1
The invention relates to the field of the treatment of metal surfaces, in particular to
the treatment by peening.
5
Shot peening is a technique that is widely used for improving certain properties of
metal surfaces, such as the fatigue life. A shot-peening treatment is typically characterized
by a degree of coverage that describes the proportion of the surface impacted by the peening
and an intensity that describes the amount of kinetic energy applied per unit area. The
10 literature in the field of shot peening prescribes limits for the degree of coverage and
intensity parameters, beyond which the peened material incurs degradation such as cracking
and reduction of the fatigue life. The conditions giving rise to these degradations are
commonly denoted by the term "overpeening".
Subsequent research into peening treatments has highlighted the possibility of
15 producing a nanostructuring of the material at a surface layer by pushing the peening
treatment beyond the limits normally prescribed. The term "nanostructuring" denotes the
obtaining of a stable phase, the grain size of which is of the order of a few tens of
nanometres. Under certain conditions, it is assumed that the nanostructuring of the material
prevents the propagation of microcracks, so that the aforementioned degradations do not
20 occur.
The nanostructuring of the material produces advantageous effects such as the
increase of the fatigue life, of the hardness, of the corrosion resistance, of the atomic
diffusivity, of the biocompatibility, the improvement of the tribological properties, etc.
Among the processes known for producing a nanostructured surface layer, note may
25 essentially be taken of:
• techniques for projecting fine or very fine particles at high or very high speed (Fine Particle
Bombarding or Air Blast Shot Peening) at normal incidence onto the sample to be treated.
These techniques are especially presented in the publication of the Iron and Steel Institute of
Japan, ISIJ International, Vol. 47 (2007), No. 1, pp 157-162;
30 • techniques for vibrating larger particles at lower speeds, known as UltraSonic Shot Peening
or Surface Mechanical Attrition Treatment. These techniques are especially presented in the
publication of the Japan Institute of Metals Materials Transactions, Vol. 45, No. 2 (2004),
pp 376-379.
WO02/10461 describes a process for generating nanostructures at the surface of a
35 metal part in which perfectly spherical balls similar to ball bearing balls are projected onto a
2
point of impact of the part under variable incidences. In order to obtain a thickness of
nanostructures of a few tens to a few hundreds of microns, it is taught to mechanically and/or
thermally stress the surface of the metal part to be treated.
WO 02/10462 describes a process for generating nanostructures in which balls are
5 projected onto a point of impact of a part along different and varied directions of incidence
by a ball projection source in order to create deformations having any direction. A layer
thickness of 10 urn is obtained with balls having a diameter of 300 um and a layer thickness
of 20 um is obtained with balls having a diameter of 3 mm.
WO 02/10463 describes a process for generating nanostructures in which ball
10 motion is triggered by the combination of a circular motion of a chamber containing the balls
and a vibrating motion along a direction perpendicular to the plane of the circular motion of
the chamber. A nanostructured layer thickness of 10 um is obtained with balls having a
diameter of 300 um and a layer thickness of 20 um is obtained with balls having a diameter
of 3 mm.
15 EP1577401 describes vibrating rods which produce impacts on a material. The
maximum impact speeds are equal to 3.6 m/s.
In a first embodiment, JP2003201549 teaches how to project a stream of particles
onto a metal part along a normal incidence. In another embodiment, the document teaches
how to generate a vibrating motion in order to produce projections, which involves relatively
20 small projection speeds.
According to one embodiment, the invention provides a process for the surface
treatment of a metal part, comprising:
exposing a surface of the metal part to a stream of substantially spherical particles, so that
25 any portion of said surface receives said particles along several primary incidences, the
primary incidences of the particles on a portion of the surface being essentially distributed in
a cone or a conical film which has an outer half apex angle between 10° and 45°, until a
surface layer of nanostructures, having in particular an average thickness of greater than
50 um is obtained,
30 the particles having a diameter of less than 2 mm and greater than 0.1 mm and being
projected at a speed between 40 m/s and 100 m/s.
One idea at the heart of the invention is to create one or more streams of particles
capable of hitting a surface to be treated along varied and controlled incidences in order to
stress a large number of atomic slip planes of the material. According to one embodiment,
35 the surface layer of nanostructures has an average thickness of greater than 50 um, the
3
boundary of the surface layer of nanostructures being determined to be a region of the metal
part where the hardness is greater than a threshold that is dependent on the metal material
from which the part is made.
According to one embodiment, said hardness threshold is defined by a hardening of
5 the material with respect to a prior art upon surface treatment which is equal to 50% of the
hardening obtained at the treated surface of the metal part. In other cases this threshold may
be defined as a function of other parameters, especially the position of a crystalline phase
transition in the material when such a transition takes place.
According to other advantageous embodiments, such a process may have one or
10 more of the following features.
According to one embodiment, the particles have a diameter of greater than
0.3 mm and less than 1.4 mm.
According to one embodiment, the incidences of the particles are distributed
substantially continuously in the cone or the conical film.
15 In one embodiment, the cone or the conical film has an outer half apex angle of
between 10° and 30°.
According to one embodiment, the stream of particles comprises a jet of particles
projected along a central direction, the metal part being fixed to a support so as to present
said surface oriented obliquely with respect to said central direction, the support being
20 rotated about an axis coaxial with the central direction of the jet of particles.
According to one embodiment, the inclination of the surface of the part with
respect to the central direction is between 10° and 30°, preferably close to 15°.
According to one embodiment, the particles are projected at a speed of between 50
and 80 m/s.
25 According to one embodiment, the particles have a hardness greater than the
hardness of the surface of the part before treatment.
According to one embodiment, the invention thus provides a metal part comprising
a surface treated by the aforementioned process, said surface comprising a surface layer of
nanostructures having an average thickness of greater than 50 um, the boundary of the
30 surface layer of nanostructures being determined to be a region of the metal part where the
hardness is greater than a threshold that is dependent on the metal material from which the
part is made.
4
According to one embodiment, said hardness threshold is defined by a hardening of
the material with respect to a prior art upon surface treatment which is equal to 50% of the
hardening obtained at the treated surface of the metal part.
