Procedia Engineering 74 ( 2014 ) 421 – 428

Available online at www.sciencedirect.com

1877-7058 © 2014 Elsevier Ltd. Open access under CC BY-NC-ND license. Selection and peer-review under responsibility of the Politecnico di Milano, Dipartimento di Meccanicadoi: 10.1016/j.proeng.2014.06.293

ScienceDirect

XVII International Colloquium on Mechanical Fatigue of Metals (ICMFM17)

Fatigue behaviour of sintered duplex stainless steel

T. Ta skia*, Z. Brytana†, K. Labiszb

aDivision of Materials Processing Technology and Computer Techniques in Materials Science, Institute of Engineering Materials and Biomaterials, Silesian University of Technology, Konarskiego St. 18a, 44-100 Gliwice, Poland

bDivision of Biomaterials Engineering, Institute of Engineering Materials and Biomaterials, Silesian University of Technology, Konarskiego St. 18a, 44-100 Gliwice, Poland

Abstract

The stainless steel despite of tradition of use over the 100 years are steel interesting in terms of metallurgical improvement and development of new products. Stainless steel products undergo intensive development, especially for duplex stainless steel grades manufacture be conventional casting method and metal forming processes, thus, influences an increase in the interest of those materials by alternative technologies of metal forming, which is powder metallurgy. The powder metallurgy and manufacturing processes of metal powders and sintered components production undergoing rapid development. This trend applies also for sintered duplex stainless steels. The paper investigates fatigue behaviours of vacuum sintered duplex stainless steel produced by mixing in appropriate proportions powders of ferritic and austenitic stainless steel and elemental alloying powders and then sinter hardening in a vacuum. The mechanical properties of studied steel were evaluated in terms of tensile strength, hardness, toughness, plasticity. Fatigue tests were carried out in symmetric plane bending at stress ratio R= – 1 with frequency of about 24Hz. Fatigue crack propagation micromechanisms were investigated by means of a scanning electron microscope fracture surface analysis.

© 2014 The Authors. Published by Elsevier Ltd. Selection and peer-review under responsibility of the Politecnico di Milano, Dipartimento di Meccanica.

Keywords: Powder Metallurgy, Duplex stainless steel; Mechanical properties;

* Corresponding author. Tel.: +48 32 237 2509; fax: +48 32 237 2281. E-mail address: tomasz.tanski@polsl.pl

© 2014 Elsevier Ltd. Open access under CC BY-NC-ND license. Selection and peer-review under responsibility of the Politecnico di Milano, Dipartimento di Meccanica

422 T. Tański et al. / Procedia Engineering 74 ( 2014 ) 421 – 428

1. Introduction

Duplex stainless steels are called as duplex because they have a two-phase microstructure consisting of ferritic and austenitic grains. When duplex stainless steel is melted it solidifies from the liquid phase to a completely ferritic structure and as the material cools to the room temperature the transformation of about half of the ferritic grain to austenitic grains take place, thus resulting in the microstructure of roughly equal amounts of austenite and ferrite. The duplex microstructure gives this family of stainless steels a combination of attractive properties. Duplex stainless steels are about twice as strong as regular austenitic or ferritic stainless steels. Duplex stainless steels have significantly better toughness and ductility than ferritic grades. Regarding the corrosion resistance of stainless steels for chloride pitting and crevice corrosion resistance, their chromium, molybdenum and nitrogen content are most important. Duplex stainless steel grades have a range of corrosion resistance, similar or higher to the range of austenitic stainless steels. Duplex stainless steels show very good stress corrosion cracking resistance, a property they have inherited from the ferritic stainless steels. Regarding costs, the duplex stainless steels have lower nickel and molybdenum contents than their austenitic counterparts of similar corrosion resistance. Due to the lower alloying content, duplex stainless steels can be lower in cost, especially in times of high alloy surcharges [1-4].

