news & views

destructive, the acoustic intensity decays
exponentially with distance into the crystal,
and the incident wave is perfectly reflected
(an evanescent mode). Importantly, the
nature of the interference can be controlled
by varying the spacing L relative to the
acoustic wavelength λ (Fig. 1c). For certain
values of the ratio L/λ, transmission
through the crystal is near zero (identically
so as N → ∞). These so-called bandgaps
are characteristic of wave propagation in
periodic materials7,8.
Such phononic crystals8–10 can
be used to manipulate sound just as
photonic crystals are used to control
light. Phononic structures are typically
created by introducing periodic variations
in the density and/or elastic constants
of an otherwise homogeneous material
on length scales commensurate with
the wavelength of sound (1 cm to 10 m
in air). These static structures cannot
be tuned between different frequencies
and fail irreversibly when damaged. To
address these limitations, Zhang and
colleagues recognized that sound waves
need not be passively directed by material
structures, but can exert forces capable
of altering these structures. Making use
of this dynamic feedback whereby matter
directs sound and sound directs matter,
they demonstrate how acoustic energy
can be harnessed to direct the assembly of
phononic materials.
The experimental system is remarkably
simple: a handful of particles made from
drinking straws are distributed within
a long tube and exposed to an audibly
unpleasant 3.9 kHz tone from a computer
speaker. Driven by acoustic radiation
forces (Fig. 1a), these subwavelength
scatterers move along the tube and
organize to form stable configurations
in one dimension. As explained by a
quantitative model, each particle in the
final configuration experiences an identical
acoustic force that pushes the assembly
slowly down the tube. Although the steady-
state structures are not periodic, they are
nevertheless characterized by a phononic
bandgap with a band edge located at the
driving wavelength. The spacings between
particles in these ‘pseudo-crystals’ can be
tuned by varying the wavelength of the
source and recover when mechanically
perturbed. As a result, the phononic
bandgap created by these structures is
similarly tunable and healable.
Looking forward, this initial
demonstration motivates the exploration
of related opportunities involving richer
input signals, higher dimensions, and other
types of waves. Even in one dimension,
more complex structures with additional
functionality may be accessible using
multiple driving frequencies to direct order
on different length scales. While one-
dimensional structures on the centimetre
scale are readily fabricated by other
means, the extension of this approach to
two and three dimensions and to shorter
wavelengths (ultrasound, hypersound)
could enable the facile assembly of complex
phononic structures on the 100-μm scale.
Moreover, the analogies between sound and
light waves suggest that photonic materials
are also promising targets for dynamic
self-assembly. Though highly speculative,
it’s fun to imagine the wondrous variety of
‘living’ structures that may emerge when
scattering units are moved by the waves
they make. ❐
Kyle J. M. Bishop is in the Department of Chemical
Engineering at Columbia University, New York,
New York 10027, USA.
e-mail: kyle.bishop@columbia.edu
References
1. Grzybowski, B. A., Stone, H. A. & Whitesides, G. M. Nature
405, 1033–1036 (2000).
2. Fialkowski, M. et al. J. Phys. Chem. B 110, 2482–2496 (2006).
3. Umbanhowar, P. B., Melo, F. & Swinney, H. L. Nature
382, 793–796 (1996).
4. Driscoll, M. et al. Nat. Phys. 13, 375–379 (2016).
5. Bachelard, N. et al. Nat. Mater. 16, 808–813 (2017).
6. Landau, L. D. & Lifshitz, E. M. Fluid Mechanics 2nd edn
(Pergamon, 1987).
7. Markoš, P. & Soukoulis, C. M. Wave Propagation (Princeton Univ.
Press, 2008).
8. Gorishnyy, T., Maldovan, M., Ullal, C. & Thomas, E. Phys. World
18, 24–29 (December, 2005).
9. Lu, M. H., Feng, L. & Chen, Y.-F. Mater. Today 12, 34–42
(October, 2009).
10. Hussein, M. I., Leamy, M. J. & Ruzzene, M. Appl. Mech. Rev.
66, 040802 (2014).

MARAGING STEELS

Making steel strong and cheap

Introducing high-density ordered nanoprecipitates into martensitic steel increases strength at modest cost.