According to one embodiment, the surface layer of nanostructures has an average
5 thickness of greater than 100 um.
According to one embodiment, the invention also provides a surface treatment
device for a metal part, comprising:
a projection means capable of producing a stream of substantially spherical particles having
a diameter of less than 2 mm and greater than 0.1 mm and thus are projected at a speed of
10 between 40 m/s and 100 m/s,
a support capable of holding a metal part, the support comprising a surface exposed to the
stream of particles, and
an actuator capable of modifying an orientation of the support with respect to the stream of
particles so that the primary incidences of the particles on a surface of the support are
15 essentially distributed in a cone or a conical film that has an outer half apex angle of between
10° and 45°.
According to one embodiment, the projection means is capable of producing a jet
of particles projected along a central direction, the surface of the support being oriented
obliquely with respect to said central direction, the actuator being capable of pivoting the
20 support about an axis that is coaxial with the central direction of the jet of particles.
Certain aspects of the invention are based on the idea of designing a process for
nanostructuring the material which has a high productivity in order to produce relatively
thick nanostructured surface layers in a relatively short time. Certain aspects of the invention
are based on the idea of producing relatively homogeneous nanostructured surface layers.
25 ^Certain aspects of the invention are based on the idea of designing a process for
nanostructuring the material which can be applied to varied geometries, in particular concave
shapes. Certain aspects of the invention are based on the idea of designing a process for
nanostructuring the material which is relatively easy and economical to implement.
The invention will be better understood, and other objectives, details, features and
30 advantages thereof will become more clearly apparent in the course of the following
description of several particular embodiments of the invention, given solely by way of
illustration and nonlimitingly, with reference to the appended drawings.
In these drawings:
5
• Figure 1 is a schematic representation of a process for nanostructuring a metal
surface.
• Figure 2 is a schematic perspective view of a peening machine suitable for
implementing the processes according to the embodiments of the invention.
5 • Figure 3 is a schematic representation of a particle jet produced by the machine
from Figure 2.
• Figure 4 is a diagram of the operation of the machine from Figure 2.
» Figure 5 is a graph representing the change in the hardness of a metal part as a
function of the depth below the treated surface, for several peening conditions.
10 • Figure 6 is a graph representing the change in the thickness of a nanostructured
surface layer as a function of the degree of coverage for the peening conditions from
Figure 5.
• Figure 7 is a graph representing the change in the treatment time as a function
of the degree of coverage for several shot sizes.
15 • Figure 8 is a graph representing the change in the surface hardness and in the
thickness of a nanostructured surface layer as a function of the degree of coverage for a
peening condition.
• Figure 9 is a graph representing the change in the thickness of a nanostructured
surface layer as a function of the degree of coverage for various modes of attachment of the
20 treated part.
• Figure 10 is a graph representing the change in the thickness of a nanostructured
surface layer as a function of the inclination of a support in the machine from Figure 2, for
several peening conditions.
• Figure 11 is a graph representing the change in the hardness of parts made of
25 various metal materials as a function of the depth below the treated surface.
• Figure 12 is a graph representing the change in the thickness of a nanostructured
surface layer as a function of the degree of coverage for various metal materials.
• Figures 13 and 14 are graphs representing the change in the surface hardness
and in the thickness of a nanostructured surface layer as a function of the degree of coverage
30 for two different rates of projection.
• Figure 15 is a graph representing the change in the surface hardness and in the
thickness of a nanostructured surface layer as a function of the degree of coverage for
another peening condition.
6
• Figure 16 is a schematic cross-sectional representation of a part treated by a
peening process representing the region of influence of an impact.
• Figures 17 to 20 are optical micrographs of nanostructured surface layers.
• Figure 21 is a graph representing the change in the hardness of a metal part as a
5 function of the depth below the treated surface, for several peening conditions with another
hardness measurement method.
• Figure 22 is a schematic cross-sectional representation of a metal part having a
nanostructured surface layer as a function of the depth below the treated surface on which
the measured hardness curve is superposed.
10 • Figure 23 is a graph representing the change in the surface hardness of a part
treated by peening and the change in the thickness of a nanostructured surface layer as a
function of the degree of coverage.
Described below are embodiments of peening processes that make it possible to
15 obtain a nanostructured surface layer on a metal part. Unless otherwise indicated, the
experimental results presented below are obtained with flat metal samples.
With reference to Figure 1, a process for nanostructuring a metal surface 1 is
schematically represented. In the left-hand view, before treatment, the size of the grains 2 of
the material all the way to the surface 1 is typically a few tens to a few hundreds of um. In
20 the right-hand view, after treatment, the grain size of the material at a surface layer 3 is
reduced to a few tens of nm, for example around 20 run, whilst grains of larger size continue
to exist more deeply in the material. Subsequently, an axis z perpendicular to the surface 1
and oriented towards the inside of the material starting from the surface is defined. The
surface serves as a reference of the dimensions. The transition of the size of the grains
25 between the surface layer 3 and the unmodified deep material is in reality more gradual than
in the drawing.
The nanostructuring of the material in the layer 3 is stable up to a temperature of at
least 600°C. A metal part coated with such a nanostructured layer may be used in various
industries, for example in applications where the wear resistance and the fatigue resistance
30 are critical properties.
With reference to Figure 2, a peening machine 10 which may be used to produce
such a nanostructured layer is now described.
The machine 10 comprises a projection nozzle 11 supplied from a shot reservoir
and from an air compressor (which are not represented) in order to produce a jet of shot
7
projected at a speed V which may vary depending on the size of the shot particles. As a
variant, the projection of the shot particles may also be carried out using a vane turbine,
according to the known art. Common peening equipment makes it possible to obtain speeds
, ranging from 20 m/s to around 120 m/s.
5 The shot used preferably consists of particles obtained by atomization. Such
particles may be produced in a large amount at a relatively advantageous cost and have quite
good sphericity, for example greater than or equal to 85%. Their cost is substantially lower
than that of ball bearing balls, the process for the manufacture of which is virtually unitary in
order to achieve a sphericity of greater than 99%.
