SCHREYER HONORS COLLEGE DEPARTMENT OF MATERIALS …
Post on 11-Apr-2022
5 Views
Preview:
Transcript
THE PENNSYLVANIA STATE UNIVERSITY
SCHREYER HONORS COLLEGE
DEPARTMENT OF MATERIALS SCIENCE AND ENGINEERING
DEVELOPMENT OF SMALL-SCALE MECHANICAL TESTING TECHNIQUES FOR
LIMITED VOLUME METALS
ANDREW HENRY JOHNSON
SPRING 2020
A thesis
submitted in partial fulfillment
of the requirements
for a baccalaureate degree
in Materials Science and Engineering
with honors in Materials Science and Engineering
Reviewed and approved* by the following:
Allison M. Beese
Associate Professor of Materials Science and Engineering and Mechanical Engineering
Thesis Supervisor
Robert Allen Kimel
Assistant Professor of Materials Science and Engineering
Associate Head for Undergraduate Studies in Materials Science and Engineering
Honors Adviser
* Electronic approvals are on file.
i
ABSTRACT
Uniaxial tensile testing is a ubiquitous method for determining material strength that has
been used since the 18th century. Current metallic tensile testing standard ASTM E8/E8M-16a is
used in US industry and academia to ensure test results can be accurately compared between labs
and companies. The procedure defined by ASTM E8 provides different sample geometries;
however, these may be too large to be used to study samples with limited volume, such as
metallic glasses and weld regions. To investigate the ability to measure consistent bulk properties
with different sized samples, several miniature tensile test geometries have been designed by
researchers. Although material properties are intrinsic, their measurement can be affected by the
geometry of samples. In this work, samples of AISI Stainless Steel 304 in five different
miniature geometries were tested in a miniature load frame to examine any changes in properties
as a function of sample geometry. Samples were deformed to fracture in quasi-static uniaxial
tension and strain was measured using digital image correlation. Testing revealed that, with the
scope of samples and settings used, miniature samples have higher 0.2 % offset yield, ultimate
tensile strength, and elongation at fracture than conventional samples. Elongation at fracture was
also shown to increase with the ratio of the square root of initial sample gauge region cross-
sectional area to length.
ii
TABLE OF CONTENTS
LIST OF FIGURES ................................................................................................. iii
LIST OF TABLES ................................................................................................... v
ACKNOWLEDGEMENTS...................................................................................... vi
Chapter 1 Introduction ............................................................................................. 1
Chapter 2 Background Literature Review ................................................................. 3
2.1 Introduction ...........................................................................................................3 2.2 History ...................................................................................................................3 2.3 Uniaxial Tension Testing .......................................................................................4
2.3.1 Calculation of Mechanical Properties ...........................................................9 2.4 Miniature Uniaxial Tensile Testing in Literature .....................................................10
2.4.1 Metallic Glasses ...........................................................................................11 2.4.2 Additive Manufacturing ...............................................................................12 2.4.3 Weld Characterization .................................................................................14 2.4.4 In-Service Component Evaluation ................................................................15 2.4.5 Nuclear Materials ........................................................................................16 2.4.6 Geometry-Property Relations .......................................................................18
2.5 Structure, Properties, and Processing of SS304 .......................................................19 2.6 Engineering Considerations ....................................................................................21
Chapter 3 Materials and Procedure ........................................................................... 24
3.1 Sample Composition and Processing ......................................................................24 3.2 Sample Geometries ................................................................................................25 3.3 Tensile Testing Procedure ......................................................................................29 3.4 DIC Strain Measurement ........................................................................................34
Chapter 4 Experimental Results and Discussion ....................................................... 36
4.1 Stress-Strain Curves and Mechanical Property Data................................................36 4.2 Possible Explanations for Deviation from Predicted Behavior .................................40
Chapter 5 Summary and Conclusions ....................................................................... 43
Chapter 6 Future Work ............................................................................................. 45
BIBLIOGRAPHY .................................................................................................... 46
iii
LIST OF FIGURES
Figure 1. A typical screw-driven electromechanical uniaxial tension machine. During testing, the moving crosshead rises to produce tensile force on the sample. Figure from ref. [3]......6
Figure 2. Diagram and dimensions of standardized rectangular tensile test geometries per ASM
E8/E8M – 16a. Sample thicknesses should match those of the sheet or plate from which they were taken. Figure adapted from ref. [2]. ......................................................................8
Figure 3. Diagram and dimensions of standardized pin-loaded tension test specimen geometry per
ASM E8/E8M – 16a. Sample thicknesses should match those of the sheet or plate from
which they were taken. Figure adapted from ref. [2]. ....................................................8
Figure 4. Typical stress-strain curve for a metallic tensile sample with labelled property values.
Figure adapted from ref. [3]. ........................................................................................10
Figure 5. Drawings of specimens tested by Karnati et al. Figure adapted from ref. [11]. .......13
Figure 6. Steel weld schematic, exhibiting diverse microstructures as a function of weld zone.
