A polymer-like ultrahigh-strength metal alloy

doi:10.21203/rs.3.rs-3649839/v1

Download PDF

Physical Sciences - Article

A polymer-like ultrahigh-strength metal alloy

https://doi.org/10.21203/rs.3.rs-3649839/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 04 Sep, 2024

Read the published version in Nature →

Version 1

posted

You are reading this latest preprint version

Futuristic technologies like morphing aircrafts and superstrong artificial muscles are hinged on metal alloys being as strong as an ultrahigh-strength steel (with a high yield strength σ_y>1 GPa) yet as flexible as a polymer (with an ultralow elastic modulus E ~10 GPa)^1-3. However, achieving such “strong yet flexible” alloys has proven challenging^4-9. The difficulty lies in an inevitable trade-off between strength and flexibility^5,8,10, which precludes a high-strength alloy from being of polymer-like ultralow modulus. Here we report a Ti-50.8 at.% Ni strain glass alloy showing an unprecedented combination of an ultrahigh yield strength σ_y ~1.8 GPa with a polymer-like ultralow elastic modulus E ~10.5 GPa, together with a superlarge rubber-like J-shaped elastic strain of ~8%. As a result, it possesses the highest flexibility figure of merit σ_y/E ~0.17 which far exceeds that of existing structural materials. This alloy was fabricated by a simple 3-step thermomechanical treatment, which leads to not only ultrahigh strength but also ultralow modulus through forming a unique “dual-seed strain glass” (DS-STG) microstructure, being a strain glass matrix embedded with a small amount of R and B19' martensites. In-situ x-ray diffractometry reveals that the DS-STG enables a nucleation-free reversible transition between strain glass and R and B19’ martensites during stress loading/unloading, thereby leading to ultralow modulus and large recoverable strain with narrow hysteresis. Our finding may open a new horizon for designing and mass-producing strong and flexible alloys and such alloys may lead to important applications.

Physical sciences/Nanoscience and technology/Nanoscale materials/Structural properties

Physical sciences/Materials science/Structural materials/Metals and alloys

Physical sciences/Materials science/Condensed-matter physics/Phase transitions and critical phenomena

Exotic metal alloys showing simultaneously steel-like ultrahigh strength and polymer-like ultrahigh flexibility have long been desired for many emerging technologies such as morphing aircrafts^1,2 and superman-type artificial muscles³. Such an unconventional “strong yet flexible” property combination (red regime of Fig. 1a) will enable a large recoverable shape change under a small driving power and simultaneously possess strong resistance against fracture or yielding failure under large load. As a result, such alloys would enable a morphing wing in an aircraft or superstrong artificial muscle in a humanoid robot, to name just a few possibilities.

However, achieving such unconventional alloys has remained challenging due to the trade-off relation between strength and flexibility (the latter being measured by elastic compliance or inverse of elastic modulus)^5,8,10, as manifested by the grey band (known as Ashby plot⁴) of Fig. 1a. It reveals a common observation^4,10,11 that steels can be made very strong (with a high yield strength σ_y>1 GPa) but they are stiff (with a high Young’s modulus E ~200 GPa), whereas organic materials such as fiber-reinforced polymers (FRP) have the opposite property of being flexible (with a typical low E ~10 GPa) but they are weak (σ_y< 0.3 GPa). The strength-flexibility trade-off stems from the fact that the two properties are oppositely correlated to the bonding strength of a material, which is represented roughly by elastic modulus^5,8,10. Therefore, this inevitable trade-off relation precludes a 1 GPa-class steel-like high-strength alloy from being of 10 GPa-class polymer-like low modulus.

Over the past decades, significant efforts have been made to seek metal alloys with simultaneously high strength and low modulus^4-9,12-17; but an alloy showing both steel-like high yield strength (σ_y>1 GPa) and polymer-like low modulus (E ~10 GPa) still remains unattainable. So far, several alloys based on shape memory alloys (SMA)^5,16,17 are reported to reveal 1 GPa-class high strength and a moderately low modulus E ~30 GPa, and a conventional Mg-Sc strain glass alloy⁷ has recently been shown to possess a lower modulus E ~20 GPa but with a lower strength σ_y ~0.3 GPa. Despite these efforts, existing alloys still fall into the conventional grey band of Fig. 1a, and the desired polymer-like ultrahigh-strength property (red regime of Fig. 1a) still remains elusive.