10 Alternatively, other conventional peening media can be used, such as conditioned
cut wire, glass beads or ceramic beads.
The projection nozzle 11 is fixed facing a mobile support device 12 constructed in
the following manner: a metal disk 13 is mounted on the shaft of a rotary motor that is not
represented, for example an electric motor, in order to be able to pivot with respect to a fixed
15 frame 19. The central pivoting axis of the disk 13 is coaxial with a central projection axis of
the nozzle 11. Positioned on the disk 13 is an inclinable support 14, the angle of inclination
of which with respect to the disk 13 can be adjusted by means of a screw. Fastened around a
central portion of the inclinable support 14 are fastening clamps 15 provided with screws 16
parallel to the support 14. The screws 16 may be tightened onto a part to be treated in order
20 to fasten the part between the clamps 15 and may be loosened in order to withdraw the part
after treatment.
With reference to Figure 3, a jet of particles 20 produced by the projection nozzle
11 is schematically represented. The jet 20 has an approximately conical shape with a half
apex angle p. The angle P may be measured, for example, as the ratio between the radius p of
25 an impacted region 21 and the distance L from the region 21 to the orifice 22 of the nozzle
11-
With reference to Figure 4, the operating principle of the peening machine 10 is
now described. For a flat sample parallel to the support 14, the surface portion located
around the central axis 25 of the jet 20 receives the particles at an angle of incidence a with
'".' 30 respect to the local normal direction 26. The surface portion located around an edge of the jet
20 receives the particles at an angle of incidence (a-P) with respect to the local normal
direction 27. The surface portion located around the opposite edge of the jet 20 receives the
particles at an angle of incidence (a+P) with respect to the local normal direction 28.
8
When the support device 12 turns during the projection of the particles, any portion
of the sample located in the jet 20 is hit at incidences located in a more or less wide conical
film. This conical film is thin towards the centre of the jet where it coincides exactly with the
angle a and broader towards the periphery of the jet, where it includes all the angles between
5 (a-P) and (a+P). If p > cc, the conical film degenerates into a cone. During the rotation of the
support device 12, a treated surface region may be hit at all the angle of elevation values
located in the conical film. This property of the machine 10 makes it possible to produce
nanostructured layers on different metals with a relatively high productivity, as will be
recounted in the tests below. In the tests below, the angle p is equal to around 8° and the
10 distance L to around 300 mm. Of course, it is not excluded for a small portion of the
particles to be projected along atypical trajectories outside of the main directions of the jet
20.
The tests which will be described below were carried out with various types of
shot, the main properties of which are mentioned in Table 1, according to the SAE J444
15 standard. The nominal diameter of a type of shot is defined as the median diameter of the
distribution: 50% by weight of the particles of the type of shot considered have a diameter of
less than the nominal diameter, and 50% have a larger diameter.
Table 1: properties of the shots according to the SAE J444 standard
Nominal
Type of diameter
shot (D) Distribution (fraction of the particles of larger size than)
S550 1.40 0 >85 >96
S330 0.85 °___S5 >85 >96
S280 0.71 °__f§ >85 >96
>85 >97
S170 0.425 0 <10
o„™ „-,„ 0 <10 >80 >90
S070 0.18
.425 0.35 0.30 0.18 .125
mm J 2.00 | 1.70 | 1.40 | 1.18 | 1.00 [ 0.85 | 0.71 | 0.60 | 0.50 I I 1 I I
20
Test 1
Table 2 recounts the results of a first test carried out with the machine 10 on flat
samples of E24 steel (low-alloy steel: 0.2% C, 1.5%Mn, 98.2% Fe) fastened by clamping to
the support 14 with an inclination a = 15°. Recorded in this table are the type of shot used in
25 the test, the projection speed V, the degree of coverage R, the thickness of the nanostructured
layer Zn obtained, the Vickers hardness of the sample on its face exposed to the peening, the
9
Vickers hardness of the sample on its opposite face, and the ratio between the two
hardnesses, known as the hardness gain.
The degree of coverage R is a measurement of the proportion of the surface
impacted by the peening. In the present description, it is defined as follows: the reference
5 100% indicates that an amount of shot which is statistically sufficient to impact 98% of the
exposed surface was projected. Beyond 100%, a linear law is applied with respect to this
reference amount. A degree of coverage of 1000% therefore indicates that ten times the
reference amount has been projected. At constant flow rate, the degree of coverage is
therefore also a measurement of the treatment time of the sample.
10 The thickness of the nanostructured layer zn was obtained by two methods:
observation by optical microscopy and observation of the hardness profile of the material as
a function of the depth z.
Via optical microscopy, the thickness measured is an arithmetic mean of nine
observations of the thickness of the visually amorphous layer corresponding to the
15 nanostructured region 3. The width of the sample treated is scanned over three regions and
three measurements are taken per region, which ensures the reproducibility of the
measurement method.
The microscope observations are then correlated to hardness profiles, in order to
confirm that the visually amorphous region observed indeed corresponds to the peak of
20 hardness originating from the hardening by the effect of the nanometre-sized grains.
The method used for producing the hardness profile consists in making an
indentation line with a step of 50 urn starting from the outermost surface with a micro
Vickers hardness tester having a pyramidal tip with a load of 100 g (HV 0.1) which
possesses a lens. The surface of the sample and the nanostructured layer are visualized as in
25 optical microscopy. The hardness profile is thus obtained from a depth of 50 |j,m to 500 \im.
The values communicated are an average of three indentation lines in order to have a reliable
and reproducible measurement.
The connection between the hardness profile and the thickness z^ may be explained
more precisely with the aid of Figure 5. Figure 5 represents the hardness profiles obtained by
30 the method explained above in test 1 samples with R=3000%. The curve 30 corresponds to
the type S170 shot. The curve 31 corresponds to the type S280 shot. The curve 32
corresponds to the type S330 shot. The curve 33 corresponds to the type S550 shot. On all
the curves 30 to 33, a region of very high hardness 34 appears, which corresponds to the
nanostructured layer 3 and a second region 35 appears where the hardness decreases more
10
gradually with the depth and which corresponds to the strain hardening of the material. The
boundary of the nanostructured layer 3 must therefore correspond to a steep change of slope
of the hardness. This point is verified in Figure 5 where the thicknesses z^ obtained by visual
observation have been plotted as a dot-and-dash line for each type of shot.