J. W. Morris Jr

T
here are two certain things
regarding the metallurgy of steel:
new alloys that combine improved
properties with lower cost are always
highly demanded; and, despite 3,000 years
of research in the art and science of the
subject, many issues are still unresolved,
leaving space for new ideas. In a work
recently published in Nature, Zhaoping Lu
and co-workers1 have taken the well-
established concept of ultrastrong maraging
steels, simplified the alloy content and the
nature of the hardening precipitates, and
shown that it is still possible to achieve
very promising mechanical properties,
arguably opening a new pathway to the
development of structural steels with
exceptional combinations of strength,
ductility and cost.
Ultrahigh strength steels (UHSSs),
which exhibit strengths above 1,500 MPa,
are used in the most challenging structural
applications, such as aircraft landing gear,
rocket cases, high-performance shafts and
tubes, high-strength fasteners, and others.
Strength is not the only critical issue; it
must be combined with toughness to resist
fracture, ductility for forming and for
margin of safety in service, and resistance to
cyclic fatigue and environmental cracking.
Of course, hovering in the background is the
cost of the manufactured product, which is
often the determining factor in the ultimate
materials selection.
Iron (steel) has a face-centred cubic
(fcc) crystal structure at high temperature,
but is body-centred cubic (bcc) at low
temperature. If it is cooled quickly
(quenched), a martensitic transformation
occurs through the spontaneous shear
of the parent fcc into the product bcc
structure, creating a microstructure with
a high density of internal defects. If these
defects are crystal dislocations (dislocated
lath martensite), the steel ordinarily has
good strength and ductility in the as-
quenched condition. In fact, roughly
half of the strength of an UHSS can be

NATURE MATERIALS | VOL 16 | AUGUST 2017 | www.nature.com/naturematerials 787

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
news & views

Ti Ni Fe can cut through them, shearing them

Metal Ni
and leaving a crystallographic fault
^
^ behind (Fig. 2). As the precipitates grow,
^ Al they increasingly resist shear, eventually
C reaching a size such that the dislocations
loop around them (the Orowon
^ mechanism8) with dislocation arms
^ ^ merging on the far side of the precipitate to
M2C Ni3Ti NiAl α-iron recreate the dislocation line. Alloy strength
starts to decrease with further coarsening
Alloy Precipitate Yield strength Tensile strength Total elongation Fracture toughness
(over-ageing).
AF14104 M2C 1527 MPa 1699 MPa 18% 128 MPa m1/2
To maximize alloy strength, we need
AerMet 1005 M2C 1778 MPa 2081 MPa 14% 99 MPa m1/2 to maximize obstacle strength β and
Maraging C3005 Ni3Ti 1943 MPa 2005 MPa 8.6% 68 MPa m1/2 minimize obstacle spacing ls (ref. 8).
Maraging T3006 Ni3Ti 1963 MPa 2032 MPa 10.5% 74 MPa m1/2 The most effective way to increase β is
PH 13–87 NiAl 1413 MPa 1482 MPa 13% 127 MPa m1/2 either to increase the local strain field
Ref. 1 NiAl 1947 MPa 2197 MPa 8.2% — repelling the dislocations or to increase
the energy of the fault created when a
dislocation passes through. The strain
Figure 1 | The crystal structures of hexagonal α-M2C (alloy carbide), hexagonal η-Ni3Ti (DO24),
field is created by lattice mismatch and
cubic NiAl (B2), and body-centred cubic α-iron. The table shows example mechanical properties of
a difference in elastic modulus between
conventional ultrahigh strength steels as well as the values from the work under discussion.
precipitates and the matrix. A large
lattice misfit increases obstacle strength,
but may decrease obstacle density since
attributed to the as-cooled microstructure, differ in the nature of the precipitates the associated strain energy lowers
which is hardened by the high dislocation used: the secondary hardening steels use the equilibrium volume fraction of
density and the solute species (solution alloy carbides such as CrC2 and MoC2 precipitates. Moreover, when the lattice
hardening)2,3. When the quenched (M2C), while maraging steels employ mismatch is large (as it is for both the
supersaturated solid solution is reheated to intermetallic precipitates such as Ni3Ti M2C precipitates in secondary hardening
a certain temperature, precipitates nucleate and NiAl. Figure 1 shows the structures of steels and the dominant Ni3Ti precipitates
and grow. the precipitates and example mechanical in maraging steels), precipitates tend to
Two basic classes of steels have properties of selected steels. nucleate heterogeneously on dislocations,
survived the multi-property selection The basic features of precipitation limiting their ability to fill the volume and
process to become key competitors for hardening are common to all UHSSs8. minimize spacing.
UHSS applications: secondary hardening Since the surface energy dominates The precipitate of interest in the present
steels such as AF14104 and AerMet 1005, the free energy of nanoparticles, work is the NiAl, whose properties are
and maraging steels such as C3005 (high nuclei and nanoprecipitates are always qualitatively different from those of M2C
cobalt), T3006 (cobalt-free) and PH 13–87 crystallographically coherent with the (refs 4,5,9,10) and Ni3Ti (refs 5,6,11). M2C
(high-strength stainless steel). All of parent matrix in their earliest stages, even and Ni3Ti have hexagonal lattices (Fig. 1),
them have significant alloy additions, if significant strain is needed to match and the misfit strain plays a major role in
are processed to have a dislocated the matrix lattice. Once formed, the hardening. But NiAl is a simple binary
martensitic microstructure, and are nanoprecipitates grow in size (coarsen) to ordering of the same bcc lattice as the
hardened by an internal distribution of lower the overall surface energy. When the matrix. The only lattice mismatch is from
nanosized precipitates. However, they precipitates are very small, dislocations the difference in the edge length. Despite
small misfit strain, NiAl is known to be
a b an efficient hardening precipitate12. In
their work, Lu and co-workers achieved
much higher strengths by increasing the
F = 2Tcos(Ψ/2) = 2Tβ Al and Ni contents. In their 18Ni–3Al
F T T
alloy, the lattice misfit is reduced by a
partial substitution of Al by Fe atoms
T T
Ψ/2 (Ni(Fe,Al) precipitate) and almost
γd vanishes. The strength of the precipitate
R comes, rather, from the exceptionally high
energy (0.5 J m–2) of the planar antiphase
boundary faults (Fig. 2b), created when
a lattice dislocation shears through the
Figure 2 | Idealized dislocation pinned by a distribution of precipitates. a, The dislocation (black line) precipitate and disturbs its chemical order.
bows out between obstacles (blue spheres) into a circular arc with a radius R (R = T/τb, where τ is the As the authors point out, the small or
applied shear stress, b is the magnitude of the Burgers vector of the dislocation, and T is its line tension). vanishing misfit strain also facilitates the
β is the obstacle strength (dimensionless). The two arms of the dislocation at an obstacle pull on it with homogeneous nucleation of the Ni(Fe,Al)
the force F = 2Tβ, and cut through, or loop around, when β reaches the critical value βc. b, Force balance phase throughout the volume, creating
at maximum strength for an ordered precipitate particle in the plane, where d is the diameter of the a very dense distribution of hardening
precipitate and γ is the antiphase boundary energy. precipitates. As a result, their 18Ni–3Al