Here we show that a Ti-50.8 at.% Ni “dual-seed strain glass” (DS-STG) alloy can overcome the strength-flexibility trade-off and demonstrates simultaneously a steel-like ultrahigh yield strength σ_y ~1.8 GPa and a polymer-like ultralow Young’s modulus E ~10.5 GPa (corresponding to an ultrahigh elastic compliance S=1/E ~0.1 GPa^-1) (Fig. 1a), together with a superlarge rubber-like J-shaped elastic strain of ~8% (Fig. 1b). As a result, it exhibits an unprecedented flexibility figure of merit σ_y/E ~0.17 which far exceeds that of existing structural materials (Fig. 1c), and the alloy behaves like a superstrong polymer (Fig. 1d). In the following, we shall first demonstrate its unique mechanical properties, then show its 3-step thermomechanical fabrication route towards its unique DS-STG microstructure, and finally discuss the origin of mechanical properties in relation to the DS-STG microstructure.

Unique mechanical properties of DS-STG alloy

Figure 1a shows that the DS-STG alloy overcomes the well-observed strength-flexibility trade-off relation of existing structural materials (shown as the grey band of Fig. 1a) and exhibits an unprecedented combination of ultrahigh-strength-steel class yield strength of σ_y ~1.8 GPa and a polymer-like ultralow Young’s modulus E ~10.5 GPa (corresponding to ultrahigh elastic compliance of S=1/E ~0.1 GPa^-1). Such a property combination places the DS-STG alloy into a hitherto inaccessible “strong yet flexible” regime (red regime) in Fig. 1a.

Figure 1b shows that the DS-STG alloy demonstrate a J-shaped stress-strain curve with a superlarge pseudoelastic strain ε_re ~8% before reaching its elasticity limit at σ_y~1.8 GPa for wire samples and σ_y~1.3 GPa for plate samples. The 1.8 GPa yield strength of DS-STG alloy even exceeds that of most ultrahigh-strength steels like quenched spring steel, and the 8% pseudoelastic strain value not only far exceeds the small elastic strain ε_re ~0.2-1% for typical metal alloys^18-20 but also is larger than that (ε_re ~1-5%) of many organic structural materials like wood, bamboo, bones, polymer, and fiber-reinforced polymers^21-24. It is noted that the large J-shaped elasticity behavior associated with a small stress hysteresis of ~15% resembles that of ultra-flexible materials like rubbers²⁵, and it is very different from the superelasticity of conventional shape memory alloys, which is characterized by a stress plateau with a much larger initial modulus of E~30-75 GPa and a huge hysteresis (~50-90%)^26-28. The strongly hysteretic behavior of shape memory alloys is known to be undesirable for many applications^28-30.

The combination of ultrahigh strength and ultralow modulus endows the DS-STG alloy with a superhigh flexibility figure of merit σ_y/E ~0.17, far exceeding that of existing structural materials including typical metal alloys and organic materials, as shown in Fig. 1c. Flexibility figure of merit σ_y/E is a measure of ultimate material flexibility before yielding failure¹⁰, and existing structural materials usually have a much lower σ_y/E ~0.01-0.06.

Figure 1d and Supplementary Movie 1 show visual evidence for the unconventional “strong yet flexible” (i.e., polymer-like ultrahigh-strength) property of the DS-STG alloy, which contrasts with the conventional behavior of two reference materials: a “strong & stiff” spring steel (with high modulus E ~190 GPa and high strength σ_y ~1.36 GPa), and a “weak & flexible” FRP (with low modulus E ~11 GPa and low strength σ_y ~0.13 GPa). Thus, the DS-STG alloy combines the high strength of a steel with the high flexibility of a polymer, and it behaves like a superstrong organic material, which has long been desired by many emerging technologies.

