Master Thesis
Development and Evaluation of a Small Punch
Testing Device
Jan Benjamin Ottosson
Linköping 2010
LIU-IEI-TEK-A–10/00870–SE
Division of Engineering Materials
Department of Management and Engineering (IEI) Linköpings universitet, SE-581 83 Linköping, Sweden
Development and Evaluation of a Small Punch
Testing Device
Master’s thesis project conducted at
Division of Engineering Materials
Linköpings universitet
by
Jan Benjamin Ottosson LIU-IEI-TEK-A–10/00870–SE
Supervisor and Examiner: Håkan Brodin
Avdelning, Institution
Division, Department
Department of Management and Engineering Division of Engineering Materials
Linköpings universitet S-581 83 Linköping, Sweden Datum Date 2010-006-10 Språk Language Svenska/Swedish Engelska/English Rapporttyp Report category Licentiatavhandling Examensarbete C-uppsats D-uppsats Övrig rapport
URL för elektronisk version
ISBN
—
ISRN
LIU-IEI-TEK-A–10/00870–SE
Serietitel och serienummer
Title of series, numbering
ISSN
—
Titel
Title Development and Evaluation of a Small Punch Testing Device
Författare
Author
Jan Benjamin Ottosson
Sammanfattning
Abstract
In the turbine industry today, thermal barrier coatings are a commonly used, these are 0.1-2mm thick. So to be able to do some type of mechanical testing to receive material data so one can build an opinion regarding the health of the material. One needs a procedure that can work with small specimens and achieve clear results that can be transformed and compared with known data and known procedures. One of those methods is Small Punch Testing.
This thesis describes one way to develop and test a functioning prototype of a Small Punch Testing device. The thesis includes; the reason it was developed in the beginning and how it has been developed throughout the decades, also in which areas the main research is made. It also shortly describes a working procedure in Ansys to get a Finite Element Method [FEM] model working.
This method showed itself as useful, when just a small sample is at hand. The trials in this thesis also show that repetitive test can be done with good results which can be compared with real and FEM analysis data such as σuts.
Inom turbin industrin idag så är keramiska värme barriärer vanligt förekom-mande dessa är normalt 0,1-2mm tjocka. För att kunna utföra mekanisk provning som grund för att bilda en åsikt om materialets kondition. Så behöver man en metod som kan åstadkomma tydliga data med små provbitar, Small Punch Test-ing är en av dem.
Den här rapporten beskriver hur man kan gå tillväga för att få en fungerande prototyp. Den tar upp metodens ursprung och hur den har utvecklats under år tiondena, också mot vad den nuvarande forskningen riktar sig. Den beskriver även kort hur man ställer upp en finita element metod [FEM] modell i Ansys.
Metoden visade sig användbar när man bara har en liten provbit att tillgå. Försöken visade att repetitiva tester kan göras med bra resultat som går att jäm-föra med verkliga och FEM analys data.
Nyckelord
Abstract
In the turbine industry today, thermal barrier coatings are a commonly used, these are 0.1-2mm thick. So to be able to do some type of mechanical testing to receive material data so one can build an opinion regarding the health of the material. One needs a procedure that can work with small specimens and achieve clear results that can be transformed and compared with known data and known procedures. One of those methods is Small Punch Testing.
This thesis describes one way to develop and test a functioning prototype of a Small Punch Testing device. The thesis includes; the reason it was developed in the beginning and how it has been developed throughout the decades, also in which areas the main research is made. It also shortly describes a working procedure in Ansys to get a Finite Element Method [FEM] model working.
This method showed itself as useful, when just a small sample is at hand. The trials in this thesis also show that repetitive test can be done with good results which can be compared with real and FEM analysis data such as σuts.
Inom turbin industrin idag så är keramiska värme barriärer vanligt förekom-mande dessa är normalt 0,1-2mm tjocka. För att kunna utföra mekanisk provning som grund för att bilda en åsikt om materialets kondition. Så behöver man en metod som kan åstadkomma tydliga data med små provbitar, Small Punch Testing är en av dem.