5 More specifically, for the four types of shot tested in Figure 5, the boundary of the
nanostructured layer 3 observed visually corresponds substantially to the region in which the
hardness is equal to the median value between the hardness value at the surface, which is
here represented by the first measurement point at a depth of 50 p.m, and the hardness value
far from the surface, where the material has not been substantially affected by the peening,
10 which is represented by the last measurement point at 500 |j.m.
A quantitative definition of the nanostructured layer 3 may therefore be provided as
a function of the hardness curve: the nanostructured layer 3 is the region in which the
hardening of the material produced by the peening treatment is greater than or equal to 50%
of the maximum hardening obtained at the surface of the sample. This empirical definition
15 has been verified experimentally for the degrees of coverage of greater than 750%, as will be
explained below.
Figure 6 represents the change in the thickness zn observed visually as a function of
the peening treatment time, measured by the degree of coverage R, by the four types of shot.
Curve 36 corresponds to type S170 shot. Curve 37 corresponds to type S280 shot. Curve 38
20 corresponds to type S330 shot. Curve 39 corresponds to type S550 shot. Curves 36 to 39
demonstrate a detection threshold of the nanostructured layer 3 and a saturation threshold of
its thickness. In particular, it is seen that the thickness no longer changes significantly
beyond the R=3000% within the context of test 1.
Figure 6 demonstrates that all the shots from test 1 make it possible to obtain a
25 thickness z^ that exceeds 100 urn, or even 140 um. This figure also demonstrates two
advantages of the type S280 and S330 shots (curves 37 and 38). On the one hand, the
nanostructured layer 3 appears significantly at a lower degree of coverage R, around 300%,
than with the larger particles (S550) or smaller particles (SI70). On the other hand, the
thickness Zn reaches its peak at a higher level than that obtained with larger particles (S550)
30 or smaller particles (SI70).
Not obtaining a maximum thickness ^ with the largest particles (S550, curve 39)
may be considered surprising. This observation can however be explained by the competition
effect that exists, when the size of the projectiles increases, between on the one hand the
increase in the kinetic energy per particle, which involves a deeper and more intense plastic
11
deformation of the material at each particle impact, and on the other hand the increase of the
mean spacing between the impacts, which involves a less even spatial distribution of the
impacts.
This competition is illustrated schematically in Figure 16, where the region of
5 influence of an impact, also referred to as the nanocrystallization lobe, is represented by a
semisphere. Whereas close impacts produce a thickness Zn that is relatively uniform over the
entire treated surface, impacts that are relatively spaced apart give rise to edge regions where
the material is deformed over a relatively small thickness z0 and central zones where the
material is deformed over a relatively large thickness zx. The thickness z„ that can be
10 observed lies between z0 and zi.
Another property on which the size of the particles has an observable effect is the
uniformity of the thickness ^ along the treated surface. This property may be characterized
by the standard deviation I of the thickness z„. Table 3 recounts the values measured in the
samples from test 1, micrographs of which are reproduced in Figures 17 to 19. For the
15 chosen degree of coverage, it appears that the largest type S550 shot provides a mean
thickness ZQ comparable to the thickness obtained with the type S330, but a doubling of the
standard deviation £. Figures 17 to 19 also make it possible to observe nanocrystallization
lobes.
Table 3: Standard deviation of the nanostructured thickness in test 1
Nano
Vickers thickness
Degree of hardness at Nano standard
coverage R the surface thickness ZQ deviation 6
Fig. Type of shot (%) (HV) (urn) (urn)
17 S170 1000 263 72.05 11.1
18 S280 1000 290_ 119.7 12^
19 S330 1000 290 159.76 ' 19.6
20 S550 1000 292J 175.5 [ 40_
20
Moreover, depending on the nature of the projection nozzle 11, the time needed to
obtain a given degree of coverage may increase with the size of the particles. Figure 7
represents, for a conventional peening nozzle model, the change in the degree of coverage R
with the projection time t for two different particle sizes, all conditions being otherwise
25 equal. Curve 40 relates to type S550 and curve 41 to type S280. In test 1, in order to form a
thickness of 100 um, 107 s are needed with type S550 versus 30 s with type S330 and 75 s
with type S280. It is therefore seen that the optimal type of shot in terms of productivity, that
12
is to say that produces the greatest nanostructured thickness per unit time, lies below the
S550 particle size.
Test 1 therefore shows that counter-productive effects of the large particles begin
to arise with the type S550 shot and that it is not advantageous to use even larger sizes.
5 Figure 8 demonstrates the relationship between the nanostructured thickness z,, and
the hardening observed at the surface of the treated sample. Curve 42 represents the
thickness Zn (left-hand axis) and the curve 43 the Vickers hardness at the surface (right-hand
axis) as a function of the coverage R for type S280 in test 1. Curve 43 demonstrates a strainhardening
effect which causes a first increase in hardness in a region 45 starting from the
10 initial hardness 44 without however forming nanometre-size grains, and an effect of the
nanostructuring of the material which causes a second increase in the hardness in a region
46.
Test 2
In order to evaluate the optional effect of clamping the part by the clamps 15 in test
15 1, a test 2 was carried out with the type S280 shot under conditions similar to test 1 by
adhesively bonding the sample to the support 14 without applying any clamping stress
thereto. Figure 9 shows the change in the thickness Zn as a function of the coverage R in test
2 (square symbols) superposed on curve 42 from test 1. No significant difference emerges
between the results of the two tests, neither in the thickness measurements, nor in the
20 hardness measurements, which means that the fastening of the part by clamping in test 1 has
no causal relationship with the nanostructuring effects observed.
Test 3
In order to evaluate the effect of orienting the part to be treated with respect to the
jet of shot, a test 3 was carried out with the type S170, S280 and S330 shots under conditions
25 similar to test 1 by varying the angle a between 0° and 45° and the rotation of the support
device 12. The nanostructured thicknesses obtained in this test 3 are recorded in Table 4.