A 3-step thermomechanical fabrication route towards DS-STG alloy and the corresponding microstructure feature after each step

The DS-STG alloy was fabricated through a 3-step thermomechanical processing route from a starting solution-treated Ti-50.8Ni alloy, as shown in Fig. 2a. The simplicity of this processing method makes it scalable to industry lines. The corresponding microstructure after each step is described below.

Step-1 processing: formation of deformation-stabilized martensite

The starting solution-treated Ti-50.8Ni alloy (Fig. 2a(i)) is in its B2 parent state (strictly speaking an unfrozen strain glass state^31-33 (Extended Data Fig. 1a)) at room temperature since it has a sub-zero martensitic transformation temperature of M_s ~250 K. Such a B2 state exhibits a high elastic modulus E~71 GPa and an incomplete superelasticity with a large remanent strain and hysteresis (Extended Data Fig. 1b) during loading/unloading even at a moderate stress of ~0.3 GPa. This is a well-observed behavior of Ti-Ni alloys when dislocation slip occurs during a stress-induced martensitic transformation¹⁴. Thus, the starting B2 alloy is neither a strong alloy nor a polymer-like alloy.

In Step-1 processing, the B2 alloy is severely deformed by 50% tensile elongation (Fig. 2(ii)). This cold-working process results in a deformation-stabilized B19’ martensite which is preserved at room temperature (Extended Data Fig. 2a). This cold-working-strengthened B19’ martensite exhibits a quasi-linear elastic behavior with a high yield stress of ~1.3 GPa and a moderate elastic modulus of E~37 GPa, as shown in Extended Data Fig. 2b. Such mechanical properties are consistent with previous reports on tensile-deformed Ti-Ni alloys^12,34; thus the deformation-stabilized B19’ martensite alloy is a high-strength alloy but not an ultralow modulus alloy.

Step-2 processing: formation of a dual-crossover strain glass (DC-STG) state

In Step-2 processing, the deformation-stabilized B19’ martensite alloy is annealed at 573 K for 10 min (Fig. 2(iii)). Annealing at such a temperature not only fully annihilates the metastable B19’ martensite produced in Step-1 processing, but also results in a unique “dual-crossover strain glass” (DC-STG) state. This DC-STG is an unfrozen R strain glass (R-STG) at room temperature and it undergoes a strain glass transition around T_g ~251 K, which is evidenced by the same signatures as those of a normal strain glass transition reported in the literature^31-33, including a frequency (w) dependent T_g following Vogel-Fulcher law (inset of Fig. 2b) and the invariance of average B2 structure across T_g, together with the appearance of nano-sized R strain domains. But being different from a conventional R strain glass which keeps R nanodomain down to 0 K, this DC-STG is able to crossover to R and B19’ dual martensites at lower temperatures, as evidenced by in-situ transmission electron microscopy (TEM) and X-ray diffractometry (XRD) (Fig. 2b). It is noted that the crossover transitions are incomplete and significant amount of strain glass remains even at the lowest temperature tested. This unique R strain glass that can crossover into R and B19 dual martensites is thus named “dual-crossover strain glass” or DC-STG.

It should be noted that this DC-STG state, albeit a forerunner state to the ultralow-modulus DS-STG state, is not a low modulus state by itself. It shows a non-linear and hysteretic superelastic behavior with a moderate initial elastic modulus of ~33 GPa and a high yield strength of ~1 GPa (Extended Data Fig. 3).

Step-3 processing: formation of dual-seed strain glass (DS-STG) state

In Step-3 processing, the DC-STG alloy undergoes a moderate tensile elongation of ~12% (Fig. 2a(iv)). This process introduces a small amount of R and B19’ martensite “dual-seeds” (50-100nm in size) into the room-temperature DC-STG matrix (manifested as nanosized R domains), as revealed by HRTEM image (Fig. 2c, top). The resultant strain glass state containing R and B19’ dual-seeds is thus named “dual-seed strain glass” (DS-STG). Clearly the realization of this unique DS-STG owes to the capability of its forerunner state DC-STG to crossover into two martensites R and B19’, as revealed in Fig. 2b. Such a unique DS-STG state endows this alloy a polymer-like ultralow modulus E ~10.5 GPa, together with a J-shaped superlarge pseudoelastic strain ε_re ~8% and small hysteresis of ~15% (Fig. 1b). The mechanism of such unconventional elastic behavior will be revealed in Fig. 3.