788 NATURE MATERIALS | VOL 16 | AUGUST 2017 | www.nature.com/naturematerials

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 news & views

alloy has a yield strength near 1,950 MPa,
and a tensile strength near 2,200 MPa, with
total elongation of 8.2% and a uniform
elongation of 3.8%. The combination of
strength and ductility is competitive with
the best reported for both maraging and
secondary hardening steels (Fig. 1).
Finally, the results presented by Lu
and colleagues document that Ni(Fe,Al)
can be a very effective hardening
precipitate, leading to UHSSs that are also
economically attractive since they avoid
expensive species such as Co and Ti. But
there are a number of other issues that
must be addressed before these alloys
become viable candidates for important
structural applications. An important
example is the fracture toughness of the
MATERIAL WITNESS
alloy, which is not reported. There is reason
for concern, since alloys strengthened by
NiAl precipitates, such as PH 13–8, are
susceptible to cleavage fracture at ambient
temperature12 and lose toughness rapidly
with strength7. However, brittle cleavage
fracture is not overcome metallurgically
by changing the precipitate distribution,
but rather by modifying the underlying
martensitic microstructure, as is done
effectively, for example, in the AerMet 100
alloy 13. A similar approach could be useful
in the 18–3 alloy discussed here. ❐ 10. Hu, Z. & Wu, X. Micron 34, 19–23 (2003).
J. W. Morris Jr is in the Department of Materials
Science and Engineering, University of California,
Berkeley, California 94720, USA.
e-mail: jwmorris@berkeley.edu
References
1. Jiang, S. et al. Nature 544, 460–464 (2017).
2. Kinney, C. C, Pytlewski, K. R., Khachaturyan, A. G. & Morris,
J. W. Jr Acta Mater. 69, 372–385 (2014).
3. Galindo-Nava, E. I., Rainforth, W. N., Rivera-Díaz-del-Castillo,
P. E. J. Acta Mater. 117, 270–285 (2016).
4. Handerhan, K. J., Garrison, W. M. & Moody, N. R. Met. Trans. A
20, 105–123 (1989).
5. Novotny, P. M. & Maurer, G. E. Adv. Mater. Process.
165, 37–40 (2007).
6. Hickey, C. F. Jr, Dix, D. W. & Kagan, D. Mechanical Property
Characterization of Vascomax T-300 (US Army Materials
Technology Laboratory, 1989).
7. Wert, D. E. & DiSabella, R. P. Adv. Mater. Process. 164, 34–36 (2006).
8. Glazer, J. & Morris, J. W. Jr Acta Metall. 36, 907–915 (1988).
9. Ayer, R. & Machmeier, P. M. Metall. Trans. A 24, 1943–1955 (1993).
11. Vasudevan, V. K., Kim, S. J. & Wayman, C. M. Metall. Trans. A
21, 2655–2668 (1990).
12. Leitner, H., Schober, M., Schnitzer, R. & Zinner, S. Mat. Sci. Eng. A
528, 5264–5270 (2011).
13. Sato, K. Improving the Toughness of Ultrahigh Strength Steel PhD
thesis, Univ. California, Berkeley (2002).