Table 4: nanostructured thickness in test 3 for R = 3000%
Rotation a (") 0 15 30 45
Without 0 97 • 92 72
With S170 0 130 102 105
Without 131 156 134 153 ZnU, im).
With S280 133 189 171 160
Without 111 168 134 144
With I S330 [ 1 1 6 I 236 I 183 I 125 |
13
It is observed for each type of shot that the thickness z^ varies with the angle a in
order to reach a peak very clearly at around a = 15°. It is also observed that the rotation of
the support does not produce any significant effect for a = 0° but substantially increases the
thickness z^ when the support is inclined. These observations show that the production of
5 impacts of the particles at incidences varied at any point of the treated surface substantially
increases the productivity of the nanostructuring process. In particular, these instances are
distributed in a cone or a conical film which has an outer half apex angle of between around
10° and 45° within the context of this test.
Figure 10 graphically represents the results from Table 4 with rotation of the
10 support. Curve 50 corresponds to type SI70 shot. Curve 51 corresponds to type S280 shot.
Curve 52 corresponds to type S3 30 shot.
In order to evaluate the effect of the nature of the treated material on the
nanostructuring process, other tests were carried out with different materials. In theory,
different materials have a different receptivity to severe plastic deformation and therefore to
15 nanocrystallization mechanisms. The grain refinement procedures under severe plastic
deformation depend on many intrinsic and extrinsic factors, such as the structure and the
stacking fault energy (SFE) of the material. The higher the SFE energy of the material, for
example such as pure iron, the more difficult the activation of the various slip planes and the
generation of dislocations necessary for the grain requirement procedures are made. The
20 crystallographic structure of the metal and the optional presence of other elements such as
carbon or other alloy elements, especially in the form of precipitates that favour the
formation of dislocations, therefore have an influence on the productivity of the
nanostructuring process.
Test 4
25 A comparative test was carried out with samples of 304L stainless steel and a
32CDV13 structural steel under conditions similar to test 1 with type S280 shot. Figure 11
illustrates the results of these tests in terms of hardness profile for R=3000% in a
representation analogous to Figure 5. Curve 53 corresponds to the 304L stainless steel.
Curve 54 corresponds to the 32CDV13 structural steel. The hardness profiles of these
30 materials correspond to the trends observed in test 1. Regions 34 and 35 of Figure 11 have
the same meaning as in Figure 5. Curve 31 from test 1 (E24 steel) is plotted by way of
comparison. The thicknesses Zn observed are visually 143 urn for E24, 176 urn for 32CDV13
structural steel and 155 um for 304L stainless steel.
14
Once again, the validity of the empirical quantitative definition given above for
curves 54 and 31 is observed.
This definition clearly corresponds to curve 53 (304L steel) when the reference for
the hardening is chosen at a depth of 300 um. The choice of reference is explained by the
5 change of microstructure specific to the 304L steel, during the peening of the material, and
more particularly during a first step of the peening corresponding to a step of strainhardening
of the material.
During the first step of the peening of the material, a certain amount of austenite of
the 304L steel is converted to strain-induced martensite. This conversion to strain-induced
10 martensite gives rise to a significant increase in the hardness. On curve 53, a significant
reduction in the hardness is visible between 300 and 350 um. This reduction in the hardness
corresponds on the whole to the austenitic phase transition zone and the phase having a high
content of strain-induced martensite. In a second step, the nanostructured layer 3 appears in
the martensitic phase. Thus, starting from a thickness of greater than 350 um, the sample of
15 304L steel has its original hardness of the austenite and for a thickness of less than 300 um
the hardness of the material is increased both by the nanostructured layer and by the presence
of strain-induced martensite. Thus, the reference hardness used for determining the
nanostructured layer is the hardness at the deepest layers of the strain-induced martensite,
which is here around 300 um.
20 Test 5
A comparative test was carried out with samples of pure iron containing 0.03C
(99.8% Fe) under conditions similar to test 1 with type SI70 shot. The pure iron is assumed
to be one of the least favourable materials for grain refinement due to its ferritic structure
and its high SFE energy (around 200 mJ/m2). Curve 55 from Figure 12 represents the
25 thickness Zn resulting from this test, observed visually, as a function of the degree of
coverage R. Curve 36 from test 1 (E24 steel) is plotted by way of comparison.
It is thus confirmed that the E24 steel nanocrystallizes more rapidly (appearance of
the nanostructured layer at R=750%) than pure iron (appearance of the nanostructured layer
at R=1000%) and has a thicker nanostructured layer (Zn=130 um versus Zn=100 um) at
30 saturation. Test 5 shows that the process makes it possible to obtain nanostructured layers
thicker than 100 um for most of the materials that can be envisaged.
Test 6
15
In order to evaluate the effect of smaller particles, tests were carried out with
samples of pure iron containing 0.03C (99.8% Fe) and type S070 shots. The other conditions'
are similar to test 1.
Figure 13 illustrates the results obtained with a projection speed V = 60 m/s in a
5 representation similar to Figure 8. Curve 60 represents the Vickers hardness at the surface
and curve 61 the thickness z„ observed visually. It is observed that the thickness z„ saturates
at a level close to 60 urn from R= 3000% onwards. With small particles such as type S070,
this degree of coverage may be rapidly achieved, for example in less than 300 s with a
common peening material.
10 Figure 14 illustrates the results obtained with a projection speed V= 92 m/s in a
representation similar to Figure 8. Curve 62 represents the Vickers hardness at the surface
and curve 63 the thickness z„. It is observed that the thickness z„ saturates at a level close to
80 to 90 urn from R= 3000% onwards.
These results should be compared with those presented in Figure 4 of the
15 International ISIJ publication cited above, where the Fe-3.3Si alloy used has a ferritic
crystalline structure comparable to Fe-0.03C. Test 6 demonstrates the obtaining of a greater
nanostructured thickness with a degree of coverage, a particle size and a projection speed
that are all lower than in this publication. It is noted that the comparison of degrees of
coverage requires a calibration due to different definitions in the two cases. The use of a
20 lower projection speed may prove advantageous for reducing the roughness of the treated
sample or protecting a material more vulnerable to microcracks.