Low magnification image of the DS-STG alloy (Fig. 2c, bottom) reveals that it contains high density dislocations and a large amount of B2 mechanical twins, which are formed by the severe cold-working during the 3-step processing. These deformation-hardening features are characteristic of severely deformed Ti-Ni alloys^16,35-37, and can account for the ultrahigh yield strength of the DS-STG alloy (Fig. 1b).

To confirm the general applicability of the 3-step thermomechanical processing in achieving polymer-like ultrahigh-strength property, we also tested a wire sample of the Ti-50.8Ni alloy by the same 3-step route but the severe deformation of the sample was done with a wire-drawing machine instead of a tensile machine. As can be seen in Fig. 1b, the wire sample exhibits the same ultralow modulus E ~10.5 GPa as that of the plate sample, but with an even higher yield strength σ_y ~1.8 GPa due to the favorable influence of cold-drawing. Therefore, the present 3-step processing route is applicable to both wire and plate samples with Step-1 deformation being either tensile elongation and cold-drawing.

Origin of the polymer-like ultrahigh strength of DS-STG alloy

In-situ tensile XRD experiment (Fig. 3) shows that the DS-STG alloy with the above “dual-seeds in a STG matrix” microstructural feature (Fig. 2c) undergoes a unique nucleation-free reversible transition between STG and R and B19’ martensites during a stress loading/unloading cycle up to 1.3 GPa and this process leads to the observed ultralow elastic modulus. Before loading (i.e., σ =0 GPa), the DS-STG, being a DC-STG matrix embedded with a small fraction of R and B19’ martensite seeds, reveals a broadened B2 peak (representing the average structure of the random STG nanodomains³⁸) with a long tail covering R and B19’ positions in its XRD profile. With increasing stress, the R and B19’ martensite peaks grow smoothly without needing a critical stress until the whole sample transforms almost fully into B19’ martensite at σ =1.3 GPa. Clearly this is a result of R and B19’ martensite seeds that bypass the nucleation barrier of the stress-induced transition, consequently leading to the polymer-like ultralow elastic modulus (E ~10.5 GPa). Upon unloading from 1.3 GPa, a smooth reverse transition occurs from B19’ to a mixture of R and B19’, and then to the original state of DS-STG. Thus the original DS-STG state is recovered upon unloading. The large pseudoelastic strain (ε_re ~8%) is a natural result of this reversible transformation, as the maximum transformation strain to form B19’ martensite in Ti-Ni alloys is about 8-9%²⁶.

Another important consequence of the nucleation-free DS-STG to R and B19’ transition is its J-shape and narrow hysteresis features in the stress-strain curve of the DS-STG alloy, because the R and B19’ martensite seeds can bypass the nucleation barrier during a stress-induced transition. This results in ultralow initial slope in the stress-strain curve, i.e., the “J-shape”, and small hysteresis. The small hysteresis is also reflected by the similarity of the XRD profiles between loading and unloading at the same stress level. The smooth, J-shape, and narrow-hysteretic behavior of DS-STG, contrasts with that of normal stress-induced STG to martensite transition (Extended Data Fig. 3) or a conventional stress-induced martensitic transformation (Extended data Fig. 1b ), where the inevitable nucleation leads to not only a much higher elastic modulus of E ~30-70 GPa but also a strongly nonlinear and hysteretic stress-strain behavior with a stress plateau^14,16.

Fig. 4a shows that a DS-STG state is essential to achieve a polymer-like ultralow modulus. Such a state is obtained by a 3-step processing with Step-2 annealing at T_a ~573 K. When the Step-2 annealing temperature T_a deviates from the optimal 573 K, the alloy deviates from a DS-STG state and this results in an increase in elastic modulus and a decrease in recoverable strain.