The demand for lower production costs, especially in the automotive industries, resulted in increased use of sintered components even for highly stressed fatigue loaded components, like parking gears, camshafts, etc. The main sintered materials for automotive components is low alloyed steel, but the sintered stainless steels plays an important role for demanding corrosion applications like exhaust flanges and HEGO boss applications. The sintered duplex stainless steels give a unique opportunity to join corrosion resistance, high toughness, high plastic properties and mechanical resistance in one material, thus are such interesting material for huge amount of possible automotive sintered applications [19, 29, 23]. Sintered duplex stainless steels may be obtained within a single sintering cycle through the controlled addition of alloying elements promoting formation of austenite or ferrite to single-phase powders, both ferritic and austenitic trying to predict the final structure on the bases of Schaffler’s diagram [5]. Alloying element may be added in the form of single elements or in combined form and the sintering cycle is done in vacuum at argon backfilling and nitrogen is under pressure is used to obtain rapid cooling rate directly from sintering temperature. Sintered stainless steel, in order to achieve high mechanical properties must be must be sintered at high temperatures applying inert atmosphere [17, 18, 21, 22]. Depending on chemical composition sintered duplex stainless steels must be cooled form sintering temperature with controlled cooling rate due to the possibility of precipitations of brittle intermetallic sigma phase which highly negative influence on stainless steel properties. The presence of this brittle phase reduces the ductility and produce chromium depleted areas leading to decrease of the corrosion resistance and toughness. Sinter-hardening method mainly applied to steels undergoing a martensitic transformation during cooling, thanks to rapid cooling from sintering temperature can be applied to sintering duplex stainless steels thus creates the possibility of producing complex dual-phase microstructure with controlled mechanical properties and corrosion resistance in one sintering cycle with no need of the additional heat treatments [6-10].

The main purpose of this paper is the investigation on the basic mechanical properties and microstructures of different duplex stainless steels compositions manufactured from base powder of prealloyed single phase stainless steel and the addition of alloying elements powders. The mechanical properties of studying steels were evaluated in terms of tensile strength, hardness, toughness, plasticity with correlation to presented porosity and its morphology. Fatigue tests were carried out in symmetric plane bending at stress ratio R= – 1 with frequency of about 24Hz.

2. Experimental procedure

To produce sintered duplex stainless steel different compositions have been tested, using alloyed ferritic AISI 410L (0.14%Ni, 12.2%Cr, 0.88%Si, 0.09%Mn, 0.04%C) base water atomized powder of Hoganas Corporation (table 1) as a starting powder. Stainless steel base powders were mixed with the addition of alloying elements powders such as Cr (as Fe-Cr powder) and Ni, Mo and Cu as elemental powders in the right quantity to obtain sintered steel with chemical composition corresponding to duplex one. Chemical compositions of producing mixture were placed in austenitic-ferritic area of the Schaeffler’s diagram. During premix preparation and prediction of the final structure based on Schaffler’s diagram, thus CrE and NiE (CrE= %Cr + %Mo + 1,5 x %Si + 0,5 x %Nb; NiE= %Ni + 30 x %C + 0,5 x %Mn) equivalents are obtained introducing the wt. % quantity of the corresponding element

423 T. Tański et al. / Procedia Engineering 74 ( 2014 ) 421 – 428

in the formula. Naturally isothermal projected phase diagram of ternary Fe-Cr-Ni system was taken intoconsideration and the proper range of coexistence of austenite and ferrite was controlled. Analysed sintered stainless steel composition is given in Table 1. During preparation of composition lubricant Acrawax was used in a quantity of 0.65 wt. %. Premixes were prepared in tubular mixer for 20 min. and then uniaxially compacted using a floating die at 700MPa. The dewaxing process was performed at 550°C for 60 minutes in a nitrogen atmosphere. Samples were then sintered in a vacuum furnace with argon backfilling at temperature 1250°C for 60 min. After sintering rapid cooling were applied using nitrogen under pressure of 0.6MPa with a cooling velocity of 6ºC/s calculated in the range of 1250-400°C. The sintered stainless steel samples reached ~7,2 g/cm3.

Table 1. Chemical composition of investigated sintered stainless steel

Element concentration, wt. % CrE NiE CrE/NiE

Ni Cr Si Cu Mn Mo C

8.10 22.72 0.70 - 2.00 0.60 0.03 29.8 9.03 3.20

Microstructure observations were carried out using a light microscope and scanning electron microscopy equipped in EDS probe. Evaluations of the phase composition were made using X-ray diffractometer with the filtered copper lamp rays CuK . Density of sinters was evaluated using the water displacement method. Mechanical properties were evaluated based on the tensile test performed according to EN 10002-1 standard on samples prepared according to ISO 2740 standard, the “dog-bone” samples and the Charpy impact test were performed on unnotched samples according to EN 10045. The fatigue test samples were sintered according to ISO 3928. Fatigue tests were carried out in symmetric plane bending at the stress ratio R = – 1 with frequency of about 24Hz using SCHENCK POWN testing machine. The maximum number of cycles was 107. Batches of 20 specimens were tested. Fatigue crack propagation micromechanisms were investigated by means of a scanning electron microscope fracture surface analysis. The profile surface roughness Ra was measured by a profilometer of Taylor-Hobson Sutronic 25 on as sintered samples.

