Scholarly article on topic 'Recent developments in small punch testing: Tensile properties and DBTT'

Recent developments in small punch testing: Tensile properties and DBTT Academic research paper on "Materials engineering"

CC BY
0
0
Share paper
OECD Field of science
Keywords
{"Small punch testing" / "Yield stress" / "Ultimate tensile strength" / "Ductile to Brittle Transition Temperature (DBTT)" / "Miniature testing techniques"}

Abstract of research paper on Materials engineering, author of scientific article — M. Bruchhausen, S. Holmström, I. Simonovski, T. Austin, J.-M. Lapetite, et al.

Abstract Neutron irradiation and temper embrittlement in nuclear power plants (NPPs) lead to microstructural changes in structural materials which induce a shift of the ductile to brittle transition temperature (DBTT) towards higher temperatures. Monitoring of the DBTT in NPP components receives therefore considerable attention — in particular in the context of long term operation. In that context small specimen testing techniques are developed for characterizing structural materials with a limited amount of material. One of the most used of these miniature testing techniques is the small punch (SP) test which is based on disc shaped specimens. Although SP testing has been used for more than 30years, there is still no commonly agreed procedure for deriving basic material properties from SP test data. We describe the current status of the SP test with regard to data evaluation procedures for obtaining yield stress, ultimate tensile strength and DBTT from SP tensile/fracture data. The methods for deriving the quantities characterizing the SP force-deflection curve and their use for determining basic mechanical properties are discussed. Possible reasons for the difference between the DBTT determined from Charpy and SP tests are presented. Data from the present study as well as from the literature suggest that neither notch nor strain rate effects can explain the observed discrepancies. Based on data from ongoing research projects the importance of Finite Element Analysis (FEA) for studying SP tests is presented for the example of tube specimens derived from fuel claddings. Finally an overview over the currently available standards and standardization developments is given.

Academic research paper on topic "Recent developments in small punch testing: Tensile properties and DBTT"

Accepted Manuscript

Recent developments in small punch testing: Tensile properties and DBTT

M. Bruchhausen, S. Holmström, I. Simonovski, T. Austin, J.-M. Lapetite, S. Ripplinger, F. de Haan

PII: S0167-8442(16)30167-7

DOI: http://dx.doi.org/10.1016/j.tafmec.2016.09.012

Reference: TAFMEC 1772

To appear in: Theoretical and Applied Fracture Mechanics

Please cite this article as: M. Bruchhausen, S. Holmström, I. Simonovski, T. Austin, J.-M. Lapetite, S. Ripplinger, F. de Haan, Recent developments in small punch testing: Tensile properties and DBTT, Theoretical and Applied Fracture Mechanics (2016), doi: http://dx.doi.org/10.10167j.tafmec.2016.09.012

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Recent developments in small punch testing: tensile properties and DBTT

M. Bruchhausena*, S. Holmstroma, I. Simonovskia, T. Austina, J.-M. Lapetitea, S. Ripplingera, F. de Haan1

aEuropean Commission, Joint Research Centre (JRC), Westerduinweg 3, 1755 LE Petten, The Netherlands

ostructural changes in )BTT) towards higher tem-: attention — in particular in re developed for characterizing

Abstract

Neutron irradiation and temper embrittlement in nuclear power plants (NPPs) lea structural materials which induce a shift of the ductile to brittle transition temper peratures. Monitoring of the DBTT in NPP components receives therefore cor the context of long term operation. In that context small specimen testing techniq structural materials with a limited amount of material.

One of the most used of these miniature testing techniques is the small punch (SP) test which is based on disc shaped specimens. Although SP testing has been used for more than 30 years, there is still no commonly agreed procedure for deriving basic material properties from SP test data. We describe the current status of the SP test with regard to data evaluation procedures for obtaining yield stress, ultimate tensile strength and DBTT from SP tensile/fracture data. The methods for deriving the quantities characterizing the SP force-deflection curve and their use for determining basic mechanical properties are discussed.

Possible reasons for the difference between the DBTT determined from Charpy and SP tests are presented. Data from the present study as well as from the literature suggest that neither notch nor strain rate effects can explain the observed discrepancies.

Based on data from ongoing research projects the importance of Finite Element Analysis (FEA) for studying SP tests is presented for the example of tube specimens derived from fuel claddings. Finally an overview over the currently available standards and standardization developments is given.

Keywords: Small Punch Testing, yield stress, ultimate tensile strength, Ductile to Brittle Transition Temperature (DBTT), miniature testing techniques

1. Introduction

During the investigation of irradiated materials from fission and fusion programs limiting the exposure of the experimentalists to irradiation is a high priority. Consequently the use of miniature specimens receives significant attention in the nuclear community. The high cost of irradiation experiments is a further incentive for using small specimen testing techniques. The Small Punch (SP) test initially developped in the U.S. and Japan in the 1980s is one of these miniature testing techniques.

In an SP test a small hemispherical tip or a ball ("punch") is pushed through a disc-shaped specimen along its axis of symmetry. SP tests can either be carried out as creep tests, where a constant force is applied

* Corresponding author

Email address: matthias.bruchhausen@ec.europa.eu ( M. Bruchhausen)

and displacement is measured as a function of time, or as tensile/fracture tests, where a constant displacement rate is applied to the punch and the force is measured as function of time [1].

At the beginning of the development two specimen thicknesses of 0.25 mm (derived from TEM specimens) [2-4] and 0.5 mm [5-7] have been used. Although both geometries are still in use today [8], the 0.5 mm thickness specimens are more common.

The triaxial, time dependent stress state in the specimen and the sensitivty of the test geometry make determining even basic mechanical properties from SP testing a challenge. Although significant effort has already been put into deriving mechanical properties from SP data, the evaluation of SP tests is still a topic of research [8-13].

The current paper describes recent developments related to SP testing with a focus on the determination of tensile material properties and the ductile to brittle tran-

Preprint submitted to Theoretical and Applied Fracture Mechanics

September 15, 2016

sition temperature (DBTT). The current situation with regard to international standards is also reviewed.

2. Small Punch tensile tests

2.1. Set-Up

A typical scheme of an SP test setup is shown in Fig. 1. The disc shaped specimen (in red) is clamped between two dies. In an SP tensile test, the punch is pushed with a constant displacement rate through the specimen. Fig. 1 shows a solid punch in a single piece as recommended in the current CEN Workshop Agreement (CWA) [1] and used by many authors [10, 12, 14, 15]. An alternative, frequently used configuration is based on a punch with a flat or concave tip pushing a ball through the specimen [7, 13, 16]. The latter configuration has the advantage that the tip (i.e. the ball) can be replaced after every test. Changing the ball after each test avoids potential problems caused by wear of the punch which might lead to less reproducible results.

The force needed to push the punch through the specimen is recorded and plotted as a function of either the displacement of the punch tip/ball or — as shown in Fig. 1 and recommended in the CWA [1] — as a function of the specimen deflection measured on the specimen surface opposite to the point of contact between the punch and the specimen. The displacement cannot be measured directly at the tip of the punch but has to be inferred either directly from the cross-head displacement or from a clip gage or a similar device attached to the push rod or the punch. In both cases the displacement signal has to be corrected for force line compliance.

This problem does not occur when the specimen deflection is used instead of the displacement of the punch. In such a case a rod is touching the specimen from below. The rod simply transfers the deflection of the specimen to an LVDT or a similar device. Since the force applied on the rod is very small, no compliance correction is necessary for the rod. There might be a small compliance from the entire setup though. A hollow ceramic rod can be used which can contain a thermocouple in direct contact with the specimen surface allowing determination of the test temperature.

Ideally, the compliance corrected displacement and the deflection only differ because of the specimen thinning. A detailed discussion of the implications of the different approaches has recently been published [13].

Figure 1: A typical SP test setup. The basic dimensions are listed in Tab. 1

punch diameter = 2 mm du/dt = 0.005 mm/s T = -100 °C

2500 2000 ,1500 ' 1000 500

0 0.5 1 1.5

u [mm]

Figure 2: Typical SP force-deflection curve for a ductile material [17]). The roman numbers indicate the different zones of the curve.

2.2. Characteristics of the force-deflection curve

A typical force-deflection curve for a ductile material is shown in Fig. 2. The force-deflection curve is gen-

Feature Symbol Size

punch diameter d 2.5 mm

diameter of receiving hole (lower die) d2 4 mm

chamfer length (lower die) l 0.2 mm

chamfer angle (lower die) a 45°

specimen diameter d1 8 mm

specimen thickness h 0.5 mm

Table 1: Principal dimensions of the SP specimen and setup according to the CWA 15627, part B [1].

erally divided in different stages [18-21]: zone I corresponds to the indenting of the specimen surface and elastic bending. During zone II plastic bending spreads through the specimen. In zone III the specimen behaviour is dominated by membrane stretching and in zone IV by necking and crack initiation. In zone V fracture softening occurs and final fracture occurs in zone VI.

For the evaluation of an SP tensile test data a num ber of characteristic values determined from the force deflection F(u) curve are used [1, 22] (Fig. 3):

Fm, the maximum force,

um, the deflection at maximum force,

Fe, the elastic-plastic transition force,

• £frac, the fracture energie £frac = f0Ufiac F(u) du.

For the determination of £frac the force F is integrated over the deflection u up to the point ufrac where fracture occurs. Different approaches for defining ufrac are discussed in context with the determination of the DBTT in section 4.

3. Tensile Material Properties

3.1. Correlating uniaxial and SP properties

For determining the yield stress, <y and the ultimate tensile strength, <UTS of a material from a tensile SP test empirical correlations between the material properties and the characteristic points on the force-deflection curve are used (e.g. [10, 16, 19, 23]):

^UTS ^UTS

ai he +a2

Fm hoUm

pi Fm+p2

„1500

Jin...

punch diameter = 2 mm

- du/df = 0.005 mm/s

T = -100 °C

F / e/

/1 I u u e m

u [mm]

Figure 3: Characteristic points in the force-deflection curve [17]).

The correlation factors ai,fii,fi'i depend on the dimensions of the test rig such as punch diameter or diameter of the lower die. Different authors use these correlations with [10, 23] or without [4, 8, 13] the constant terms a2, ¡32 and j3'2. In both cases, however, the normalization of Fm with h0um (i.e. Eqn. 2) is the preferred formulation for <UTS [8, 10].

Normalizing with h2 in the case of <y and h0um in the case of < UTS makes also sense from a physical point of view. Fe is determined at a point of the force-deflection curve dominated by bending deformation. The force required for bending a plate increases quadaratically with its thickness. In contrast, during the stage of the test where Fm is determined the specimen is in a membrane stretching/necking regime, so the force can be expected to depend proportionally on specimen thickness.

3.2. Determination ofFe

While Fm and um can easily be obtained from the force-deflection curve, the elastic-plastic transition point is not as well defined since yielding occurs during the test successively in different areas of the specimen.

Figure 4: Determination of Fe by means of bilinear fitting on a P92 specimen at room temperature. The intital specimen thickness h0 was 0.5 mm.

Figure 5: Determination of Fe by off-set methods on a P92 specimen at room temperature. The slopes of the linear functions where determined from the first part of the bi-linear fit in Fig. 4.

Different proposals for defining and determining Fe are currently discussed [10, 13, 22].

Some methods are based on a bilinear fit of the first part of the force-deflection curve, normally from the start of the test up to a deflection u corresponding to the initital specimen thickness h0. There are two approaches for obtaining the bi-linear curve. The CWA [1] suggests using a function minimizing the error between a bi-linear function f (u) and the force-deflection curve F(u). This corresponds to the situation in Fig. 4. The deflection uA and the values f(uA) and f(h0) are the fitting parameters. In the following this is referred to as the "two secants method". Fe can then be identified either directly by the intersection point of the two linear functions (Fe(two secants), Fig. 4) or its projection on the force-deflection curve as currently recommended in the CWA (Fe(CWA) in Fig. 4). It is clear from Fig. 4 that Fe(two secants) is generally higher than Fe(CWA).

The "two tangents" methods is very similar to the "two secants" method in that it is based on approximating the first part of the force-deflection curve by a bilinear function. Tangents on the force-deflection curve are determined at the start of the test (u = 0) and at the point where u = h0 [4]. The intersection point between the two linear tangents is then used for determining Fe just as in the case of the two secants method.

Determining the two tangents is always to some extent subjective since it requires determining two sections of the force-deflection curve that are sufficiently linear to be approached by tangents. On top of this, the first part of the curve frequently shows some irregularities from the settling in of the punch or other components that make determining the tangent even more difficult. The two secants method in contrast relies on an ob-

jective algorithm based solely on the curve data which makes it more robust. In the current work we therefore give preference to the two secants method over the two tagents method.

As an alternative to these bi-linear methods offset methods are used by some authors [10, 16, 23] where Fe is determined in much the same way as Rp0.2 from uniaxial tensile tests. In these approaches a straight line which is parallel to the slope at the beginning of the force-deflection is drawn through an offset point (e.g. h0/10 or h0/100, see Fig. 5). The intersection of this linear function with the force-deflection curve is then identified with the elastic-plastic transition point Fe.

3.3. Application to <UTS and <y

Eqns. 1 to 3 are empirical correlations which means that the factors ai, anddepend on the individual test rig and are not calculated directly from the geometry of the test setup but derived experimentally. To investigate the transferability of these factors from one test rig to another with nominally equivalent geometry, we have used factors reported in the literature together with test data acquired at the JRC on Gr. 91 steel from the FP7 project MATTER [24] to calculate <UTS and <y. The material was delivered as a 60 mm thick plate produced by Industeel (heat 20057). In order to cover a range of values for <UTS and <y the tests were carried out at temperatures between room temperature (RT) and 650 °C. For these tests the displacement was measured instead of the deflection and a temperature dependend compliance correction was carried out. To minimize problems with with "setting" of the punch at the beginning of the test, the punch was pressed with a force of ca. 30 N on the specimen during heating up. The tests were carried

Factor Literature value [10] Optimized value

ßl 0.277 0.326

ß2 0 -27.04

ßl 0.065 0.093

ß2 268.81 -11.86

Table 2: Factors according to Eqs. 2 and 3 from the literature [10]. Values optimized for the current study are also given.

out on a uniaxial test rig was kept in force control until the final temperature was reached. Access to the test data can be obtained from [25]. Reference data for <UTS and <y were taken from uniaxial tensile tests [26].

The factors taken from the literature are from an extensive study including a wide range of materials [10]. In that study a punch diameter of 2.4 mm was used instead of the 2.5 mm punch in the present study and recommended by the CWA (Tab. 1). Otherwise the test geometries in both studies were compliant with Tab. 1.

Fig. 6 shows the <UTs predicted from SP testing compared to the uniaxial tensile test results from the same material batch. The values for the correlation factors are given in Tab. 2. It can be seen that the pre-defined material constants for Eq. 3 do not produce satisfactory estimates for < UTS, significantly overestimating < UTS at high temperature. However for Eq. 2 the literature factors work well and the estimated < UTS from SP testing is in average less than 7% lower than the value from uniaxial testing over the entire temperature range.

For comparison Fig. 7 shows the <UTs estimate from the same data but with factors andoptimized for the current data set (optimized values in Tab. 2). Optimized fitting factors lead to considerably better agreement between < UTS from uniaxial specimens and from SP specimens. Since all used specimens in the present study had the same intitial thickness h0, the two approaches in Eqs. 2 and 3 effectively only differ in so far as Eq. 2 takes also ductility um into account. In the present case this does not have a major impact but it is significant in the case of less ductile materials.

The large difference between the two estimates in Fig. 6 might be explained by the different punch diameters used in both studies. Normalizing Fm with um seems to correct for this. In fact, using membrane stretching theory [27] it can be calculated that reducing the punch diameter from 2.5 mm to 2.4 mm reduces Fm to 95% and um to 98% of their orginial values. This suggests that the lower Fm for the smaller punch diameter is at least partly compensated by the lower um if Eq. 2 is used.

Figure 6: SP estimates for <UTS with factors from the literature ([10], Tab. 2) against values from uniaxial tensile tests. The two curves refer to normalization of Fm by umh0 (Eq. 2) and by hjj (Eq. 3).

Fig 8 shows for three methods of determining Fe (see Figs. 4 and 5) < y determined from SP data with coefficients from the literature (Tab. 3) against the uniaxial data. The slope of the points largely follows the unity-line but there is a large offset which leads to unsatisfactory estimates. The estimates can be improved significantly by optimizing the paramters ai (Fig 9). There is no large difference between the different methods for determining Fe.

The scatter for the SP estimates of < y is larger than for < UTS. This is consistent with the outcome of the inter-laboratory comparison reported in [8] where in every laboratory the scatter for the determination for Fe was significantly higher than in the case of Fm.

4. Ductile to Brittle Transition Temperature

(DBTT)

One of the drivers leading to the development of the SP technique was the determination of the Ductile to Brittle Transition Temperature (DBTT) and its shift to lower temperatures as a consequence of temper embrit-tlement and neutron irradiation [2, 5, 6]. The standard method for determining DBTT is by means of Charpy impact testing on notched specimens with the dimensions (10 x 10 x 55 mm3). Embrittlement in nuclear power plants (NPPs) is monitored by using dedicated Charpy specimens installed at well defined locations in the plant and having been exposed to the same irradiation and thermal conditions as the actual reactor components. In the context of lifetime extension of NPPs these specimens have become an invaluable asset. Replacing

Method Factor Literature value [10] Optimized value

CWA ai «2 0.476 0 0.382 28.8

two-secants ai «2 0.442 0 0.405 -34.9

ho/10 ai «2 0.346 0 0.288 10.7

Table 3: Literature values for the factors in Eq. 1 for three different methods to determine Fe. Values optimized foi

unity 1/(u h.)

v m 0'

200 200

300 400 uniaxial a

500 600 [MPa]

Figure 7: SP estimates for <xuTs against uniaxial values with factors optimized for the current data, Tab. 2. The two curves refer to normalization of Fm by umh0 (Eq. 2) and by (Eq. 3).

o -i—'

200 200

V two secants

0 h0/10

400 uniaxial a

500 [MPa]

Figure 8: SP estimates for <xy with factors from the literature ([10], Tab. 3) against values from uniaxial tensile tests. Fe was determined by three different methods (see Figs. 4 and 5).

200 200

e current study

are also given.

-unity

CWA v two secants h0/i0

400 uniaxial a

500 [MPa]

Figure 9: SP estimates for <xy against uniaxial values with factors optimized for the current data, Tab. 3. Fe was determined by three different methods (see Figs. 4 and 5).

even some of the usual Charpy tests with SP tests would make more service exposed material available for continued monitoring during a prolonged service life.

For determining the DBTT by SP testing, the fracture energy £frac has to be determined as a function of temperature, where:

Efrac —

rufrac

F(u) du

The integration is carried out from the beginning of the test up to the deflection ufrac where fracture occurs. Several definitions of Ufrac have been used by different authors:

• um, the deflection at the maximum force, has been used as Ufrac [2, 8, 28].

• The current CWA defines Ufrac as the deflection where the force has dropped by 20% after reaching its maximum Fm [1].

punch diameter = 2 mm T = -196

0.4 0.6 u [mm]

Figure 10: Typical SP force-deflection curve for a brittle material [29]). The arrows highlight crack initialisation before the maximum force is reached [6].

These definitions are well established in the case of ductile failure with smooth force-deflection curves like the one in Fig. 3. In the case of brittle fracture, the situation is more complex. Fig. 10 shows an example of brittle failure where different drops of the force occur before the global maximum force is reached. These force drops are indicative for crack initialisation [6]. In these cases other definitions for ufrac have been proposed:

• The "first drop" criterion defines ufrac as the deflection where the first force drop occurs [19, 28].

• In analogy to the 20% force drop rule [1] the consecutive force drops have been summed until the cummulated force drop has reached 20% [14].

When tested at very low temperatures the specimens fracture easily i.e. at low fracture energies. At higher temperatures ductile fracture leads to higher fracture energies. Thus the overal behaviour shows a temperature dependence of Efrac which is very similar to the one observed in Charpy tests with a lower shelf at low temperature a higher shelf at high temperature and a well defined transition temperature in between.

However, numerous studies have shown that the DBTTSP observed from SP testing is significantly lower than the DBTT from Charpy testing. The relationship between these two values is often expressed as [1, 2, 19, 30]:

DBTTSP = a x DBTT

Other authors use a formulation with two parameters [5, 20,31,32]:

DBTT = a* x DBTTSP + ¡3*

Figure 11: Surface image of a notched SP specimen reconstructed from XCT volume data

The reasons for the discrepancy between the DBTTs from Charpy and from SP testing are not fully understood. Possible reasons could be [22]:

1. Size effect: Charpy specimens have characteristic dimensions in the mm-cm range while the thickness of an SP specimen is 0.5 mm or less.

2. Notch effect: The notch in Charpy specimens acts as a stress concentrator not present in SP tests.

3. Strain rate effect: The falling hammer in a Charpy test leads to strain rates that are orders of magnitude higher than in an SP test; the duration of a Charpy test is in the order of 1-10 ms whereas an SP fracture tests takes a few minutes.

4. Loading mode: different amount of triaxility. Stress state in SP is highly bi-axial, stress triaxi-ality is lower compared to Charpy tests.

In the following the effects of the notch and the displacement rate will be looked at more closely.

4.1. Notched Specimens

To investigate a possible impact of the notch on the DBTTSP, a diametrical sharp notch was laser cut in a number of specimens from Gr. 91 steel. The material was originally produced for the FP6 EUROTRANS project (domain 4: DEMETRA) by Industeel, ARCELOR group (batch number S50460). The X-ray Computed Tomography (XCT) technique described in [33] was used for characterizing the notches. The basic notch dimensions are listed in Tab. 4. An XCT image of a notched specimen is reproduced in Fig. 11.

SP tensile tests have been carried out on a series of these notched specimens [34]. A tanh-type function has been fitted to the calculated fracture energies Efrac to compute the DBTT as suggested in [28]:

Eus + Els EUS - ELS t T - DBTT \ Efrac =---+---tanh I

IT - DBTT\ '( 2DB )(7)

Specimen Identifier Depth [urn] Width [um] Angle [◦] Bottom radius [um]

AD-002 110.42 40.03 17.18 5.17

AD-003 114.22 46.11 17.25 4.67

AD-004 106.87 37.43 19.47 5.72

AD-005 115.76 44.63 21.26 6.07

AD-006 128.45 62.00 16.09 6.13

AD-007 112.37 35.99 15.14 6.13

Table 4: Geometric characteristics of six of the notches as determined by XCT. The reported values are the depth of the notch measured from its root to the specimen surface, the full width of the notch close to the specimen surface, the notche's opening angle and its root radius. Each given quantity is the average of three measurements along the notch.

-200 -180 -160

first crack

DBTT Sp = -126 °C E = 0.065 J

-140 -120 -100 t [°C]

Figure 12: DBTTSP determined from notched specimens with a 2 mm punch.

The lower and upper shelf energies ELS and EUS and the temperature parameter AT are the fitting parameters. The computed DBTTSP = -126 °C value coincides very well with the value found with unnotched specimens from the same batch of Gr. 91 (from the FP6 EUROTRANS, domain 4: DEMETRA) which in turn agreed very well with the transition temperature derived from fracture strains [28].

These observations indicate that a diametrical notch does not have a significant impact on the DBTTSP. This conclusion is consistent with observations made on specimens with comparably blunt diametrical EDM notches or with two perpendicular, diametrical scratches [35]. Tests with circular EDM notches did not show an impact of the notch on the transition temperature either [36]. However, another study came to the conclusion that an 0.25 mm wide and 0.2 mm deep dimateri-cal notch led to a significant rise in DBTTSP [37]. The different materials and notch geometries used in the investigations might be the reason for these contradicting observations.

It is important to note that the tests on the notched specimens reported in the current work as well as those in [28] were carried out with punches with a tip diameter of 2 mm instead of the more frequently used 2.5 mm. The different punch diameters might explain the relatively high DBTTSP found in these studies compared to other SP studies on Gr. 91 [8, 38]. The variation in fracture behaviour in the transition region observed for different punch diameters (brittle for 2 mm and ductile for 2.5 mm [28, 39]) confirms this hypothesis. However, tests from another study did not show any difference in the DBTTSP determined with 2.0 mm and 2.5 mm punches [37].

4.2. Displacement Rate

Another major difference between the Charpy tests and an SP tensile test is the strain rate. A Charpy specimen is several mm thick and the test lasts 1-10 ms. The maximum strain rate in a Charpy test is typically in the order of 103 s-1 [40].

An SP specimen is 0.5 mm thick and the test takes a few minutes to complete. The maximum strain rate ¿mPax can be estimated from the displacement rate of the punch v [1]:

¿mpax = 1000 m-1v where v is in [m]

For a typical displacement rate of v = 0.008 mm/s this leads to a maximum strain rate of 0.008 s-1. Although neither of the tests is carried out at constant strain rate conditions, it is clear that the strain rate in a Charpy test is several orders of magnitude higher than in an SP tensile test.

To address this question SP tests with varying displacement rates have been carried out. In one study on Gr. 91 steel no significant difference in the DBTTSP was found when the displacement rate was increased from 0.005 mm/s to 0.5 mm/s [28]. In another study on a reactor pressure vessel steel the punch displacement rate was modified from 0.005 mm/s to 100 mm/s. In this

study a shift of the DBTTSP of 31 °C was detected. However, the higher displacement rate led to a lower DBTTSP. This unexpected observation is attributed to an adiabatic heating effect [35]. It does not explain the lower DBTTSP compared to the DBTT determined from Charpy tests.

5. Simulation

During SP tests the deformation state in the specimen changes continuously. Consequently, it is not straightforward to extract tensile material properties from a force-deflection curve. Numerical tools such as Finite Element Analysis (FEA) are often used to get better insight into the test [3, 10, 36], including evaluation of applicable theoretical equations [16], crack propagation [11], creep [41] and the significance of specimen displacement definition [13].

A further example, for the application of FEA to SP testing is the recent development of a model for tube specimens (Fig. 13), representing specimens from nuclear fuel cladding tubes [42]. Comparison of the tu specimen with the flat ones shows that a marginall higher maximum force (Fm) is obtained for the tube specimens. However, displacements at Fm for the tube specimen are significantly lower, cf. Fig. 14 and Fig. 15. This is attributed to the curvature of the specimen. Friction influences the force-displacement response only after approximately half way into the Zone II, in line with [13, 21]. For both, flat and tube specimens, friction linearly increases Fm. Good agreement between the simulation and experimental results is achieved. The best agreement is obtained with a friction coefficient of 0.2. In the presented cases a simple elastic-plastic material model of Gr. 91 material is used without incorporating damage, Fig. 16. Sometimes the literature reports that more advanced material models (e.g. with damage) need to be incorporated in order to obtain good Fm estimates [16]. This is logical since the amount of plastic deformation during the SP test can be significant. However, in [42] even a simple elastic-plastic material model resulted in Fm forces comparable to the experimental data.

The ductility exhaustion parameter A can be used to estimate where the creep damage corresponding to inservice multi-axial conditions occurs [43]. A is based on the Rice-Tracey [44] rigid plastic deformation model for growth of voids under a triaxial field of stress. To estimate its applicability to SP test where a strong bi-axial stress state occurs, maximum values of A parameter in the whole specimen are calculated. Usually, AM, based on the maximal principal strain Eq. (9), is used. An al-

Figure and ba

SP tube specimen (blue), dies (red and grey)

ternative definition AEq, based on the equivalent strain (10), is proposed here Eq. (11).

^Max.principal

1.65 ■ e

-1.5 -

£Eq = -3"[(e11 - e22)2 + fe2 - e-33)2 + (10)

(£33 - en)2 + 6^ + 6^23 + 6^]

1.65 ■ e

A maximum value of 1.0 indicates a plastic collapse of a equivalent uniaxial test specimen according to the Rice-Tracey [44]. Points where AM and AEq reach a value of 1.0 are indicated in Fig. 14 and Fig. 15 with * and ° markers, respectively. Comparing the two figures, AM=1.0 correlates well with the failure of the SP specimens while AEq=1.0 correlates well with highest force Fm in the force-displacement curves of the tubular specimens. For flat specimens AEq=1.0 values are obtained slightly prior to Fm. The points AEq=1.0 lie within zone IV of the force-displacementT curve where cracking starts in ductile materials [21].

6. Standards

Currently there is no international standard covering the most widely used applications of SP testing. A

1800 1600 1400 1200 ! 1000 800 600 400 200 0

Displacement [mm]

Figure 14: Simulated and experimental force-displacement curves of flat specimens. Markers indicate where AM=1.0 (* markers) and AEq=1.0 (° markers).

Figure 1

1800 1600 1400 1200

¡I 1000

£ 800 600 400 200 0

FEM: tube, h0=0.45mm, ^=0.001 FEM: tube, h0=0.45mm, ^=0.1 FEM: tube, h0=0.45mm, ^=0.15 FEM: tube, h0=0.45mm, ^=0.2 Exp. 1: tube, h0=0.45mm Exp. 2: tube, h0=0.45mm Exp. 3: tube, h0=0.45mm

Displacement [mm]

rimental force-displacement curves of icate where AM=1.0 (* markers) and

Figure 15: Simulated tubular specimens. AEq—1.0 (o marl

Japanese standard exists but it is limited to creep testing and only a small part is available in English [45].

In the U.S. SP standards exist for characterizing materials used in surgical implants by tensile SP tests at room temperature (ASTM F2183-02, ASTM F2977-13). For material characterization they recommend using quantities derived from the force-displacement curves (i.e. Fe, Fm, Efrac) that are also used within the structural engineering community. However, they focus on assuring reproducibility and ranking but do not provide tools to derive material properties from SP testing. Other issues such as DBTT and creep testing that may be relevant towards the power generation or nuclear industries are not covered. ASTM E 643-15 is a

6: Gr. 91 tensile curve.

standard for testing metallic sheet material by means of a ball punch deformation test. The test is very similar to a small punch test. It covers the specimen thicknesses from 0.2 mm to 2 mm which includes the thickness of

ISP specimens. However, the specimen diameter has a mimimum width of 89 mm and the punch has diameter of 22.2 mm. The test is used for charaterizing the deformability of metallic sheets but does not cover the derivation of basic material properties. Current activities under the auspices of ASTM Subcommittee E10.02 (Behavior and Use of Nuclear Structural Materials) go in that direction [46].

The most recent European standardization document on SP testing is the CEN workshop agreement (CWA) from 2007 [1]. A CWA is a pre-normative document agreed upon by the participants in a CEN workshop. It is not voted by the CEN members and is not a standard but meant to prepare the future development of a standard. A proposal for developing an SP standard within CEN has recently been accepted and introduced in the working programme as work item (WI) EC101162. A new working group (WG) will be installed within ECISS/TC 101 (Test methods for steel (other than chemical analysis)) to draft an EN standard on SP testing [47]. The standard is expected to cover tensile/fracture as well as creep testing and to include TEM specimens (0.25 mm thickness) as well as the more commonly used specimens with 0.5 mm thickness.

To ease the collection and exchange of data, the new EN standard will include a section dedicated to data formats. This part of the activity will build on a series of CEN Workshops on formats for engineering materials data. Given the lack of any widely adopted tech-

nology for exchanging engineering materials data, the CEN Workshops rely on existing documentary testing and product standards from which to derive data models and accompanying formats. To date, the CEN Workshops have delivered data formats for ambient temperature tensile testing (based on ISO 6892 Part 1) and materials pedigree data [48]. Ongoing [49] and future CEN Workshops will extend the test type coverage to fatigue (ISO 12106), uniaxial creep (ISO 204), creep crack growth (ASTM E1457), creep-fatigue (ASTM E2714-13), and creep-fatigue crack growth (ASTM E2760-10). Whereas the CEN Workshops are focusing on existing testing and product standards, the development of the small punch data formats will be integral to the development of the testing standard. At a time when all aspects of engineering materials manufacture and qualification rely on digital systems, the parallel development of the standard testing procedure and accompanying data formats will set a precedent for the way in which mechanical testing standards could (and perhaps should) be developed.

foster the exchange of data and help create synergies between different organizations.

7. Summary and Outlook

The SP testing technique was initially developed as a small specimen technique for the characterization of irradiated materials of the fission and fusion programs with regard to their mechanical properties and in particular the shift of the DBTT. However, it has become a technique that is used for a wide range of material (life) assessment and characterization tasks in the power industry. During the test, the specimen is in a time-dependent, triaxial stress state which makes correlating the results from SP testing to those from standard uniaxial tests challenging.

In the last years significant progress has been made and some material properties such as the ultimate tensile strength can be determined reliably from SP tensile data. On other issues like the determination of yield stress and the reliable transfer of the DBTT determined from SP tests and established standard test still needs further research.

Finite element analysis is expected to be an essential tool for the further development of the technique as it gives insight into the test method itself at a higher level of detail than can be achieved experimentally.

The SP technique is more sensitive to the geometry of the test rig than established techniques using larger specimens. Establishing international standards for SP testing is therefore necessary to ensure comparability of test results between different organizations. Harmonizing data formats by including them in the standard will

8. Acknowledgement

[2] T. Mis tests it ated fe

[3] M. M;

The research leading to these results was partly funded by the European Atomic Energy Community's (Euratom) Seventh Framework Programme FP7/2007-2013 under grant agreement No. 269706 (MATTER project) and carried out in the framework of the EERA (European Energy Research Alliance) Joint Programme on Nuclear Materials.

9. References

[1] Small Punch Test Method for Metallic Materials, CEN Work-;reement, CWA 15627:2007 E.

T. Adachi, M. Saito, Y. Hamaguchi, Small punch or evaluating ductile-brittle transition behavior of irradi-ferritic steels, J. Nucl. Mater. 150 (1987) 194-202. Manahan, A. Argon, O. Harling, The development of a miniaturized disk bend test for the determination of postirradiation mechanical properties, J. Nucl. Mater. 103 & 104 (1981) 1545-1550.

[4] X. Mao, H. Takahashi, Development of a further-miniaturized specimen of 3 mm diameter for TEM disk (diameter 3 mm) small punch tests, J. Nucl. Mater. 150 (1) (1987) 42 - 52. http://dx.doi.org/10.1016/0022-3115(87)90092-4.

[5] J.-M. Baik, J. Kameda, O. Buck, Small punch test evaluation of intergranular embrittlement of an alloy steel, Scripta Matallur-gica 17 (1983) 1443-1447.

[6] J. Kameda, O. Buck, Evaluation of the ductile-to-brittle transition temperature shift due to temper and embrittlement and neutron irradiation by means of a small-punch test, Materials Science and Engineering 83 (1986) 29-38.

[7] G. Lucas, The development of small specimen mechanical test techniques, J. Nucl. Mater. 117 (1983) 327-339.

[8] E. Altstadt, H. Ge, V. Kuksenko, M. Serrano, M. Houska, M. Lasan, M. Bruchhausen, J.-M. Lapetite, Y. Dai, Critical evaluation of the small punch test as a screening procedure for mechanical properties, J. Nucl. Mater. 472 (2016) 186-195. doi: http://dx.doi.org/10.1016/j.jnucmat.2015.07.029.

[9] R. Lacalle, J. Alvarez, F. Guiterrez-Solana, Analysis of key factors for the interpretation of small punch test results, Fatigue & Fracture of Engineering Materials & Structures 31 (2008) 841849.

[10] T. Garcia, C. Rodriguez, F. Belzunce, C. Suarez, Estimation of the mechanical properties of metallic materials by means of the small punch test, Journal of Alloys and Compounds 582 (2014) 708-717.

[11] C. Soyarslan, B. Gtilfimen, S. Bargmann, P. Hahner, Modeling of fracture in small punch tests for small- and large-scale yielding conditions at various temperatures, International Journal of Mechanical Sciences 106 (2016) 266 - 285. doi:http: //dx.doi.org/10.1016/j.ijmecsci.2015.12.007.

[12] I. Cuesta, C. Rodriguez, T. Garcia, J. Alegre, Effect of confinement level on mechanical behaviour using the small punch test, Engineering Failure Analysis 58 (2015) 206-211. doi:http: //dx.doi.org/10.1016/j.engfailanal.2015.09.008.

[13] M. Moreno, G. Bertolino, A. Yawny, The significance of specimen displacement definition on the mechanical properties derived from Small Punch Test, Materials and Design 95 (2016) 623-631. doi:http://dx.doi.org/10.1016/j.matdes. 2016.01.148.

[14] K. Turba, R. Hurst, P. Hahner, Anisotropic mechanical properties of the MA956 ODS steel characterized by the small punch testing technique, J. Nucl. Mater. 428 (1-3) (2012) 76-81.

[15] C. Rodriguez, E. Cardenas, F. Belzunce, C. Betegon, Fracture characterization of steels by means of the small punch test, Experimental Mechanics 53 (2013) 385-392.

[16] S. Haroush, E. Priel, D. Moreno, A. Busiba, I. Silverman, A. Turgeman, R. Shneck, Y. Gelbstein, Evaluation of the mechanical properties of SS-316L thin foils by small punch testing and finite element analysis, Materials & Design 83 (2015) 75 -84. doi:http://dx.doi.org/10.1016/j.matdes.2015. 05.049.

[17] J.-M. Lapetite, M. Bruchhausen, Small punch tensile/fracture test data for Gr. 91 material at -100 °C and a displacement rate of 0.005 mm/s, JRC Petten (2015). doi:http://dx.doi. org/10.5290/1900105.

[18] E. Fleury, J. Ha, Small punch tests to estimate the mechanical properties of steels for steam power plant: I. Mechanical strength, International Journal of Pressure Vessels and Piping 75 (1998) 699-706.

[19] M. A. Contreras, C. Rodriguez, F. Belzunce, C. Betegon, Use of the small punch test to determine the ductile-to-brittle transition temperature of structural steels, Fatigue & Fracture of Engineering Materials & Structures 31 (2008) 727-737. doi:http: //dx.doi.org/10.1111/j.1460-2695.2008.01259.x.

[20] M. Saucedo-Munoz, T. Matsushita, T. Hashida, T. Shoji, H. Takahashi, Development of a multiple linear regression model to estimate the Ductile-Brittle Transition Temperature of ferritic low-alloy steels based on the relationship between Small Punch and Charpy V-notch tests, Journal of Testing and Evaluation 25 (2000) 352-358.

[21] I. Penuelas, I. Cuesta, C. Betegon, C. Rodriguez, F. Belzunce, Inverse determination of the elastoplastic and damage parameters on small punch tests, Fatigue and Fracture of Engineering Materials and Structures 32 (11) (2009) 872-885. doi:http: //dx.doi.org/10.1111/j.1460-2695.2009.01387.x.

[22] M. Bruchhausen, S. Holmstrom, I. Simonovski, T. Austin, S. Ripplinger, J.-M. Lapetite, F. de Haan, Recent developments in small punch testing: tensile and fracture testing, in: Anales de Mecainica de la Fractura, Vol. 33 of Encuentro del Grupo Espanol de Fractura, 2016, pp. 17-29.

URL http://www.gef2016.es/docs/

anales-de-mecanica-de-la-fractura-33.pdf

[23] K. Matocha, Small-punch testing for tensile and fracture behavior: Experiences and way forward, ASTM Special Technical Publication STP 1576 (2015) 145-159. doi:http://dx.doi. org/10.1520/STP157620140005.

[24] T. Lebarbei, S. Marier, P. Agostini, C. Fazio, S. Gavrilov, Presentation of FP7 MATTER project: General overview, in: Proceedings of the ASME 2011 Pressure Vessels & Piping Division Conference, Baltimore, Maryland, USA, 2011.

[25] M. Bruchhausen, MatDB catalog of small punch tensile/fracture data from tests performed between room temperature and 650 °C, version 1.0, European Commission JRC, [Catalog] (2016). doi:http://dx.doi.org/10.5290/9.

[26] M. Bruchhausen, MatDB catalog of uniaxial tensile test data for P91 material, version 1.0, European Commission JRC Institute for Energy and Transport, [Catalog] (2016). doi:http://dx. doi.org/10.5290/3.

[27] J. Chakrabarty, A theory of stretch forming over hemispherical

punch heads, Int. J. mech. Sci. 12 (1969) 315-325.

[28] M. Bruchhausen, S. Holmstrom, J.-M. Lapetite, S. Ripplinger, On the determination of the ductile to brittle transition temperature from small punch tests on Grade 91 ferritic-martensitic steel, (unpublished results, submitted to International Journal of Pressure Vessels and Piping).

[29] J.-M. Lapetite, M. Bruchhausen, Small punch tensile/fracture test data for Gr. 91 material at -196 °C and a displacement rate of 0.5 mm/s, JRC Petten, http://dx.doi.org/10.5290/ 1900103, v1.1, [data set] (2015).

[30] K. Turba, R. Hurst, P. Hahner, Evaluation of the ductile-brittle transition temperature in the NESC-I material using small punch testing, International Journal of Pressure Vessels and Piping 111-112 (2013) 155-161. doi:http://dx.doi.org/10. 1016/j.ijpvp .2013.07.001.

[31] M. Saucedo-Munoz, T. Hashida, T. Shoji, V. Lopez-Hirata, Correlation between Small Punch and CVN impact tests for evaluation of cryogenic fracture characteristics of isothermally-aged nitrogen-containing austenitic stainless steels, Materials Research 15 (2012) 218-223.

[32] J. Foulds, R. Viswanathan, Determination of the toughness of inservice steam turbine disks using small punch testing, Journal of Materials Engineering and Performance 10 (2000) 614-619.

[33] M. Bruchhausen, J.-M. Lapetite, S. Ripplinger, T. Austin, Small punch tensile/fracture test data and 3D specimen surface data on Grade 91 ferritic/martensitic steel from cryogenic to room temperature, Data in Brief 9 (2016) 245-251. doi:http:// dx.doi.org/10.1016/j.dib.2016.08.061.

[34] M. Bruchhausen, MatDB catalog of small punch tensile/fracture test data for P91 material, version 1.0, European Commission JRC Institute for Energy and Transport, [Catalog] (2016). doi: http://dx.doi.org/10.5290/2.

[35] J. Adams, R. C. Hurst, J. B. Borradaile, M. R. Bache, Influence of specimen geometry and strain rate on ductile-brittle transition characterisation of reactor pressure vessel steels using the small punch technique, in: American Society of Mechanical Engineers, Pressure Vessels and Piping Division (Publication) PVP, Vol. 6B-2015, 2015.

[36] K. Turba, B. Gulcimen, Y. Li, D. Blagoeva, P. Hahner, R. Hurst, Introduction of a new notched specimen geometry to determine fracture properties by small punch testing, Engineering Fracture Mechanics 78 (2011) 2826-2833.

[37] K. Matocha, The use of small punch tests for determination of fracture behaviour of ferritic steels, Procedia Engineering 86 (2014) 885-891. doi:http://dx.doi.org/10.1016/j. proeng.2014.11.110.

[38] B. Gulcimen, A. Durmu§, S. Ulku, R. Hurst, K. Turba, P. Hahner, Mechanical characterisation of a P91 weldment by means of small punch fracture testing, International Journal of Pressure Vessels and Piping 105-106 (2013) 28-35.

[39] M. Bruchhausen, MatDB catalog of small punch tensile/fracture test data with two punch diameters, version 1.0, European Commission JRC Institute for Energy and Transport, [Catalog] (2016). doi:http://dx.doi.org/10.5290/4.

[40] W. Lei, X. Yan, M. Yao, Numerical analysis of strain rate field below notch root of Charpy V-notch test specimen under impact loading condition, Engineering Fracture Mechanics 46 (4) (1993) 583-593.

[41] R. Lancaster, W. Harrison, G. Norton, An analysis of small punch creep behaviour in the y titanium aluminide Ti-45Al-2Mn-2Nb, Materials Science and Engineering A 626 (2015) 263-274. doi:http://dx.doi.org/10.1016/j. msea.2014.12.045.

[42] I. Simonovski, S. Holmstrom, M. Bruchhausen, Small punch tensile testing of cladding tube specimens: Finite element anal-

ysis and experiment (unpublished results, submitted for publication to International Journal of Mechanical Sciences).

[43] S. Holmstrom, A. Laukkanen, K. Calonius, Finding critical damage locations by i-filtering in finite-element modelling of a girth weld, Materials Science and Engineering A 510511 (C) (2009) 224-228. doi:http://dx.doi.org/10. 1016/j.msea.2008.04.107.

[44] J. Rice, D. Tracey, On the ductile enlargement of voids in triaxial stress fields, Journal of the Mechanics and Physics of Solids 17 (3) (1969) 201-217. doi:http://dx .doi.org/10.1016/ 0022-5096(69)90033-7.

[45] The Society of Materials Science, Japan (Ed.), Standard for Small Punch Creep Test - Estimation of Residual Life for High Temperature Component, Vol. ISBN978-4-901381-38-3, 2012.

[46] ASTM WK47431 New practice for Small Punch test method for metallic materials ((accessed on 10/06/16)).

URL http://www.astm.org/DATABASE.CART/

WORKITEMS/WK47431.htm

[47] ECISS/TC 101 (WI=EC101162) Small Punch test method for metallic materials ((accessed on 10/06/16)).

URL https://standards.cen.eu/dyn/www/

f?p=204:22:0::::FSP_ORG_ID:733642&cs=

14BA103751C289759D79531BE065B6A8D

[48] T. Austin, Realizing the unexplored potential of materials data, Materials Today 19 (2) (2016) 60-61. doi:http://dx.doi. org/10.1016/j.mattod.2015.12.002.

[49] CEN workshop on standards compliant formats for fatigue data - FATEDA ((accessed on 10/06/16)). URL http://www.cen.eu/work/areas/ict/ebusine; pages/ws-fateda.aspx

test ss/

Highlights

• Overview over recent developments in small punch tensile/fracture testing.

• Determination of yield, tensile strength and DBTT by small punch testing.

• Impact of notched specimens, displacement rate and punch diameter on DBTT.

• Review of current small punch standards and standardization activities.