Fig. 4b explains why the DS-STG state is obtained only when the Step-2 annealing temperature T_a is around 573 K. This is because the DS-STG’s forerunner state DC-STG is achieved only around T_a ~573 K, as shown in the phase diagram of Fig. 4b (established from data from Fig. 2b, Extended Data Fig. 5 and 6). When T_a deviates significantly from 573 K, the forerunner DC-STG will no longer exist and consequently a R+B19’ dual-seed strain glass cannot be achieved after Step-3 processing (i.e., 12% elongation).

The phase diagram (Fig.4b) shows that a significantly lower T_a (e.g., T_a ~473 K) will lead to a mono-crossover strain glass (MC-STG) after Step-2 processing, which is a R strain glass that can crossover (partially) into one martensite B19’ at low temperature. After Step-3 processing (12% elongation) it will change into a mono-seed strain glass (MS-STG, i.e., with B19’ seeds only), which has a higher modulus of ~17 GPa (~1.5 times high than that of DS-STG) and a smaller recoverable strain of ~6%.

At a significantly higher T_a ~773 K, which is a common annealing temperature for TiNi alloys²⁶, strain glass transition is no longer existent and the alloy is characterized by a normal 2-step martensitic transformation into R and B19’ martensite. After Step-3 processing (12% elongation) the alloy will change into an aligned B19’ martensite, which has an even higher modulus of ~30 GPa and an even smaller recoverable strain of ~4%.

The lower elastic modulus of the DS-STG state as compared with non-DS-STG states can be explained by the phase instability of the system in the vicinity of three different phases. This results in a nearly flat free energy landscape which facilitates the transition from one phase to another and consequently the lattice becomes the softest. Such an effect has been observed in many ferroelectrics and relaxor materials associated with multiple ferroelectric phases⁴⁰.

A polymer-like ultrahigh-strength metal alloy DS-STG was fabricated by a simple 3-step thermomechanical processing route. It exhibits an unprecedented combination of polymer-like ultralow elastic modulus of ~10.5 GPa and steel-like ultrahigh yield strength of ~1.3-1.8 GPa, together with a large J-shaped elastic strain of ~8%. This unconventional alloy overcomes the long-standing strength-flexibility trade-off which has precluded such a long-sought property from being achieved. The ultralow elastic modulus originates from the unique DS-STG state that enables a nucleation-free reversible transition between strain glass and R and B19’ martensites, and the ultrahigh strength results from deformation-induced strengthening effect. The polymer-like ultrahigh-strength alloy may open the door for a wide range of applications in emerging technologies such as morphing aerospace vehicles, superman-type humanoid robots and advanced biomedical devices. The industry-scalable fabrication route may enable mass-production of such alloys.

Data availability

All data generated or analyzed during this study are included in the published article and Extended Data Figures and are available from the corresponding authors upon request.

Acknowledgements

This research was supported by National Key R&D Program of China (2022YFB3808700), National Natural Science Foundation of China (51831006, 52071257, 12204368 and U2241242) and National 111 Project 2.0 (BP0618008). XR acknowledges the support from JSPS Kakenhi (23H01675).

Author contribution

Y.J. and X.R. conceived the idea. Z.X. and Y.J. designed this project. Z.X., Y.J. fabricated the samples. Z.X. performed mechanical testing and property/phase-transition characterizations. Z.X., T.M., C.L., and L.H. performed the transmission electron microscopy profiles. Z.X. analyzed all results with all the authors. Z.X., Y.J., and X.R. wrote the manuscript with input from all the authors.

Competing interests

The authors declare no competing interest.