3. Results and discussion

The X-ray qualitative analyses confirmed that the structure of obtained sintered steels consists of austenite and ferrite phases (fig. 1). The Averbach and Cohen method was used to calculate the quantity of individual structural components in the microstructure. The phase content of sintered samples was 46% of ferrite and 54% of austenite.

According to metallographic examinations the presence of a fine microstructure with no presence of precipitates can be seen (Fig. 2). The lack of precipitations shows that applied technology and the way of achieving mixtures results in the right microstructure. Austenite and ferrite are well mixed with an observed balance between these two structures present in the sample. The mechanical properties of sintered stainless steel are shown in Table 2.

a) b)

Figure 1. a) X-ray diffraction pattern, b) microstructure of sintered duplex stainless steel

0

200

400

600

800

1000

1200

1400

1600

1800

40 44 48 52 56 60 64 68 72 76 80 84 88 92 96 100

Reflection angle 2 θ

Inte

nsi

ty[C

PS

]

Fe

(

20

0)

α

Fe

(21

1)

α

Fe

(22

0)

α

Fe

(11

0)

α

Fe

(111

)γ

Fe

(20

0)

γ

Fe

(2

22)

γ

Fe

(22

0)

γ

Fe

(31

1)

γ

200 µm

424 T. Tański et al. / Procedia Engineering 74 ( 2014 ) 421 – 428

Table 2. The mechanical properties of sintered duplex stainless steel

Hardness / HRA UTS / MPa YS 0.2% / MPa El. / % IE, J Ra / μm

43 500 301 16.3 135±15 3.0±0.5

The Staircase method [11,12] applied in the present study to estimate fatigue strength of sintered steel use a simple protocol in which a specimen is tested at a given starting stress for a specified number of cycles or until failure, whichever comes first. If the specimen survives, the stress level is increased for the next specimen; likewise the stress is decreased if the specimen fails. This protocol is continued for a batch of specimens with Dixon and Mood’s equations applied to the results in order to estimate the mean fatigue strength and its standard deviation at the specified number of cycles. The method is remarkably accurate and efficient in terms of quantifying the mean fatigue strength. Unfortunately, it is difficult in practice to provide accurate estimates of the standard deviation of the fatigue strength using this method for small-sample test programs typical of high-cycle or ultra-high-cycle fatigue testing. The Dixon and Mood method, based on the maximum likelihood estimation (MLE), provides approximate formulas to calculate the mean ( S,FL) and the standard deviation ( S,FL) of a fatigue limit. The method assumes that the fatigue limit follows a normal distribution. The two statistical properties are determined by either using only the failures or only the survivals, dependent on the least frequent event that had the smaller total numbers. The stress levels S spaced equally with a chosen increment Sd are numbered i where i=0 for the lowest stress level S0. Note that the fixed stress increment should be in the range of half to twice the standard deviation of the fatigue limit. Denoting by ni the number of the less frequent event at the stress level i, two quantities A and B are calculated:

(1)

(2)

The calculation of the mean value is given by:

(3)

where the plus sign (+) is used if the more frequent event is survival and the minus (-) sign is used if the more

frequent event that is the failure. The standard deviations are then estimated as:

(4)

or

(5)

The first specimen (1C) has been tested at 120 MPa and after 1,E+07 cycles it didn’t break. After the first specimen the amplitude stress has been increased up to 180 MPa where the specimen (C4) didn’t survive. Subsequently, in according to the Staircase method, the stress level was decreased if the specimen fails or was increased if the specimen survives. The test results have been reorganized following the schema up-and-down (Table 3) and the number of the less frequent event at the stress level i and two quantities A and B were estimated (Table 4). The data test of specimens 1C (stress level i = 120MPa) and 4C (stress level i = 180MPa) wasn’t take into account because the first one has the level stress to low and the second one has the level stress too high to be included in evaluation. The total number of tests specimens included in the Staircase method was 18. The total broken specimens were 10, the total survived specimens was 8 and the number of the less frequent event was 8. The lowest initial level stress S0 was then 140 MPa and the stress increment Sd was 10 MPa. Following the Staircase