Test 7
Test 7 was carried out with samples of pure iron containing 0.03C (99.8% Fe) and
type SI70 shots. The other conditions are similar to test 1.
25 Figure 15 illustrates the results obtained with a projection speed V= 57 m/s in a
representation similar to Figure 8. Curve 64 represents the Vickers hardness at the surface
and curve 65 the thickness z^. The numbers 44, 45 and 46 have the same meaning as in
Figure 8. It is observed that the thickness Zn saturates at a level close to 100 urn.
Test 8
30 A second series of tests will now be described. In this second series of tests, the
hardness profiles of samples were measured with a more precise method in order to provide
a definition of the nanostructured layer based solely on the hardness curve of the material.
Table 3 recounts the results of the second series of tests carried out according to the
same conditions as test 1 presented in Table 2.
16
The method used for producing the hardness profile during this second series of
tests consists in making an indentation line with a step of 10 um starting from 20 urn from
the outermost surface to a depth of 100 um. The indentation line is then continued with a
step of 50 um to a depth of 300 um. The indentation line is made with a micro Vickers
5 hardness tester having a pyramidal tip with a load of 25 g (HV 0.025) which possesses a
lens. It is a Buehler Micromet 5104 microhardness tester comprising a motorized table
having a step of 1 um and Buehler Omnimet Mhtsa control and measurement software. The
hardness profile is thus obtained from a depth of 20 um to 300 um. The values
communicated are an average of three indentation lines in order to have a reliable and
10 reproducible measurement. In the same way as in the preceding tests, the surface of the
samples and the nanostructured layer are visualized by optical microscopy. The observation
of the samples is carried out using a Zeiss axio scope Al microscope, a Qimaging
Micropublisher 5.0 RTV camera, a Zeiss EC EPIPLAN X10/0.2HD lens and Axiovision 4.8
software.
15 The first three columns of Table 3 correspond to the first three columns of Table 2.
The fourth column mentions the thickness of the nanostructured layer, denoted by z^ with
reference to the hardness. Indeed, in test 8, the thickness of the nanostructured layer z^ was
obtained by a method solely based on the hardness profile as a function of the depth z. For
this, a hardness threshold is determined by calculating the median value of the hardness
20 between the hardness measured on the surface layer and the hardness of the sample in the
deep layer in which the material is not substantially modified by the peening.
The thickness of the nanostructured layer z^ therefore corresponds to the depth at
which the increase in the hardness is equal to half of the increase in hardness observed at the
surface of the sample after treating this surface. '
25 The fifth and sixth columns mention the hardness at the surface of the sample on
the treated face and on the untreated face. These values correspond to the first measurement
points' of the measured hardness curve, that is to say to a depth of 20 um. On the whole, the
hardness is measured closer to the surface than in test 1, so that the hardness value is higher
than in Table 2. Indeed, the size of the grains in the vicinity of the surface varies according
30 to a gradient. Thus, in one outermost surface region, the size of the grains varies between 10
and 50 nm, and in a deeper region, the size of the grains varies between a few tens of
nanometres to a few hundreds of nanometres. Furthermore, the hardness is measured with a
larger load in test 1 than in test 2. The impression made in the material therefore has larger
dimensions in test 1 and therefore generates a less precise measurement.
17
The last column from Table 3 mentions the uncertainty margin of the thickness
measurement z^ resulting from the uncertainty margin of the microhardness tester. Indeed,
the hardness measurements have an uncertainty of around ±10 Vickers for the E24 steel,
±9.5 Vickers for the 32CDV13 steel and ±13.5 Vickers for the 304L steel. For better
5 accuracy of the hardness measurement, the hardness tester load is adapted as a function of
the hardness of the material: a greater load is used for harder materials. Thus, a load of 50 g
(HV 0.050) is used for the 32CDV13 steel and for the 304L steel.
Figure 21 represents the hardness profiles obtained by the method explained above
for the samples corresponding to the samples from test 1 with R=3000%. Curve 70
10 corresponds to type SI70 shot. Curve 72 corresponds to type S330 shot. Curve 73
corresponds to type S550 shot. On all the curves 70, 71 and 73, a zone of very high hardness
appears which corresponds to the nanostructured layer 3 and a second zone appears where
the hardness decreases more gradually with the depth and which corresponds to the strain
hardening of the material.
15 By way of example, in Figure 21, the hardness value 74 measured in the deep layer
and the maximum hardness value 75 measured on the surface layer of the sample associated
with curve 70 are respectively equal to 142 and 300 Vickers. The corresponding threshold 71
has a value of 221 Vickers, which corresponds to the median value between the hardness
value 74 measured in the deep layer of the sample and the maximum hardness value 75
20 measured on the surface layer of the sample.
This threshold makes it possible to determine a thickness z^ of the nanostructured
layer having a value approximately equal to 81.5 um for the test corresponding to the S170
shot.
An uncertainty range of the thickness z^ of the nanostructured layer is therefore
25 determined from the hardness threshold and from the uncertainty range of the hardness. By
way of example, for the threshold 71 of 221 Vickers presented previously, the boundary
values of the thickness of the nanostructured layer are plotted for hardness values 85 and 86
respectively of 231 Vickers and 211 Vickers. Thus, the thickness of the nanostructured layer
lies within a range of around 69 to 92 um. The uncertainty ranges of the thickness of the
30 nanostructured layer are presented in Table 3. Thus, due to the uncertainty of the hardness,
the thickness of the nanostructured layer measured graphically itself also has a measurable
uncertainty.
As indicated previously, the second measurement method based on the hardness
agrees satisfactorily with the visual determination method: Figure 22 schematically
18
represents the regions observed on the optical micrographs of the sample corresponding to
curve 70 from Figure 21 (S170 peening at R=3000%). The hardness profile 70 as a function
of the depth z from the surface of the sample is plotted on the schematic representation of
these regions.