Matloff, L. Y. et al. How flight feathers stick together to form a continuous morphing wing. Science. 367, 293–297 (2020).
Royal Aeronautical Society, Shape shifters. https://www.aerosociety.com/news/shape-shifters/ (2021).
Tawfick, S. & Tang, Y. Stronger artificial muscles, with a twist. Science. 365, 125–126 (2019).
Ashby, M. Materials selection in mechanical design. (Elsevier, 2011).
Hao, S. et al. A transforming metal nanocomposite with large elastic strain, low modulus, and high strength. Science. 339, 1191–1194 (2013).
Zhu, J., Wang, D., Gao, Y., Zhang, T.-Y. & Wang, Y. Linear-superelastic metals by controlled strain release via nanoscale concentration-gradient engineering. Mater. Today 33, 17–23 (2020).
Liu, C. et al. A lightweight strain glass alloy showing nearly temperature-independent low modulus and high strength. Nat. Mater. 21, 1003–1007 (2022).
Xu, S. et al. Non-Hookean large elastic deformation in bulk crystalline metals. Nat. Commun. 13, 5307 (2022).
Odaira, T. et al. Flexible and tough superelastic Co–Cr alloys for biomedical applications. Adv. Mater. 34, 2202305 (2022).
Peng, J. & Snyder, G. J. A figure of merit for flexibility. Science. 366, 690–691 (2019).
Zhou, M. Exceptional properties by design. Science. 339, 1161–1162 (2013).
Xu, Z. et al. Simultaneously increasing the strength and decreasing the modulus in TiNi alloys via plastic deformation. Scr. Mater. 209, 114374 (2022).
Wu, C.-T. et al. Machine learning recommends affordable new Ti alloy with bone-like modulus. Mater. Today 34, 41–50 (2020).
Zhao, S. et al. Quasi‐linear superelasticity with ultralow modulus in tensile cyclic deformed TiNi strain glass. Adv. Eng. Mater. 24, 2200239 (2022).
Wu, G., Chan, K.-C., Zhu, L., Sun, L. & Lu, J. Dual-phase nanostructuring as a route to high-strength magnesium alloys. Nature 545, 80–83 (2017).
Liang, Q. et al. Strain states and unique properties in cold-rolled TiNi shape memory alloys. Acta Mater. 231, 117890 (2022).
Hua, P., Xia, M., Onuki, Y. & Sun, Q. Nanocomposite NiTi shape memory alloy with high strength and fatigue resistance. Nat. Nanotechnol. 16, 409–413 (2021).
Sun, W. et al. Precipitation strengthening of aluminum alloys by room-temperature cyclic plasticity. Science. 363, 972–975 (2019).
Zhang, D. et al. Additive manufacturing of ultrafine-grained high-strength titanium alloys. Nature 576, 91–95 (2019).
He, B. B. et al. High dislocation density–induced large ductility in deformed and partitioned steels. Science. 357, 1029–1032 (2017).
Song, J. et al. Processing bulk natural wood into a high-performance structural material. Nature 554, 224–228 (2018).
Zakikhani, P., Zahari, R., Sultan, M. T. H. & Majid, D. L. Extraction and preparation of bamboo fibre-reinforced composites. Mater. Des. 63, 820–828 (2014).
Abdul Khalil, H. P. S. et al. Bamboo fibre reinforced biocomposites: A review. Mater. Des. 42, 353–368 (2012).
Moniruzzaman, M., Chattopadhyay, J., Billups, W. E. & Winey, K. I. Tuning the mechanical properties of swnt/nylon 6,10 composites with flexible spacers at the interface. Nano Lett. 7, 1178–1185 (2007).
Li, X. et al. Preparation of polyurethane/polyvinyl alcohol hydrogel and its performance enhancement via compositing with silver particles. RSC Adv. 7, 46480–46485 (2017).
Otsuka, K. & Ren, X. Physical metallurgy of Ti–Ni-based shape memory alloys. Prog. Mater. Sci. 50, 511–678 (2005).
Tanaka, Y. et al. Ferrous polycrystalline shape-memory alloy showing huge superelasticity. Science. 327, 1488–1490 (2010).
Song, Y., Chen, X., Dabade, V., Shield, T. W. & James, R. D. Enhanced reversibility and unusual microstructure of a phase-transforming material. Nature 502, 85–88 (2013).
Cui, J. et al. Combinatorial search of thermoelastic shape-memory alloys with extremely small hysteresis width. Nat. Mater. 5, 286–290 (2006).
Hou, H. et al. Fatigue-resistant high-performance elastocaloric materials made by additive manufacturing. Science. 366, 1116–1121 (2019).
Sarkar, S., Ren, X. & Otsuka, K. Evidence for strain glass in the ferroelastic-martensitic system Ti_50-xNi_50+x. Phys. Rev. Lett. 95, 205702 (2005).
Zhang, Z. et al. Phase diagram of Ti_50-xNi_50+x: Crossover from martensite to strain glass. Phys. Rev. B 81, 224102 (2010).
Ji, Y. et al. Ferroic glass. in Encyclopedia of Condensed Matter Physics (Academic Press, 2023).
Chen, Y. et al. Revealing the mode and strain of reversible twinning in B19′ martensite by in situ synchrotron X-ray diffraction. Acta Mater. 236, 118131 (2022).
Moberly, W. J., Proft, J. L., Duerig, T. W. & Sinclair, R. Deformation, twinning and thermo-mechanical strengthening of Ti₅₀Ni₄₇Fe₃. Acta Metall. Mater. 38, 2601–2612 (1990).
Šittner, P. et al. On the coupling between martensitic transformation and plasticity in NiTi: Experiments and continuum based modelling. Prog. Mater. Sci. 98, 249–298 (2018).
Chen, Y. et al. Recoverability of large strains and deformation twinning in martensite during tensile deformation of NiTi shape memory alloy polycrystals. Acta Mater. 180, 243–259 (2019).
Wang, Y., Ren, X. & Otsuka, K. Shape memory effect and superelasticity in a strain glass alloy. Phys. Rev. Lett. 97, 225703 (2006).
Zhang, J. et al. Spontaneous strain glass to martensite transition in a Ti₅₀Ni_44.5Fe_5.5 strain glass. Phys. Rev. B 84, 214201 (2011).
Yang, Y. et al. Morphotropic relaxor boundary in a relaxor system showing enhancement of electrostrain and dielectric permittivity. Phys. Rev. Lett. 123, 137601 (2019).
Morawiec, H., Stróż, D., Goryczka, T. & Chrobak, D. Two-stage martensitic transformation in a deformed and annealed NiTi alloy. Scr. Mater. 35, 485–490 (1996).
Chen, H. et al. Improvement of the stability of superelasticity and elastocaloric effect of a Ni-rich Ti-Ni alloy by precipitation and grain refinement. Scr. Mater. 162, 230–234 (2019).