SEEKING COMFORT IN THE IRON AGE

The sophistication of Viking material
culture, such as the fine quality of
their decorative metalwork, has
long given the lie to the old idea
that they were barbaric brutes — a
reputation that attests to the trauma
left in Europe by the Viking raids of
the ninth to the eleventh centuries.
There have even been suggestions that
Vikings used advanced navigation
techniques based on the ability of
certain local minerals — such as
calcite or Iceland spar — to influence
polarized light1,2, although there is no
archaeological evidence to support
the idea.
But a new discovery must surely
sink the image of unreconstructed
brutishness once and for all: Vikings
preferred soft underwear. They made
their underclothing not from coarse
hemp, which the Vikings used for
sailcloth and wall hangings3, but from
finer flax.
Along with wool and nettle, hemp
and flax were the only choices available
for textile manufacture in Scandinavia
in the late Iron Age of the sixth to
the eleventh centuries. Cotton could
theoretically have been acquired from
late Roman culture, but there is no
evidence of it. Wool, needless to say,
would be rather itchy for clothing
worn against the skin, but the other
options all seem viable in principle.
Lukešová and colleagues4 have
investigated the fibres in ten textile
items — nearly all of them from
the tenth century — collected from
excavations of Viking graves in
various Norwegian provinces and
kept at the University Museum of
Bergen. All are made from plant
fibres and, with the exception of a
purse, all are items of clothing. One
appears to be a man’s shirt, while the
others are women’s clothing, mostly
shifts worn as underwear. The sex of
the wearer could be identified from
other items in the respective graves,
which indicate also that the bodies
were high-ranking members of
Viking society.
Plant fibres can be distinguished by
polarized-light microscopy. The fibres
contain bundles of cellulose chains
called fibrils, which wind helically
around the cell walls in either a left-
handed (S) or right-handed (Z) twist
(the same terminology is used for the
twisting of macroscopic strands in
textile threads). Flax and nettle have
an S twist, hemp a Z twist, and these
different fibre orientations show up as
differences in the colour of the light
transmitted through crossed-polar
filters under the microscope: a so-
called modified Herzog test. Nettle
and flax can usually be differentiated,
meanwhile, by the telltale presence of
calcium oxalate crystals in the former.
With the exception of one
ambiguous sample, all of those studied
were made of flax. It is easier to make
softer fabrics from this plant, since
on average its fibres are finer — to
PHILIP BALL
achieve the same qualities with hemp,
one would need to sort the fibres to
remove the coarser strands. The same
is true of nettle, which has the added
drawback that it can’t be cultivated as a
textile crop but must be collected from
the wild.
As ever with such archaeological
studies, the findings shouldn’t be
generalized too readily. And it
remains unclear whether the fibre
choice was aesthetic or simply
dictated by local availability. All the
same, when it comes to underwear,
given the choice we would probably
have the same preference as the
Viking noblewomen. ❐
References
1. Ropars, G., Gorre, G., Le Floch, A., Enoch, J.
& Lakshminarayanan, V. Proc. R. Soc. A
468, 671–684 (2011).
2. Ropars, G., Lakshminarayanan, V. & Le Floch, A.
Contemp. Phys. 55, 302–317 (2014).
3. Skoglund, G., Nockert, M. & Holst, B. Sci. Rep.
3, 2686 (2013).
4. Lukešová, H., Palau, A. S. & Holst, B. J. Archaeol. Sci. Rep.
13, 281–285 (2017).

NATURE MATERIALS | VOL 16 | AUGUST 2017 | www.nature.com/naturematerials 789

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.