5 Observed in Figure 22 is a nanostructured surface layer 77 corresponding to a
region in which the material is substantially amorphous and homogeneous. Layer 77
corresponds to the darker zone observed in Figures 17 to 19. Layer 77 extends from the
surface 76 of the part to a second layer 78. This second layer 78 corresponds to the region in
which grain boundaries are observed and in which the size of the grains delimited by the
10 grain boundaries increases with the depth. On the optical micrographs, layer 78 corresponds
to the region which extends from a sudden change in contrast starting from layer 77. This
second layer 78 corresponds to the strain-hardening region of the material. A third layer 79
comprises a region where the size of the grains remains constant. The hardness threshold 71
agrees substantially with the boundary 84 observed visually between the nanostructured
15 surface layer 77 and the layer 78.
The difference between the thickness values Zn observed visually listed in Table 2
and the thickness values z^ listed in Table 3 originate essentially from the relatively high
uncertainty margin of the measurements mentioned in Table 2 typically of the order of
±30 um. In reality, the visual observations listed in Table 2 encompass a portion of the
20 transition layer 78, which explains the higher thickness values.
The method of measuring thickness based on the hardness described above may
display a difference with the optical observation when the thickness of the nanostructured
layer is thin, which corresponds to the case of the samples from test 8 at a degree of coverage
of less than 750%. Another method for determining the thickness of the nanostructured layer
25 may then be used. This other method is also based on the principle of determining the
thickness of the nanostructured layer from a hardness threshold. This method starts from the
observation that, when it appears on the sample and therefore when it has a very thin
thickness, the nanostructured layer 3 has a hardness value at the surface which corresponds
to this threshold. By way of illustration, with reference to Figure 23, curve 80 represents the
30 thickness of the nanostructured layer as a function of the coverage and curve 81 represents
the surface hardness of the sample as a function of the coverage for the SI 70 peening test. A
minimum detectable thickness 82 of the nanostructured layer appears for a coverage of
150%. However, the surface hardness 83 measured during this appearance of the
nanostructured layer is 226 Vickers. This hardness threshold of 226 constitutes a realistic
19
value of the hardness threshold for determining the thickness of the nanostructured layer
after treatment with a coverage of less than 750%. This alternative value has a value close to
the hardness threshold determined with the aid of the median value at R=3000% (221
Vickers). In test 8 listed in Table 3, the hardness thresholds were determined with this other
5 method for coverage values of less than 750%. In Table 3, the values determined with this
other method have an asterisk.
These results should be compared with those presented in Figures 3(a) and 4(b) of
the Materials Transactions publication cited above. In particular, much greater thicknesses
are obtained in a much shorter time and with a much more flexible process than ultrasonic
10 shot peening (USSP) for the same shot size.
Although the results presented above are obtained with fiat metal samples, the
processes used are applicable to metal parts of any shape. In particular, in order to treat a
non-planar surface, it is possible to successively treat limited portions of the non-planar
surface, by each time orienting the treated surface portion so that the angle conditions
15 described previously with reference to the flat surface are approximately respected for each
successive portion of the non-planar surface. The expression "successive portion" is
understood here to mean a surface portion that is relatively small with respect to the local
radius of curvature, so that an average orientation of the surface portion can be defined, and
relatively large with respect to the size of the shots projected, so that a large number of
20 impacts can statistically be envisaged.
Certain non-planar geometries are capable of producing multiple impacts by one
and the same particle on the part, that is to say rebounds. However, given that the rebounds
lead to very high energy losses, it is assumed that it is the primary incidence of the particle,
that is to say the incidence before the first impact on the part which is the most significant.
25 If it is not desired nor even possible to carry out the aforementioned orientation
conditions for each surface portion of the part to be treated, it is preferable to identify the
portions of the metal part intended to be the most stressed in its final use, that will be
referred to as the working surfaces of the part. For example, the working surfaces of a gear
pinion are generally the bases of the teeth. The nanostructuring treatment of a pinion can
30 therefore be carried out, in one particular embodiment, by successively orienting the tooth
base surfaces facing the particle jet, so as to carry out the particular orientation of the
primary incidences of the particles on the tooth base surface.
A single projection nozzle has been presented in the embodiment of the machine
from Figure 2. However, it is also possible to conceive a peening machine with several
20
projection nozzles. These projection nozzles may especially be arranged so as to target the
same surface of the part along several different incidences. Projection nozzles may also be
arranged so as to target various surfaces of the part to be treated.
Other relative arrangements of the projection nozzles and of the support of the part
5 can be envisaged in order to produce primary incidences of the particles which are
distributed in a cone or a conical film having an outer half apex angle between 10° and 45°.
In particular, a displacement may be carried out at the projection nozzles.
Although the invention has been described in connection with several particular
embodiments, it is clearly obvious that it is in no way limited thereto and that it includes all
10 the technical equivalents of the means described and also combinations thereof if the latter
come under the scope of the invention.
In particular, the embodiments described in the examples relate to initially
homogeneous materials on which the peening processes described make it possible to form
relatively thick nanostructured surface layers. It is possible to characterize the degree of
15 coverage applied to a given material by the thickness of the nanostructured layer that this
coverage made it possible to obtain. Hence, the application of a similar degree of coverage to
a material having undergone other prior treatments is also capable of effectively producing
nanostructured surface layers, even if this pretreated material does not correspond to the
examples described, for example a heterogeneous material.
20 The use of the verb "to have", "to comprise" or "to include" and its conjugated
forms does not exclude the presence of elements or steps other than those mentioned in a
claim. The use of the indefinite article "a" or "an" for an element or a step does not exclude,
unless otherwise mentioned, the presence of a plurality of such elements or steps. Several
means or modules may be represented by one and the same material element.
25 In the claims, any reference sign between parentheses should not be interpreted as a
limitation of the claim.