Sample preparation

A commercial Ti-50.8 at. % Ni alloy from SaiTe Inc. Ltd. was used to fabricate both plate samples and wire samples. All samples were solution-treated in their homogeneous B2 state and followed by quenching into water. The plate samples were sealed in vacuumized quartz tube and solution-treated at 1273 K for 90 minutes; the wire samples were solution-treated in an automatic furnace kept at 1023 K that enables the moving wires to enter and exit with a heating duration of 10 min before being quenched. After the solution-treatment, the oxidation affected surface layer of all samples was removed by a chemical etchant.

Both plate and wire samples underwent a similar 3-step processing described below and it is also shown in Fig. 2a, with a difference only in the deformation mode in Step-1. In Step-1 processing the plate samples were severely deformed by 50% cold-elongation and the wire samples were severely deformed by 50% area reduction through cold-drawing. The severe deformation in Step-1 changes the starting B2 phase into a metastable martensite microstructure stabilized by high-density dislocations.

In Step-2 processing, the severely deformed samples were moderately annealed at 573 K for 10 min and followed by quenching into water. This process annihilates the metastable martensite formed during Step-1 processing and yields a unique “dual-crossover strain glass” (DC-STG) state as described in the main text. It is noted that this 573 K annealing is important for the ultralow modulus after Step-3 processing and this is a lower annealing temperature than the conventional annealing temperatures of Ti-Ni alloys which is usually between 673 K and 773 K.

