RAS. Her scientific interests focus on a wide scope of problems in materials science, including phase transformations and the theory of strength.

N. A. Kruglikov is a Junior Researcher at the Institute of Metal Physics, Ural Branch, RAS. He analyses properties of ordered alloys and intermetallics.

L. A. Rodionova is a Senior Researcher at the Institute of Metal Physics, Ural Branch, RAS. She is a specialist in electron microscopy and currently studies the structure of Nb-Cu-Sn composites.

A. Yu. Volkov is a Senior Researcher, Ural Branch, Institute of Metal Physics, RAS. He studies phase transformations, the microstructure, mechanical and electrical properties of ordered noble metal alloys.

L. G. Grokhovskaya is a Head of Laboratory at the Ekaterinburg Nonferrous Metals Processing Plant. She develops new pgm materials for technical applications and examines their properties.

G. M. Gushchin is a Head of Laboratory at the Ekaterinburg Nonferrous Metals Processing Plant. Platinum materials manufacture and pgms for high-temperature applications are his main concerns.

I. N. Sakhanskaya is a Leading Specialist at the Ekaterinburg Nonferrous Metals Processing Plant. Her main field of interests includes phase transformations and analysis of the properties of pgm alloys for the instrument-making industry.

Read more from this issue »

BACK TO TOP

SHARE THIS PAGE: