Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

Materials Science and Engineering C 29 (2009) 1357–1360

Contents lists available at ScienceDirect

Materials Science and Engineering C


j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / m s e c

Influence of the sandblasting on the subsurface microstructure of 316LVM stainless


steel: Implications on the magnetic and mechanical properties
M. Multigner a,b,⁎, E. Frutos a,b, J.L. González-Carrasco a,b, J.A. Jiménez a, P. Marín c, J. Ibáñez a
o
a
Centro Nacional de Investigaciones Metalúrgicas (CENIM-CSIC), Avd. Gregorio del Amo n 8, 28040 Madrid, Spain
b
Centro de Investigación Biomédica en Red en Bioingeniería, Biomateriales y Nanomedicina (CIBER-BBN), Madrid, Spain
c
Instituto de Magnetismo Aplicado (Departamento de Física de Materiales) UCM-ADIF, Nacional VI Km 22.5, 28230 Las Rozas, Spain

a r t i c l e i n f o a b s t r a c t

Article history: The austenitic stainless steel 316 Low Vacuum Melting is a material that is widely used in biomedical applications,
Received 31 July 2008 particularly for bone substitution or repair. Sandblasting is commonly used as surface modification for
Received in revised form 9 October 2008 biomaterials to obtain better mechanical fixation of the prosthesis. However the microstructural changes beneath
Accepted 3 November 2008
the surface associated with this modification as well as other physical properties are not well known. In this work,
Available online 12 November 2008
the results for microstructural, magnetic and mechanical features of the sandblasted austenitic steels 316LVM
Keywords:
subsurface are analyzed for the first time.
316LVM stainless steel © 2008 Elsevier B.V. All rights reserved.
Sandblasting
Microstructure
Magnetic properties
Hardness

1. Introduction structure) transformation by plastic deformation, that involves a


change from paramagnetic to ferromagnetic behaviour. This change
Austenitic stainless steel 316LVM has an extensive application as a could be regarded as a contraindication for the performance of
biomaterial since it combines good mechanical properties with magnetic resonance imaging (MRI) because of the potential image
reasonable biocompatibility. Additionally, its low cost and easy artefacts that the implant could produce and also because of the risks
mechanization, when compared to other metallic implant materials, associated with the dislodgement of the implant due to the magnetic
makes the 316LVM steel a desirable class of implant material for field attraction [7 and references therein].
orthopaedic applications. Recent studies consider that improved In this context, the aim of this study is to investigate the effect of
stabilisation of short-term implants used for bone fracture repair sandblasting on austenitic stainless with respect to its magnetic
may be crucial at the early-stage of implantation. The activity for the behaviour and to analyse the relationship between subsurface mechan-
production of randomly rough surfaces by a simple method such as ical properties and the microstructural changes induced by blasting.
sandblasting has been particularly important. Most of the research has
been concerned with improving the performance of Ti and Ti-base 2. Experimental procedures
blasted alloys [1–3]. However, studies on the effect of sandblasting on
austenitic steels 316LVM are unknown in the open literature. Disks of 20 mm diameter and 2 mm thickness of 316LVM steel,
One very important factor for this research is the fact that surface whose chemical composition is given in Table 1, were supplied by the
severe plastic deformation processes, such as high-energy shot- implant manufacturer (Surgival SL, Spain).
peening or surface mechanical attrition treatment, have been Sandblasted specimens were prepared by the manufacturer by
developed to produce a nanocrystaline surface in order to improve using Al2O3 particles of about 750 μm diameter under 350 kPa and
the surface properties of different alloys including stainless steel [4,5]. 20 min. Roughness of the as-processed specimens has been
Even sandblasting together with a thermal treatment has been proved determined with a profilometer Mitutoyo Surftest 401 averaging 3
to be able to produce a nanocrystaline layer in 304 stainless steel [6]. measurements of 4 cm in length. Unblasted samples were also
On the other hand, it has been widely reported that austenitic investigated for comparative purposes.
stainless (fcc structure) steels are susceptible to martensitic (bcc The steel's microstructure was characterized using both X-ray
diffraction and scanning electron microscopy. To evaluate the
⁎ Corresponding author. Centro Nacional de Investigaciones Metalúrgicas (CENIM-
formation of strain induced α′-martensite during the sand blasting
CSIC), Avd. Gregorio del Amo no 8, 28040 Madrid, Spain. treatments, X-ray diffraction (XRD) measurements were carried out
E-mail address: mmultigner@cenim.csic.es (M. Multigner). with a Bruker AXS D8 diffractometer in grazing incidence condition