In Step-3 processing, both plate and wire samples after Step-2 processing were moderately deformed at room temperature by a 12% elongation, which is slightly above the transformation strain of B19’ (~8-9%). This step creates a small fraction of R and B19’ martensite “dual-seeds” out of the DC-STG matrix, and such a state is called “dual-seed strain glass” (DS-STG).

Mechanical property characterization

The stress-strain curves of all samples were measured at room temperature with a Shimadzu AG-IS tensile machine operated at a strain rate of 2 10^-3s^-1. An extensometer (Epsilon Tech. 3442-006M-050M-LHT) was used to ensure accurate strain reading. The Young’s modulus (E) was calculated from the initial slope of the stress-strain curves. The yield stress (σ_y) and critical stress (σ_c) to induce a martensitic transformation were determined by the conventional double-tangent method⁴³. Recoverable strain (ε_re) was determined by the difference between the maximum and minimum strain shown in the unloading curve⁴⁴.

The stress-strain curves of typical structural materials (shown in Fig. 1b) were also tested using the same tensile machine to compare with our DS-STG alloy. These structural materials include metal alloys and organic-based materials: spring steel (65Mn), titanium alloy (Ti-6Al-4V ELI), aluminum alloy (7021-T62), magnesium alloy (AZ91D), bamboo (dry), fiber reinforced polymer (FRP, PPS GF40), polyether ether ketone (PEEK), and polyphenylene sulfide (PPS).

In-situ structural and microstructural characterization

Crystal structure characterization at different temperatures was done with an x-ray diffractometer (Shimadzu XRD-7000) equipped with a temperature sample holder. Cu-Kα radiation was used as the incident x-ray source. In-situ x-ray diffractometry under tensile deformation was done with a self-made tension holder which enables in-situ tension to a small XRD sample with a gauge dimension of 40 2 0.05 mm³.

Microscopic characterization was done with a JEM 2100F TEM equipped with a field-emission gun and a cooling holder. The data were recorded using a GATAN CCD slow scan camera and analyzed by the DigitalMicrograph software. The TEM specimens were made by mechanical thinning down to about 0.1 mm followed by twin-jet polishing with a solution of 90% methanol and 10% sulfuric acid.

Characterization of martensitic transformation and strain glass transition by DSC, DMA, and electrical resistivity

A differential scanning calorimeter (DSC-Q200) from TA Instruments was used to monitor the occurrence of martensitic transformation through monitoring its latent heat peak, and the heating/cooling rate was 10 K/min. As a strain glass transition does not produce a significant DSC peak, a dynamical mechanical analyzer (DMA-Q850) from TA Instruments was used to monitor the key signature of a strain glass transition: a Vogel-Fulcher type frequency-dependent anomaly in the storage modulus curve. The measurement was done in a single cantilever mode with an amplitude of 15 μm at 1, 2, 5, 10, 20 Hz. The cooling/heating rate was 2 K/min. Four-probe electrical resistivity measurement was used to monitor R and B19’ martensitic transformation and strain glass transition under a constant current 100 mA and a cooling/heating rate of 2 K/min.

Reference

43. Niitsu, K., Date, H. & Kainuma, R. Thermal activation of stress-induced martensitic transformation in Ni-rich Ti-Ni alloys. Scr. Mater. 186, 263–267 (2020).

44. Liu, Y., Houver, I., Xiang, H., Bataillard, L. & Miyazaki, S. Strain dependence of pseudoelastic hysteresis of NiTi. Metall. Mater. Trans. A. 30, 1275–1282 (1999).

There is NO Competing Interest.

supplementaryinformation.docx
Table 1 Data source for mechanical properties of typical metal alloys and organic materials used in Fig. 1a.
supplementarymovie1.mp4
Movie 1 A metal alloy showing simultaneously polymer-like ultrahigh flexibility and steel-like ultrahigh strength.
ExtendedDataFigs.docx

Download PDF

Journal Publication

published 04 Sep, 2024

Read the published version in Nature →

Version 1

posted

You are reading this latest preprint version

A polymer-like ultrahigh-strength metal alloy

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Conclusion

Declarations

References

Methods

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1