425 T. Tański et al. / Procedia Engineering 74 ( 2014 ) 421 – 428

method and using the formula (3) the mean fatigue strength coefficient equal to S,FL = 157,5 MPa and its standard deviation S,FL = 0,4375 of the fatigue limit were calculated for sintered duplex stainless steel.

Table 3. The staircase test data evaluation

Result MPa Number of specimens

Broken=1 Survived=0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

3 0 170 1 1 1

5 3 160 0 1 1 0 0 1 1 1

2 4 150 0 0 0 0 1 1

0 1 140 0

The fatigue strength for austenitic stainless steel, as a group, is typically about 35% of the tensile strength, while for ferritic stainless steel is about ~50 to 60%. In case of duplex stainless steels the results published by different steel producers [13] conclude that as a general rule duplex stainless steels have fatigue limits in air around their tensile 0.2% proof strength levels (YS 0.2%). Reported values refer to polished samples, and as it is well known fact that the increased smoothness of surface improves strength. The fatigue resistance is dependent on the surface finish, generally smoother un-notched designs and surfaces are beneficial. The notch sensitive materials are more prone to fatigue failure, so austenitic has a higher fatigue resistance that ferritic stainless steel. The higher strength of the duplex types may account for better notch sensitivity resistance compared to the ferritics.

The relationship between fatigue and tensile strengths expressed as a ratio of fatigue strength to ultimate tensile strengths of wrought stainless steels are close to 0.3-0.35 and as a general rule for many materials is close to 0.38. For sintered materials, however this ratio can vary widely from 0.16 to 0.47 [14]. The prediction of minimum fatigue strength from tensile strength using a value of 0.38 of fatigue to tensile strength ratio is only practicable for ferritic sintered stainless steels having sintered densities at or above 7.0 g/cm3. Such prediction is not applicable for sintered austenitic stainless steels that shows lower ratio attributed to higher rates of strain hardening [15]. For studied sintered duplex stainless steel calculated mean fatigue strength ( S,FL) can be adopted, thus divided by tensile strength (UTS) gives a ratio of 0.31.

The fatigue strength is influenced by surface finish, so the fatigue endurance limit should be multiplied by a suitable factor dependent on surface finish treatment. For polished surface the coefficient of surface finish is equal 1.0, for ground surface 0.9, for turned or cold drown surface is 0.65-0.7, for hot rolled surface 0.35-0.45 and forged surface 0.25-0.3. Taking in consideration the surface condition of samples that were tested in as sintered state characterized by high surface roughness, the coefficient of surface finish should be around 0.5-0.7. Basing on this assumption if sintered test samples will be polished the fatigue strength of sintered duplex stainless steel should be higher that mean fatigue strength S,FL = 157,5 MPa.

Table 4. Number of the less frequent event at the stress level i and two quantities A and B

Stress level i Number of the less frequent event N = ni A = i × ni B = i²× ni

3 0 0 0

2 3 6 12

1 4 4 4

0 1 0 0

N = 8 A = 10 B = 16

The fracture analysis was performed on selected samples after fracture. The specimen 6C was chosen formicrostructural analysis because it breaks down at low cycle number exactly 4,E+06 cycles at 160 MPa. The fracture surface analysis of this sample generally revealed one distinct crack growth regions that nucleated near the sample surface. Crack nucleation occurred within regions that had local high porosity and weakly-bonded particles.