0928-4931/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.msec.2008.11.002
1358 M. Multigner et al. / Materials Science and Engineering C 29 (2009) 1357–1360

Table 1
Chemical composition of the 316LVM stainless steel used in the investigation.

Element Fe Cr Ni Mo Mn Si C Cu N S
wt.% Bal 17.48 14.13 2.87 1.62 0.53 0.024 0.067 0.061 0.001

with a beam incidence angle of 1° and 2θ scan between 35 and 115°


with a step size of 0.03°. The Rietveld method is a powerful tool for the
calculation of structural parameters from diffraction patterns. For the
application of the Rietveld method to grazing incidence X-ray
diffraction (GIXRD) instrument functions were empirically parame-
terised from the profile shape analysis of a corundum sample
measured under grazing incidence under the same conditions. In
this study, we have used the version 4.0 of Rietveld analysis program
TOPAS (Bruker AXS) for the XRD data refinement.
Microstructural and surface characterization was performed by
using a scanning electron microscope (SEM) Jeol JSM-6500F equipped
with a field emission gun (FEG) emitter coupled with an energy
dispersive X-ray (EDX) system for chemical analysis. In order to
preserve the original blasted surface during sectioning and to avoid
artefacts during the measurements performed beneath the surface,
selected specimens were electrolytically coated with a fine layer of Cu.
The cross-section of blasted and unblasted samples were ground with
consecutively finer SiC papers, and finely polished with diamond
paste and colloidal silica (40 nm) to remove all disturbed metal
(Beilby layer). This surface finish makes it possible to reveal the grain
structure in the backscattered electron mode.
Magnetic measurements were performed in a vibrating sample
magnetometer PPMS-VSM Quantum Design at 300 K and with a
maximum magnetic field of 2 T. For this technique small pieces of
approximately 4 × 2 × 2 mm were mechanically removed from the
disks, giving rise to a parallelepiped with three blasted faces. Finally,
mechanical properties were determined by Vickers microhardness
measurements performed on cross sectional specimens describing a
zig-zag along lines perpendicular to the blasted surface. These tests Fig. 1. BEI images corresponding to polished cross sections showing the subsurface
were carried out in a Wilson equipment by using a load of 10 g and microstructure of: A) unblasted and B) blasted specimens.
15 s of dwell time.
Fig. 3 shows the magnetic hysteresis loops registered at room
3. Results and discussion temperature corresponding to the investigated materials. Two aspects
could be remarked on:
Surface SEM examination reveals that blasting of the alloy causes
– at high fields there is a linear response corresponding to a
the plastic deformation of the alloy, resulting in an irregular rough
paramagnetic signal, as expected for the austenitic stainless steel,
surface morphology with an average roughness, Ra, of 6.7 μm. A closer
examination of the surface reveals the presence of dark zones
containing large particles of heterogeneous size. Particles present
polygonal edges and are often broken forming agglomerates of fine
particles. EDX analysis revealed that they are remnants of the particles
used for blasting.
Cross sectional examination of the unblasted and blasted samples,
Fig. 1, revealed that, in addition to roughening of the surface,
sandblasting produces a significant change in the microstructure of
the material close to the blasted surface, which is obviously related to
plastic deformation during processing. Two zones, that have under-
gone deformations with different strain level, can be distinguished
without a clearly defined borderline. The top surface layer is about
10 μm thick and the grain size cannot be signalled out with the
magnification of the Fig. 1. However in Fig. 2 it can be appreciated that
it is characterized by randomly distributed nanometre-scale grains.
Below this layer, there is observed a deformed zone where the grains
size increases with increasing distance to the blasted surface. More-
over, the deformation is inhomogeneous in terms of depth from place
to place, most probably due to the different orientations of grains and
the heterogeneous nature of the plastic deformation within and be- Fig. 2. Close up BEI images showing grain refinement at the subsurface zone of the
tween grains. blasted specimen.
M. Multigner et al. / Materials Science and Engineering C 29 (2009) 1357–1360 1359