Den här rapporten beskriver hur man kan gå tillväga för att få en fungerande prototyp. Den tar upp metodens ursprung och hur den har utvecklats under år tiondena, också mot vad den nuvarande forskningen riktar sig. Den beskriver även kort hur man ställer upp en finita element metod [FEM] modell i Ansys.
Metoden visade sig användbar när man bara har en liten provbit att tillgå. Försöken visade att repetitiva tester kan göras med bra resultat som går att jäm-föra med verkliga och FEM analys data.
Acknowledgements
This master thesis is made as a final project in the programme for a Master of Sci-ence degree in Mechanical Engineering, at the Division of Engineering Materials. I would like to thank my supervisor Håkan Brodin and all the people at the division for their help and I would also like to thank the workshop staff.
Jan Benjamin Ottosson
Contents
1 Introduction 3
1.1 Aim and purpose . . . 3
1.2 History . . . 3 2 Method 5 2.1 Method . . . 5 2.1.1 Analytical method . . . 6 2.2 Specimen holder . . . 6 3 Results 9 4 Discussion 11 4.1 Clamp Force . . . 16
4.2 Performing the test . . . 16
4.3 Possible error sources . . . 16
4.3.1 Ansys . . . 17
5 Conclusion 19 A SPT drawings 21 B Ansys 25 B.1 What to think about when modelling a 3-D modell in Ansys . . . . 25
B.1.1 Last minute results . . . 27
Development and Evaluation of a Small Punch
Testing Device
Jan Benjamin Ottosson
2 Contents
Nomenclature and Equations
d = disc diameter [mm ] t = disc thickness [mm ] E = E modulus [GPa ] σ = tensile strength [MPa ] = strain[% ]
σuts = ultimate tensile strength [MPa ]
dF = displacement at failure [mm ]
Lu = ultimate load [N ]
Cl = die clearance D = ball diameter [mm ] A = lower die diameter [mm ] A = 1.5
D = 1
Equations 1, 2 and has been examined by S.D. Norris and J.D. Parker[1]
σuts= Lu t (2.32D − 0.9Cl + 0.56) (1) σuts= Lu t ∗ (0.14D − 0.82Cl + 2.17dF+ 0.6) (2) Cl = A − (D + 2t)
Chapter 1
Introduction
The ability of performing material investigations on small specimens is of great importance, especially when it comes to parts that are or has been in service. Small punch testing [SPT] has been developed for just that case e.g. to investigate a pipe in a nuclear reactor that is still in service, this type of testing can be seen as a non destructive test due to the small specimen size.
1.1
Aim and purpose
The aim of this thesis work is to design and evaluate a SPT device for room temperature testing and to see to which theory it corresponds, i.e. uniaxial tensile test [UTT] or other known methods. It was decided to use a device for room temperature testing to start with because it has not previously been done here at Linköping University, it can be seen as a pre-study to a high temperature device. The materials that will be examinined are Al 2524-t3, iron SS1312 and In792, these materials span from ductile to the not that ductile In792.
1.2
History
This type of technique has been around since the early eighties [2][3], foremost to study creep in high temperature alloys used in nuclear plants. The first two coun-tries where it started are the USA and Japan. There are two ways to look at SPT time dependant, for example creep examination at different temperatures than room temperature and time independent, examination made at room temperature to examine e.g. σuts
Nowadays it is mainly used in two different areas, high temperature alloys and ultra high molecular weight polyethylene [UHMWPE] in biomaterials [4]. It has it grounds in that one wants to be able to extract information from small volumes of material. It can be used for shear force examinations, but in that case one uses a flat punch instead of the ball that is used to examine σuts. In steels one also uses
SPT to find out ductile to brittle transition temperature [DBTT] [5]. They look at 3
4 Introduction
the difference between Charpy test [CVN] and SPT the conclusion that they did find was that the SPT results in a much lower transition temperature, the reasons why is closer discussed in the paper. The technique is applicable where there is a necessity to examine a product far too small to be able to produce normal sized test specimen. The thermal barrier coating [TBC] of turbine shovels is a good example, due to the small thickness of the TBC, 0.1-0.25mm. A common type of SPT uses transmission electron microscope [TEM] sample sized specimens i.e. 3mm in diameter and between 0.1 and 0.5mm in thickness [6]. This is the SPT commonly known as small punch test, which shall examined closer. One could believe that the disc thickness is not of importance, but in [7] they have looked closer at the influence of disc size and one can see that the deflection stays the same but the thicker the specimen the more force is needed, which one would hope to see. In [8] it is mentioned that a aspect ratio, disc diameter/thickness should be less than 60 otherwise the specimens showed signs of local deformation and are not acceptable for use, with a aspect ratio over 118 local wrinkling and plastic instability occurs which is useless when a repeatable result is needed. They also mention a thickness tolerance of +-0.0013mm but +-0.0051mm is a nominal acceptable value. Otherwise there is one standardized type with a disc diameter of 8mm and a thickness of 0.5mm used for metals [9][10] and for UHMWPE a disc diameter of 6.4mm and a thickness of 0.5mm[11]. The small sample size makes it compatible with other non destructive tests. It also makes it possible to actually take a sample to examine from a product in use, which is a big advantage. As far as it comes to the design of the test equipment that was seen as the best to go for, has it grounds in two papers [12][13]. But the actual design is made from scratch with the inner dimensions corresponding to the others.
Chapter 2
Method
2.1
Method
At first a literature study was done, to see what has been done so far, and how it has been made i.e. history. Are there any standards applicable? Then followed the actual CAD process to bring forth a working prototype, this took longer than expected because that it is my first prototype that I have made and I had very little knowledge in the CAD programme Pro/Engineer 4 before I started. The whole apparatus will be fitted in a 858 Mini Bionix II system with a 10kN loading-cellR which will be more than sufficient to provide accurate test data. To have a more exact measurement of the deflection an extensometer will be added under the die to measure directly the specimens deflection without any regards to ball compression etc. In two of the materials that shall be examined, iron SS 1312 and Al 2524-t3, there will also be a UTT done to see how the two different methods correspond to each other. The TEM sized samples will be produced through different methods, the Al 4010, Al 2524-t3 and SS 1312 will be punched from UTT specimens, In792 discs will be produced through cutting a disc and there after follows grinding to the correct thickness and from those plates the TEM sized specimen will be cut with electronic discharge machining [EDM]. All the specimens are grinded with 500 SiC paper and In792 finished off with 1200 SiC paper. When the discs had been produced the thickness of each was established with a digital dial indicator. The thickness was decided with three significant numbers. Due to the design of this device a thickness of maximum 0.25mm is a must. Larger thickness will lead to something equivalent to deep drawing of aluminium cans, highly undesirable.
Performing the test
The actual test was performed with a speed of 0.15mm/min which results in in-formation about force and deflection during the test. The clamping force for Al 2524-t3 ended up at 750-850N if more was applied the plate was plastically de-formed. For SS 1312 and In792 a higher clamp force was applied 1150-1350N.
6 Method
2.1.1
Analytical method
The analytical methods used for comparison is the computer software Ansys Work-bench v.12 and equations 1 and 2 derived by S.D. Norris and J.D. Parker for more informaiton about the equations see [1].
Ansys
A model was created in Ansys Workbench Static Structural which is a modern FEM software with graphical interface, to have more numerical data to compare with. When it came to achieve a working model in Ansys, there were a lot of problems getting it right due to all the constraints working in the model. All these questions had to be answered before implementing them: friction between different types of materials acting together? what type of load should be applied? 2-D or 3-D? Which type of mesh gives most relying result in the model? The last question could actually be answered after the model was working by trial and error. Some types of mesh gave some really disturbing results for example, deformation that cannot occur. Due to the complexity of the model friction coefficients had to be decided for each contact surface Cambridge Engineering Selector[14] was used in combination with [15] to establish them. Two types of elements tetragonal and quad was tried, also different mesh concentrations, from 4500 up to 6500. The difference in deflection is 0.6% which for my application is negligible and in the stresses measured according to von-Mises the difference was 2% which also is negligible. So the simulations are run with the lower mesh with quad mesh, to lower the time needed to do the calculations. The model used in this case is in 3-D to make it simpler to understand. The mesh used is in Ansys referred to as SOLD186, quad with 20 nodes. In Appendix B deeper information about the procedures in Ansys can be found.
2.2
Specimen holder
The first thing done was a literature study to find information, if there was a standard, if not which dimensions where most commonly used for the TEM sized specimens. Drawings were not found but applicable dimensions, but then there was the shape, outer size and function to determine. At this stage consultaions was made, with different persons working at the university and my supervisor and discussed different prototypes and concluded that according to this principle that the current prototype is built by is the most probable to succeed. Another obstacle encountered was how narrow should the different dimensions be, especially the specimen clamper with its small diameter hole and the area under the sample with a radius of 0.2mm to prevent shearing to occur. The material for the device was also to be determined and after many different propositions Uddeholm Vanadis 4R Extra was chosen for the critical parts(A.1., A.3.) due to its good properties, high wear resistance and high ductility which leads to high resistance against adhesive wearing and chipping. Through hardening, a hardness up to 66 HRC is achievable, which is in our application desirable. For the non critical parts steel will be used.
2.2 Specimen holder 7
Besides that the force will be measured as well as the deflection. To be able to do this a holder for the SPT device had to be made. Here some problem due to the physical limitations of the 858 Mini Bionix II was encountered. To holdR the specimens under clamp force. It was designed so that disc springs can be used to apply a different clamp force if it becomes necessary to change between different materials, which is most likely to be done between aluminium and In792. To be able to screw it together some type of grip had to be implemented, milling machines have a similar shape as the prototype has. So specifications for how C-Spanners work was included into the design. As punch Cr-steel balls with a hardness of 60/66HRC will be used and as ball pusher a rod of 99.7% Al2O3 will be used all this to minimize malfunction and interference during testing.
Figure 2.1. Device setup
Figure 2.1 shows the device setup in the machine, showing the black load cell and with its puncher.
Chapter 3
Results
With the designed small punch testing equipment repetitive tests have been made and they show consistency which is a requirement to be able to draw conclusions from the received data. When the data for the different materials was obtained, an average was calculated too and that value was compared withdata and the result from Ansys.
An example is shown in Figure 3.1 which represents a typical curve.
Figure 3.1. A typicall Load/Displacement curve
There is a concave part of the Load Displacement curve that can be derived from the fact that the plate receives a larger contact area with the ball and at the same time undergoes work hardening.
This shape of curve makes it more or less impossible to be able, by any means, to receive a yield stress so the focus has been set on σuts. To calculate σuts for
Al2524-t3 and SS1312 Equation 1 has been used and Equation 2 for In792. 9
10 Results
Table 3.1. Test data
Material Calulated σuts Ansys σuts Handbook σuts UTT σuts
2524-t3 312 528 405-450 515
SS 1312 320 666 360-460 320
In 792 1320 1274 845
These are the data recived from tests with the other data available on the examined materials. The handbook gives a span for σuts, the UTT for my specific
materials are also represented.
Figure 3.2. Curves for the three examined materials
Chapter 4
Discussion
The calculated σuts for In792 is really close to the value from Ansys.
The Al 2524-t3 shows good correlation between the Ansys result and the UTT, but a significant difference to the calculatded value.
For SS 1312 the calculated σuts and the UTT it is a perfect match but between
Ansys and UTT the value received from Ansys is more than twice as high as for the other values.
As stated in [7] equation 1 and 2 are a good point of origin to find a more specific equation for the material that one is conducting research on at the moment, the need to do that can easily be deduced from the analyzed data. If there had been more consistency between the materials it would have been better. Of three different pairing options all three were hit but this probably lies in the fact that three very different types of materials was chosen, this to induce variating results which was received. The method is not perfect but when properly used a really good option.
12 Discussion
Figure 4.1. Whole curve with rupture
Figure 4.2. UTT 2524-t3
In Figure 4.1 the plastic deformation and the rupture is clearly seen but the where exactly does the transition from elastic deformation to plastic deformation occur?. Figure 4.2 shows a UTT 2524-t3 test for comparison with a familiar data type. With the three known steps, elastic deformation, plastic deformation and rupture of the material. Nevertheless, the curve in Figure 4.1 is really interesting, one should have in mind that this method affects the material in biaxial directions, not uniaxial which is the most used procedure. As seen in 3.2, this type of curve was seen in all the materials but the knee became less noticeable with increasing thickness. When the UTT values in Table 3.1 and the graphs in Figure 3.2 are
13
compared, they correlate good between each other. SS1312 with the lowest value also has the lowest value in the graph which further supports the theory that this method gives repeatable and reliable results. The graphs curvature supports pre-vious work done in the field which also supports the statement of a working device. In Figure 4.3 and Figure 4.4 fractures are shown for the different materials. As expected the crack propagation seem to have started at the top where the stresses according to Ansys are highest Figure 4.5.
14 Discussion
Figure 4.4. Fracture Al 2524-t3
15
As seen in Figure 4.6 the stresses are spread out throughout the top of the specimen which in reallity transforms into the microcracks seen in Figure 4.7.
Figure 4.6. Typicall stress distribution under force from beneath
16 Discussion
Table 4.1. Serie 2524-t3
σuts 309 322 294 313 318 301 311
Average 310
In Table 4.1 ultimate strenght is shown for AL 2524-t3, it is to be seen that the average value has a deviation of +4% and -5% which can be seen as good consistency regarding all unknown parameters that was declared at the beginning.
4.1
Clamp Force
The clamping force in this case was there to prevent slipping of the disc, so it can be seen as clamped disc. It is also corresponds to how the test are commonly performed. The actuall affect of the clamping brings some questions. As stated previously large plastic deformation was observed when to high force was applied, was the estamination of how much the force must be reduced correct? or did it still affect the material, but not at a macro level but at micro level and there for not visible. They clamping force should there for be weak enough to leave no irreversible deformation but strong enough to prevent slipping. Equation 1 and 2 do not take clamp force into regard, thus they are empiricall so they build on experiments with know boundary conditions, so with similar conditions it should work. The fact that the method is a plate with clamped edge can bring in the question if [16] can be used, that is plate bending theory. The handbook shows two different scenarios, plate with clamped edges and freely supported plate, and in SPT the scenario is a mixture of the two based on the fact that a radius of 0.2mm has been implemented to prevent shearing that can occur. In the beginning it was regarded as one solution but later on as more informatio came forth that idea was discarde.
4.2
Performing the test
The velocity of 0.15mm/min was set at the beginning of the trails but higher velocities where also tried but they caused immediate failure of the ceramic staff, so the speed was kept at 0.15mm/min. One solution to prevent it from happening could have been to use a staff consisting of a harder ceramic.
4.3
Possible error sources
As possible error sources, some type of micro work hardening due to grinding. For the Al 2524-t3 and SS 1312, surface roughness may play a part, only being grinded with 500 SiC paper not 1200 as with In792. Was the friction between ball and disc always the same. Was the clamping force exactly the same?
4.3 Possible error sources 17
4.3.1
Ansys
In Ansys some obvious error sources are the defined parameters as friction coef-ficients, clamp force, the fact that it is not a transient analasis(it does not use constant speed), mesh type and size. Most of these are taken into consideration but there is always room for errors.
Chapter 5
Conclusion
In this report small punch testing has been examined and discusses different types of conclusion one can draw from the recieved data. And the results are encouraging to that point that they are repetitive and shows good consistency.
The general equations used gives a rule of thumb about where the σuts will be.
For further analasis a recommendation to perform trials at 0.10, 0.15, 0.20 and 0.25mm thickness should be done with eight specimens of each thickness and remove the specimens with the highest and lowest value. This to minimize the affect of anomalies thar can occur during testing.
Appendix A
SPT drawings
22 SPT drawings Konstruerad av Artikel nr/Referens Ägare Utgåva Blad Titel/Benämning Skala Vyplacering jämnhet, Ra Generell yt-Ritningsnummer Pos nr Antal
Granskad av Godkänd av - datum Generell tolerans SS-ISO2768-1 Titel/Benämning, beteckning, material, dimension o.d.
IEI
Linköpings universitet Godkänd av-datum Ändr nr Ändringens art/Ändringsmeddelande M52x1.5 21.5-0.1 37±0.5 30±0.5 1.5+0.1 3.2-0.05 21+0.5 1-0.1 4±0.5 6.3 6.85±0.5 A-A SECTION R4 B-B SECTION 0.2x45° R0.2 1x45°23 Konstruerad av Artikel nr/Referens Ägare Utgåva Blad Titel/Benämning Skala Vyplacering jämnhet, Ra Generell yt-Ritningsnummer Pos nr Antal
Granskad av Godkänd av - datum Generell tolerans SS-ISO2768-1 Titel/Benämning, beteckning, material, dimension o.d.
IEI
Linköpings universitet Godkänd av-datum Ändr nr Ändringens art/Ändringsmeddelande 70.00±1 M52.00x1.5 35.00±1 21.00±0.5 8.00±0.5 20.00-0.5 1.00±0.5 4.00±0.5 10.85 6.30 15.00±0.5 B-B SECTION 1x45° 1x45° 1x45° 1x45° A-A SECTION R424 SPT drawings Konstruerad av Artikel nr/Referens Ägare Utgåva Blad Titel/Benämning Skala Vyplacering jämnhet, RaGenerell
yt-Ritningsnummer Pos nr Antal
Granskad av Godkänd av - datum Generell tolerans SS-ISO2768-1 Titel/Benämning, beteckning, material, dimension o.d.
IEI
Linköpings universitet Godkänd av-datum Ändr nr Ändringens art/Ändringsmeddelande 21+0.1 8±0.2 3.1+0.05 1.05±0.01 1+0.1 15±0.5 19±0.5 A-A SECTION C-C SECTION 1x45°Appendix B
Ansys
B.1
What to think about when modelling a 3-D
modell in Ansys
In this thesis Ansys Workbench v. 12 Static Structural was used. The problems do not only lie in the choosing of mesh, in my case that was the easy part. The real problem was to find what specific commands that has to be turned on when working with a 3-D non-linear model with large deformation. How to define the proper connections and constraints. Here follows a short list with the most critical settings in my modell.
In Commands file; CNCHECK, AUTO
In the outline tree Behaviour: Asymmetric
Formulation: Augmented Lagrange Normal Stiffness: Manual
Normal Stiffness Factor: between 0.01-0.04 Large Deflection: ON
Weak Springs: Program Controlled
In Solutions be careful for what you want, think twice when defining in Scope which bodies/surfaces one want to receive data from, most probably not the whole system.
26 Ansys
For the analasis made, some material parameters had to be stated in Ansys and a non-linear model was uses and then Ansys need the following data.
Density
Young’s Modulus Poisson’s Ration Yield Strenght Tangent Modulus
Figure B.1. Defined modell in Ansys
Figure B.1 shows how the applied forces and limitations was applied. The only thing changing between the test was the material of the disc in this model. This was done to have so few variables as possible.
The clamping force was set to 2000 N, but using the limitation of displacement, so the compacting of the disc was limited to maximum 0.015mm. The punch force of 500 N was set on the fact that most of the other tests done in this field has used a 500 N load cell to perform this type test.
B.1 What to think about when modelling a 3-D modell in Ansys 27
B.1.1
Last minute results
A last minute change was done, this change was to increase the step length so it corresponds to resluts received in the SPT tests for each material, that is the time it took for the specimen to reach maximum force.
For the different materials better data was the result, higher accuracy compared with handbook data was achieved for SS1312 and AL 2524-t3 the new data can be seen in the table below.
Table B.1. Ansys last minute data
Material Ansys σuts Handbook σuts
2524-t3 339 405-450
SS 1312 362 360-460
In 792 1274 845
But In792 received the same value, so is that because of that it is a more brittle material or has it some other ground? That is for future tests to decide.
Bibliography
[1] S. D. Norris and J. D. Parker. Deformation processes during disc bend loading. Materials Science and Technology, 12:163–170(8), February 1996.
[2] Jai-Man Baik, J. Kameda, and O. Buck. Small punch test evaluation of inter-granular embrittlement of an alloy steel. Scripta Metallurgica, 17(12):1443– 1447, 12 1983.
[3] G. E. Lucas. The development of small specimen mechanical test techniques. Journal of Nuclear Materials, 117:327–339, 7 1983.
[4] S. M. Kurtz, C. W. Jewett, J. S. Bergström, J. R. Foulds, and A. A. Edidin. Miniature specimen shear punch test for uhmwpe used in total joint replace-ments. Biomaterials, 23(9):1907–1919, 5 2002.
[5] M. A. Contreras, C. Rodríguez, F. J. Belzunce, and C. Betegón. Use of the small punch test to determine the ductile-to-brittle transition temperature of structural steels. Fatigue and Fracture of Engineering Materials and Struc-tures, 31(9):727–737, 2008. Cited By (since 1996): 1.
[6] T. Misawa, T. Adachi, M. Saito, and Y. Hamaguchi. Small punch tests for evaluating ductile-brittle transition behavior of irradiated ferritic steels. Jour-nal of Nuclear Materials, 150(2):194–202, 10 1987.
[7] Zhao-Xi Wang, Hui-Ji Shi, Jian Lu, Pan Shi, and Xian-Feng Ma. Small punch testing for assessing the fracture properties of the reactor vessel steel with different thicknesses. Nuclear Engineering and Design, 238(12):3186–3193, 12 2008.
[8] T. H. Hyde, W. Sun, and J. A. Williams. Requirements for and use of minia-ture test specimens to provide mechanical and creep properties of materials: a review. International Materials Reviews, 52:213–255(43), July 2007. [9] D. T. Blagoeva and R. C. Hurst. Application of the cen (european
com-mittee for standardization) small punch creep testing code of practice to a representative repair welded p91 pipe. Materials Science and Engineering A, 510-511(C):219–223, 2009.
[10] Cwa 15627:2007 small punch test method for metallic materials. 29
30 Bibliography
[11] Standard test method for small punch testing of ultra-high molecular weight polyethylene used in surgical implants.
[12] E. N. Campitelli, P. Spätig, R. Bonadé, W. Hoffelner, and M. Victoria. Assess-ment of the constitutive properties from small ball punch test: ExperiAssess-ment and modeling. Journal of Nuclear Materials, 335(3):366–378, 2004. Cited By (since 1996): 19.
[13] Mats Eskner and Rolf Sandström. Measurement of the ductile-to-brittle tran-sition temperature in a nickel aluminide coating by a miniaturised disc bend-ing test technique. Surface and Coatbend-ings Technology, 165(1):71–80, 2/3 2003. [14] Granta Design. Cambridge engineering selector, 2009.
[15] Tribology.
[16] Ansel C. Ugural. Stresses in plates and shells, pages 120–121, Case 5,8. WCB McGraw-Hill, Internatilnal Editions, second edition edition, 1999.