426 T. Tański et al. / Procedia Engineering 74 ( 2014 ) 421 – 428

The crack initiation occurred probably in the top right part of the specimen. Fast fracture was mainly ductile rupture (top right and centre) with isolated areas of cleavage fracture (Fig. 2). Fatigue striations were observed in the crack growth regions of the fracture surface. The example of fracture surface with fatigue striations on cleavage (Fig. 3) Was for example observed in sample 8C that braked down at over 3,E+06 cycles at 160 MPa.

a) b)

Figure 2. Fracture surface of sintered stainless steel, a) the initiation area and beginning of fatigue crack propagation (top right corner of the sample), b) surface with fatigue striations on cleavage, sample braked down after 4,E+06 cycles at 160 MPa

a) b)

Figure 3. Fracture surface of sintered stainless steel, a), b) the surface with fatigue striations, sample braked down after 3,E+06 cycles at 160 MPa

a) b)

Figure 4. Fracture surface of sintered stainless steel, a) the impurities and oxides between original powder particles, b) the surface with brittle fracture and fatigue striations, sample braked down after 3,18E+06 cycles at 160 MPa

The fatigue behaviour of sintered stainless steel depends on the plastic and strength properties of the microstructure as well as porosity. The fracture surface analysis resulted in both ductile fracture and fragile striations (Fig. 2 and 3) that probably may correspond to austenite and ferrite zones respectively. The interconnected

60 µm 150 µm

100 µm 100 µm

100 µm 80 µm

427 T. Tański et al. / Procedia Engineering 74 ( 2014 ) 421 – 428

porosity decreases the area of sample cross-section resulting in lower load-bearing cross-section. The pores acts as a stress concentration sites, where the degree of local stress increase depends on pore geometry, the distance between pores, local stress direction and their interaction [16]. Moreover, the powder particle surface may be covered by oxides and impurities that lower the surface area between adjacent particles during sintering, because they are entrapped in the sinter necks (Fig. 4a). Such mechanism encourages the nucleation of firs cracks and provide easy path for crack growth and propagation between original powder particles. The surface fracture analysis of all samples evidenced that crack propagation implies the an evident striation on cleavage (Fig 4b) and microdimples in plastic zones of the duplex microstructure (Fig. 4a).

4. Conclusions

The vacuum sintering followed by rapid cooling (sinter-hardening method) ensure duplex microstructure of well mixed grains and balance between phases. The sintered duplex stainless steel shows good mechanical properties in terms of tensile strength as well as plastic elongation. For fatigue strength testing the Staircase method was applied. Following the Staircase method the mean fatigue strength coefficient equal to S,FL = 157,5 MPa and its standard deviation S,FL = 0,4375 of the fatigue limit were calculated for sintered duplex stainless steel. The micromechanism of fracture in sintered duplex stainless steel mainly imply brittle cleavage and ductile microdimples and plastic deformation by slip. Fatigue striations were observed in the crack growth regions of the fracture surface.

Acknowledgements

The authors gratefully acknowledge to Eng. Dario Pezzini from Politecnico di Torino, Sede di Alessandria, Italy for his valuable support in fatigue testing. This research was financed with the support of the National Science Centre under grant decision number DEC-2011/01/B/ST8/06648.

References

[1] X. Chen, X. Ren, H. Xu, J. Tong, H. Zhang, Effect of superplastic deformation on the bonding property of 00Cr25Ni7Mo3N duplex stainless steel, International Journal of Minerals, Metallurgy and Materials, Vol. 19, Number 4, Apr. 2012, pp. 317-321.

[2] Y. Bao, X. Song, R. Zhang, et al., Characterization of microstructure, mechanical properties and corrosion resistance of welded joints of 2205 duplex stainless steel, TMS 2010 139th Annual Meeting & exhibition, Seattle, WA, Feb. 14-18, 2010, Supplemental Proceedings on Materials Processing and Properties, Vol. 1: Materials processing and properties, pp. 187-195.

[3] I. Mutlu, E. Oktay, Corrosion behaviour and microstructure evolution of 17-4 PH stainless steel foam, Corrosion Reviews, Vol.30 (2012) Issue 3-4, pp. 125-133.

[4] T. Mesquita, E. Chauveau, M. Mantel, et al., Lean duplex stainless steels - The role of molybdenum addition on pitting corrosion of concrete reinforcements, Revue De Metallurgie-Cahiers D Informations Techniques, Vol. 108, Issue 4 (2011), pp. 203-211.

[5] Z. Brytan, M. Actis Grande, M. Rosso, R. Bidulsky, L.A. Dobrzanski, Stainless Steels Sintered form the Mixture of Prealloyed Stainless Steel and Alloying Element Powders, Book Series: Materials Science Forum, Vol. 672 (2011), pp. 165-170.

[6] C. Moral, A. Bautista, F. Velasco, Aqueous corrosion behaviour of sintered stainless steels manufactured from mixes of gas atomized and water atomized powders, CORROSION SCIENCE, Vol. 51 Issue 8, (2009) pp.1651-1657.

[7] A. Dudek, R. Wlodarczyk, Z. Nitkiewicz, Structural Analysis of Sintered Materials Used for Low-temperature Fuel Cell Plates, Thermec 2009, PTS 1-4 Book Series: Materials Science Forum Vol.638-642, Part: 1-4, (2010) pp. 536-541.

[8] F. Martin, C. Garcia, Y. Blanco, Effect of chemical composition and sintering conditions on the mechanical properties of sintered duplex stainless steels, Materials Science And Engineering A – Structural Materials Properties Microstructure And Processing, Vol. 528, Issue 29-30 (2011), pp. 8500-8511.

[9] J. A. Cabral-Miramontes, J. D. O. Barceinas-Sanchez, C. A. Poblano-Salas, et al., Corrosion Behavior of AISI 409Nb Stainless Steel Manufactured by Powder Metallurgy Exposed in H2SO4 and NaCl Solutions, International Journal of Electrochemical Science, Vol. 8, Issue 1 (2013), pp. 564-577.

[10]. W. Fredriksson, D. Petrini, K. Edstrom, et al., Corrosion resistances and passivation of powder metallurgical and conventionally cast 316L and 2205 stainless steels, Corrosion Science, Vol. 67 (2013), pp. 268-280.

[11]. S.K. Lin, Y.L. Lee, M.W. Lu, Evaluation of the staircase and the accelerated test methods for fatigue limit distributions, International Journal of Fatigue 23 (2001), pp. 75-83.

[12]. R. Pollak , A. Palazotto, T. Nicholas, A simulation-based investigation of the staircase method for fatigue strength testing, Mechanics of Materials 38 (2006), pp. 1170-1181.

[13]. Outokumpu brochure, Duplex stainless steel, 1528EN-GB:1, December, 2013.

428 T. Tański et al. / Procedia Engineering 74 ( 2014 ) 421 – 428

[14]. R. O'Brien, Fatigue properties of P/M materials, SAE Technical Paper 880165, 1988, doi:10.4271/880165. [15]. E. Klar, P.K. Samal, Powder metallurgy stainless steels: Processing, microstructures, and properties. Materials Park, Ohio: ASM

International, 2007. [16]. R. Bidulsky, J. Bidulska, M. Actis Grande, Correlation between microstructure/ fracture surfaces and mechanical properties, Acta Physica

Polonica A, Vol. 122 (2012), pp. 548-552. [17]. L.A. Dobrza ski, W. Borek, Thermo-mechanical treatment of Fe-Mn-(Al, Si) TRIP/TWIP steels, Archives of Civil and Mechanical

Engineering Vol. 12 (3) (2012) pp. 299-304. [18]. L.A. Dobrza ski, W. Borek, Hot-rolling of advanced high-manganese C-Mn-Si-Al steels, Materials Science Forum, Vol. 706/709 (2012) pp.

2053-2058. [19]. T. Ta ski, K. Lukaszkowicz, Structure and properties of PVD coatings deposited on the aluminium alloys, Surface Engineering, Vol. 28/8

(2012) pp. 598-604. [20]. T. Ta ski, Characteristics of hard coatings on AZ61 magnesium alloys, Journal of Mechanical Engineering, Vol. 59/3 (2013) pp.165-174. [21]. L.A. Dobrza ski, K. Labisz, E. Jonda, A. Klimpel, Comparison of the surface alloying of the 32CrMoV12-28 tool steel using TiC and WC

powder, J. Mater. Process. Technol. Vol. 191/1/3 (2007) pp. 321-325. [22]. L.A. Dobrzanski, K. Labisz, A. Klimpel, Comparison of mechanical properties of the 32CrMoV12-28 hot work tool steels alloyed with WC,

VC and TaC powder using HPDL laser, Key Engineering Materials, Vol. 324-325 (2006) pp. 1233-1236. [23]. L.A. Dobrza ski, T. Ta ski: Influence of aluminium content on behaviour of magnesium cast alloys in bentonite sand mould, Solid State

Phenomena, Vols. 147-149 (2009), pp. 764-769.