been transformed to α′-martensite. First of all, in order to isolate the


effect of the sandblasting, it is enough to subtract the signal arising
from the unblasted sample to the sandblasted one, bearing in mind
that the two samples have the same geometry. Then, the mass that
gives rise to the magnetization will be

μ
m=
Ms

where, m is the mass expressed in kg, μ is the magnetic moment in


A m2 and Ms the saturation magnetization in A m2/kg. If we assume
7 · 103 kg/m3 as the density of the material, we can know the volume
of the material affected by the transformation. If, as we suppose, the
magnetization only comes from the sandblasted surface, then that is
measurable. By an easy calculation we obtain the result that the
martensite layer would be around 9 μm thick, which agrees with the
thickness of the top surface layer with nanosized grains.
It should be noticed that the magnetic moment of the sandblasted
Fig. 3. Magnetic hysteresis loops corresponding to the blasted and unblasted specimens. sample is higher but of the same order of magnitude as that of the
The paramagnetic contribution has been obtained from the extrapolation of the high field unblasted (Fig. 3) one or even higher than a completely austenite
data. Inset: Magnetic hysteresis loops before and after the heating at 700 °C for 1 h.
sample. Even though the saturation magnetization of a ferromagnetic
whereas at low fields, superimposed on the former, there is a material is usually 4 order of magnitude higher than the magnetiza-
ferromagnetic signal. For comparison, the hypothetical paramag- tion of a paramagnetic material at high magnetic fields [14], the
netic response of the 100% austenitic material, extrapolated from ferromagnetic phase must represent a very small volume fraction.
the high field values, has also been represented. From the point of view of biomedical components no significant dif-
– the sandblasted sample has greater ferromagnetic signal than the ferences in the magnetic response could be expected between the
unblasted one. unblasted and blasted specimens. For example, let us consider a
surgical screw of 4 mm diameter and 60 mm in length with the
The origin of the ferromagnetic contribution in the sandblasted theoretical magnetic response represented in Fig. 3. The highest
specimens is related to the subsurface formation of α′-martensite, gradient in a magnetic resonance equipment is just outside the
which is obviously the consequence of the severe deformation [8–10]. magnet and for a 3 T magnet the gradient, jY B, will be around 3 T/m.
In the case of the unblasted sample, however, the ferromagnetic Let us assume the simplest situation: the screw is in the axis of the
behaviour could result from the plastic deformation that take place magnet and the magnetization corresponds to a field of 1.5 T, 0.5 A m2/
during the cutting of the specimen [11] which, because of its small size kg. Taking into account the volume and the density of the steel, we obtain
and the relative high surface/volume ratio, might have been playing a μ = 2.6 A m2. Considering that μ and jB are parallel, the force that the
critical role. On the other hand, the possible presence of small magnetic field gradient would exert over the surgical screw would be
amounts of ferrite in the original material should not be ruled out. about 8 · 10− 3 N (equivalent to the weight force of 0.8 g). Consider now
To find out the origin of the ferromagnetic contribution on the the same screw but with the sandblasted surface. Assuming the said
unblasted samples a thermal treatment at 700 °C for 1 h in air has been layer of martensitic phase 9 μm thick with a saturation magnetization
performed. The reason for this experiment is that α′-martensite formed of 130 A m2/Kg, we would obtain a magnetic moment of μ = 6.2 A m2.
in austenitic stainless steel by plastic deformation can reverse to austenite Adding this value to the previous 2.6 A m2 corresponding to the bulk, we
by thermal treatments [12]. The hysteresis loops of the thermally treated obtain a total magnetic moment of 8.8 A m2, which within the supposed
samples, shown in the inset of Fig. 3, exhibit an important decrease in the gradient exerts a force of 26.4 · 10− 3 N. Although this value is three-fold
ferromagnetic contribution that indicates the reversion of the martensite. the force calculated for the unblasted surgical screw, it is equivalent to
There still remains a small ferromagnetic fraction, but the completely
reversion of the martensite needs further investigation.
GIXRD patterns for the blasted sample, Fig. 4, shows the existence
of low intensity reflections corresponding to the α′-martensite phase
beside the dominant peaks for austenite. This feature confirms that
strain induced martensite is formed in the surface of the sample. In the
same figure the broadening of the diffraction Bragg peaks is also
clearly observed. This suggests a continuous decrease of the grain size
and/or an increase in the lattice strain. The Topas 3.0 program, used
for the refinement of the GIXRD patterns, uses the double-Voigt
approach for the size–strain analysis. Due to the variation of the
microstructure through the thickness observed in Figs. 1 and 2, it was
not possible to do the calculation of the volume weighted mean
crystallize size and a mean strain value, based on individual crystalline
size and strain contributions to the line profile shape.
How can we estimate the amount of martensite induced by the
sandblasting? If we assume that the saturation magnetization of
the martensitic phase is the same as for the 304 stainless steel [13],
Ms = 130 A m2/kg,1 we can estimate the thickness of the layer that has

1
(Certainly, magnetization of 316 steel should be lower because the percentage of
Fe is smaller than in 304). Fig. 4. GIXRD patterns from (a) unblasted and (b) blasted samples.
1360 M. Multigner et al. / Materials Science and Engineering C 29 (2009) 1357–1360

determine the precise distribution of magnetic domains in the blasted


zone by Magnetic Force Microscopy and residual stresses by
synchrotron radiation are in progress.

4. Conclusions

From the above considerations, we can conclude that sandblasting


of 316LVM stainless steel produces a severe plastic deformation that is
accompanied with a grain refinement in the nanometric domain in the
first 10 μm depth from the surface and an enhancement of micro-
hardness until about 200 μm depth. In addition, X-ray diffraction and
magnetic measurements confirm that α′-martensite phase transfor-
mation takes place in the sandblasted affected zone, although the
volume fraction transformed is so small, that the changes in the
magnetic properties become almost negligible for a potential move-
ment of the implant due to an external magnetic field.
Work hardening, grain refinement and/or the presence of
α′-martensite give rise to the hardness gradient that reaches values
Fig. 5. Microhardness values as a function of depth from the blasted surface.
for the surface that are double those for the bulk.
The present results show the possibility of significantly improving
the weight force of 2.6 g, which cannot be considered enough to provoke the superficial mechanical properties without compromising the bulk
the movement of the screw. magnetic response of the austenitic stainless steel by sandblasting
From previously described results, it follows that blasting of the technique and therefore it could be considered as a promising surface
alloy causes significant microstructural changes at the subsurface modification for use in medical applications.
zone that could be important for the mechanical behaviour of the
bulk. Microhardness measurements performed along a line perpen- Acknowledgements
dicular to the surface, Fig. 5, revealed that the blasted specimen
presents a gradient in hardness with a maxima of 410 HV0.01 close to The authors wish to thank the financial support for Spanish Pro-
the surface. With increasing depth, hardness decreases achieving a ject. MAT2006-12948-C04-01 and Fundación Mutua Madrileña
near constant value after a depth of 200 μm, that is slightly higher than (Spain). Dra. M.M. thanks the “Juan de la Cierva” grant. The CIBER of
the hardness values determined in the bulk and in the subsurface area Bioingeniería, Biomateriales y Nanomedicina is supported by the
of an unblasted sample, for which 204 ± 8 and 203 ± 8 HV0.01 were ISCIII.
respectively obtained.
As regards hardness, on the one hand, the work hardening capability References
of the stainless steels is well-known and on the other hand, grain size
[1] A. Wennerberg, T. Albrektsson, B. Andersson, Int. J. Oral Maxillofac. Implants 11
gradient analysis proves that the material is deformed until 200 μm
(1996) 38.
depth. However there are at least two additional factors that could [2] V.M. Goldberg, S. Stevenson, J. Feighan, D. Davy, Clin. Orthop. Relat. Res. 319 (1995)
account for the microhardness gradient: the presence of α′-martensite, 122.
[3] C. Aparicio, F.J. Gil, U. Thams, F. Muñoz, A. Padrós, J.A. Planell, Key Eng. Mater. 254–256
which has higher hardness than austenite, [15] and the grain refinement,
(2004) 737.
as the Hall–Petch expression predicts [16,17]. Certainly the hardness [4] G. Liu a, J. Lu, K. Lu, Mater. Sci. Eng. A 286 (2000) 91.
gradient is the consequence of the sum up of all the three factors, with [5] X.H. Chen, J. Lu, L. Lu, K. Lu, Scr. Mater. 52 (2005) 1039.
different specific weights each other depending on the distance to the [6] X.Y. Wang, D.Y. Li, Wear 255 (2003) 836.
[7] L.P. Bendel, F.G. Shellock, M. Steckel, JMRI 7 (1997) 1170.
surface. [8] A.W. McReynolds, J. Appl. Phys. 20 (1949) 896.
It is clear that the refined microstructure is just beneath the surface [9] F. Lecroisey, A. Pineau, Metall. Trans. 1 (1972) 387.
but it is not so clear how the martensite is distributed inside the [10] T.S. Wang, B. Lu, M. Zhang, R.J. Hou, F.C. Zhang, Mater. Sci. Eng. A 458 (2007) 249.
[11] L.E. Samuels, G.R. Wallwork, J. Iron Steel Inst. 186 (1957) 211.
deformed volume, since a local recrystallization beneath the surface [12] S.S.M. Tavares, D. Fruchart, S. Miraglia, J. Alloys Compd. 307 (2000) 311.
could not be ruled out due to the eventual local increase of the [13] J. Childdress, S.H. Liou, C.L. Chien, J. Appl. Phys. 64 (1988) 6059.
temperature. For instance, it has been reported that in the case of the [14] B.D. Cullity, Introduction to Magnetic Materials, Addison-Wesley, Massachusetts,
1972, p.16.
Ti6Al4V alloy, temperatures as high as 1000 K can be achieved during [15] G.E. Dieter. Mechanical Metallurgy, third ed., McGraw-Hill Book Company, New
shot-peening [18]. In addition, it has been reported that the re- York, 1989, p.319.
crystalization temperature is higher than the α′ phase reversion [16] E.O. Hall, Proc. R. Soc. B 64 (1951) 747.
[17] N.J. Petch, J. Iron Stell Mater. 3 (1953) 25.
temperature [19]. So, if a recrystalization takes place in the surface, the
[18] L. Reissig, R. Völkl, M.J. Mills, U. Glatzel, Scr. Mater 50 (2004) 121.
phase would be austenite. Further experimental evidence is needed to [19] C. Herrera, R.L. Plaut, A.F. Padilha, Mat. Sci. Forum 550 (2007) 423.
clarify these points. [20] X.P. Jiang, X.Y. Wang, J.X. Li, D.Y. Li, C.-S. Man, M.J. Shepard, T. Zhai, Mater. Sci. Eng.
A 429 (2006) 30.
Furthermore, we should not rule out the presence of residual
stresses at the blasted zone that could also play an important role in
the mechanical properties, as occurs in Ti [20]. Specific experiments to

You might also like