21
Table 2: test 1, E24 steel, Rockwell hardness of the shots = 48HRC
T , Hardness Hardness of H ,
Yu • V(m/s) R(%) zn(|im) of treated untreated face Mamn,;SS
S h 0 t face(HV) (HV) gam(/o)
100 0 198 108 83%
150 0_ 211 114 85%
200 0 200 133 77%
300 0 J?12 111 91%
500 0 241 112 115%
S170 57 750 69 256 108 137%
1000 72 263 111 137% ,
1500 91 L274 116 136%
' 3000 129 308 113 173%
6000 138 '309 113 173%
- 10000 140 302 116 160%
100 0 215 130_ 65%
150 0 224 132 70%
200 _0 224 138 62%
300 67 247 139 78%
500 91 _262 137 91%
S280 52 750 101 278 138 101%
1000 120 290 113 157%
1500 134 295 116 154%
3000 143 298 114 161%
6000 178 301 113 166%
10000 172 315 114 176%
100 0 213 114 87%
150 0 233 116 101%
200 0 234 110 113%
300 111 264 111 138%
500 112 253 108 134%
S330 60 750 142 282 114 147%
1000 160 290 114 154%
1500 175 298 112 166%
3000 192 310 123 152%
6000 193 300 131 129%
• 10000 _186 304 142 114%
100 0 206 129 60%
150 0 216 144 50%
200 0 223 131 70%
300 0 227 135 68%
500 0 243 145 68%
S550 49 750 104 278 148 88%
1000 176 292 147 99%
1500 168 279 153 82%
3000 164 292 159 - 84%
6000 175 295 157 88%
I J 10000 I 173 J 308 | 167 | 8 4 %
22
Table 3: Test 8, samples corresponding to test 1, E24 steel, Rockwell hardness of the shots = 48HRC
- , . , . . Hardness Hardness of ._. , _, .„ „
YP
h
e
t A R(%) zn(nm) of treated untreated H a r d ^ f z" h m n ' z"h m f
shot (mis) face(HV) f a c e ( H V ) gain(%) frm) frm)
100 0.00 225 142 58%
150 0.00 226 133 7096
200 27.46* 234 140 67%
300 30.28* 252 143 76%
500 46.47* 276 132 109%
S170 57 750 50.00* 281 135 108%
1000 54.22 288 140 106% 49.29 62.67
1500 59.15 290 140 107% 57.74 74.64
3000 81.69 292 131 123% 69.01 92.25
6000 . 94.36 323 135 139% 90.84 96.47
10000 87.32 327 127 157% 73.23 95.77
100 0.00 240 136 76%
150 35.21* 244 136 79%
200 34.50* 253 139 82%
300 39.43* 260 135 93%;
500 67.60* 267 129 107%
S330 60 750 69.71 284 128 122% 61.26 90.14
1000 76.05 297 129 130% 69.01 96.47
1500 111.26 299 126 137% 102.11 121.83
3000 111.97 309 128 141% 97.88 123.23
6000 123.94 310 157 97% 109.50 139.43
10000 97.14 310 126 146% 90.00 113.57
100 0.00 222 135 64%
150 0.00 225 139 62%
200 0.00 227 141 61%
300 29.57* 240 144 67%
500 44.36* 248 128 94%
S550 49 750 57.74* 261 141 85%
1000 98.59 271 134 102% 76.76 161.97
1500 108.45 289 148 9596 81.69 133.09
3000 97.18 295 132_ 123% 83.80 146.47
6000 115.00 309 142 118% 85.71 140.00
J 10000 I 119.28 [ 325 | 144 | 126% | 98.57 | 150.00

CLAIMS
1« Process for the surface treatment of a metal part, comprising:
- exposing a surface (1) of the metal part to a stream (20) of particles having
a sphericity greater than or equal to 85%, so that said surface receives said
5 particles along several primary incidences, the primary incidences of the
particles on the surface being distributed in a cone or a conical film which
has an outer half apex angle (a, a+p, a-P) between 10° and 45°, until a
surface layer (3) of nanostructures is obtained,
the particles having a diameter of less than 2 mm and greater than 0.1 mm
10 and being projected at a speed between 40 m/s and 100 m/s;
wherein the stream of particles comprises a jet of particles (20) projected
along a central direction (25), the metal part being fixed to a support (14) so
that as the inclination(a) of the surface of the part exposed to the stream
with respect to the central direction is between 10° and 30°, the support (14)
15 or the projection means (11) being rotated about an axis coaxial with the
central direction of the jet of particles.
2. Process according to claim 1, wherein the the inclination(a) of the
surface of the part exposed to the stream with respect to the central
direction is 15°.
20 3. Process according to Claim 1 or 2, whereinthe particles have a
diameter of greater than 0.3 mm and of less than 1.4 mm.
4. Process according to any one of Claims 1 to 3, wherein- the
incidences of the particles are distributed substantially continuously.
5. Process according to any one of Claims 1 to 4, wherein the cone or
25 the conical film has an outer half apex angle of between 10° and 30°.
6. Process according to any one of Claims 1 to 5, wherein the particles
ere projected at a speed of between 50 and 80 m/s.
7. Process according to any one of Claims 1 to 6, wherein the particles
have a hardness greater than the hardness of the surface of the part before
30 treatment.
8. Process according to any one of claims 1 to 7, wherein the jet of
particles (20) has a conical-shape which have a outer helf apex angle (3
lower.
9. Surface treatment device for a metal part, comprising:
5 a projection means (11) capable of producing a stream of particles having
a sphericity greater than or equal to 85% and having a diameter of less than
2 mm and greater than 0.1 mm and that are projected at a speed of
between 40 m/s and 100 m/s, the projection means (11) being capable of
producing a jet of particles (20) projected along a central direction (25)
10 a support (12) capable of holding a metal part, the support comprising a
surface (14) exposed to the stream of particles, and
an actuator capable of modifying an orientation of the support with respect
to the stream of particles so that the primary incidences of the particles on a
surface of the support are distributed in a cone or a conical film that has an
15 outer half apex angle of between 10° and 45°, the surface of the support
. (14) being oriented obliquely with respect to said central direction so that as
the inclination (a) of the surface of the part exposed to the stream with
respect to the central direction is between 10° and 30°,, the actuator being
capable of pivoting the support about an axis that is coaxial with the central
20 . direction of the jet of particles.