Figure adapted from ref. [14] .......................................................................................14
Figure 7. Geometry A-type specimens. All dimensions shown are in millimeters. ................25
Figure 8. Geometry B-type specimens. All dimensions shown are in millimeters. ................26
Figure 9. Geometry C-type specimens. All dimensions shown are in millimeters. ................26
Figure 10. Geometry D-type specimens. All dimensions shown are in millimeters. ..............27
Figure 11. Geometry E-type specimens. All dimensions shown are in millimeters. ...............27
Figure 12. Geometry F-type specimens. All dimensions shown are in millimeters. ...............28
Figure 13. Geometry G-type specimens. All dimensions shown are in millimeters. ..............28
Figure 14. Custom miniature load stage used for tensile tests. ..............................................30
Figure 15. Two-pin sample grip assembly, proceeding from (a) to (h). .................................31
Figure 16. Experimental setup, including lighting, scaffolding, camera, mini load stage, and
control box. ..................................................................................................................32
Figure 17. Example DIC strain field and virtual extensometer of C1 sample halfway through
testing. .........................................................................................................................35
Figure 18. Engineering stress vs. engineering strain curves for tested samples. .....................36
Figure 19. UTS and 0.2% YS vs. gauge region volume of samples. For a given sample type, both values appear to decrease with increasing gauge region volume. ...................................39
iv
Figure 20. UTS and EL vs. gauge region volume of samples. No trends are apparent. ..........39
Figure 21. EL at fracture vs √(A_0 )/L_0 for tested samples. The two variables are positively
correlated as expected from literature. ..........................................................................40
Figure 22. Sample E2 post-fracture. Fracture occurred off-center and the remaining pieces appear
misaligned. ..................................................................................................................42
Figure 23. Close-up of sample E2 prior to loading. Roughness can be seen in the lower left-hand side of the gauge region. ..............................................................................................42
v
LIST OF TABLES
Table 1. Extensometer accuracy classifications according to ASTM E83-16. Table from ref. [6]. 7
Table 2. Composition of tested SS304 (wt%). ......................................................................24
Table 3. Measured gauge region widths and thicknesses of tensile samples prior to testing...29
Table 4. Sample reduced parallel and gauge region lengths and load rates for analyzed samples. 33
Table 5. Yield stress, UTS, elongation (EL), and elastic modulus values for samples and from the
manufacturer/literature. ................................................................................................37
vi
ACKNOWLEDGEMENTS
My thesis would not have been completed without the invaluable assistance of several
individuals. First, I would like to thank Dr. Allison Beese for her patience and guidance during
the thesis writing process. I would also like to thank Lourdes Bobbio for providing advice and
training during testing and feedback during writing. Additionally, I would like to thank Alex
Wilson-Heid, Dr. Shipin Qin, and Alex Caputo for assisting me with ordering materials,
machining samples, and analyzing data.
Regarding funding, I would like to thank Dr. Beese and the Penn State Department of
Materials Science and Engineering for their assistance in paying for materials, machining, and
facilities. I would also like to thank the department for providing me with a curriculum that
taught me many facts and equations, but most importantly how to think about materials.
Finally, I would like to thank my parents David and Jennifer Johnson for teaching me to
be curious about the world around me and for supporting me through all my academic endeavors.
Soli Deo Gloria
1
Chapter 1
Introduction
In the quest to create new and improved materials for modern applications, material
characterization is an important step. Areas of interest include developing novel materials, such
as metallic glasses, and determining processing-structure-property relations resulting from new
techniques such as additive manufacturing. Mechanical characterization is especially important
to evaluate a material’s potential for use in stressed components, from high pressure turbine
blades to irradiated pressure vessels. Without accurate and efficient measurement of physical
property data, the commercial application of new materials and processes is delayed.
To enable more efficient mechanical characterization of a wide variety of metallic
materials, this thesis explores the miniature tensile testing technique. Tensile testing of sub-size
specimens requires small volumes of material, less bulky test machines, and smaller force and
power than conventional testing. Considering these benefits, its use in characterizing expensive
or small-volume materials is proposed. Existing literature explores various miniature testing
techniques applied to metallic glasses, additive manufacturing, weld characterization, in-service
components, and nuclear materials, but recommends further work to explore the effect of test
specimen geometries on measured properties.
This thesis examines the effect of sample geometry on properties measured through
miniature uniaxial tension testing. Seven sample geometries with different gauge lengths and
widths were cut from a sheet 304 stainless steel with uniform thickness. These samples were
tested on a custom miniature load frame until fracture. Strain was measured using the non-
2
contact Digital Image Correlation (DIC) method. Force and strain data were combined and
examined to determine the elastic modulus, yield strength, ultimate tensile strength (UTS),
percent elongation, and fracture stress. Data are compared between samples and to the
mechanical properties given by the sheet manufacturer. Trends in data are then indicated and
explained. Finally, the limitations of the work are summarized, and future directions of study are
recommended.
3
Chapter 2
Background Literature Review
2.1 Introduction
This chapter serves to summarize background information for miniaturized tensile testing
to contextualize the reported work. First, the historical development of, contemporary standards
for, and theory behind uniaxial tension testing are introduced. Next, conventional testing
techniques and previous miniature tensile testing studies in a variety of fields are summarized.
This survey covers sample geometries, test stands, and extensometer techniques. Key
relationships between sample gauge region geometry and measured mechanical properties
reported in the literature are explored, following the survey of experiments by field. Next, the
structure and properties of the material tested in this thesis, 304 stainless steel (SS304), are
summarized to provide a point of comparison for experimentally acquired data. Finally, the
promising applications and industrial, environmental, and economic impacts of miniaturized
tensile testing will be considered.
2.2 History
From the advent of human technological development, engineering materials have been
chosen for applications based on their mechanical properties. The first recorded, explicit,
quantitative material standard for an application requiring mechanical strength is a 4th century
4
B.C.E. Greek artifact known as the “Stele of Eleusis.”1 This inscription detailed a mandated
material composition of 11:1 Cu to Sn for bronze spigots used the construction of structural
columns and indicates an ancient understanding of composition-property relations.1 Elastic
strength was not specifically studied mathematically until the 16th and 17th centuries when
Galileo and Hooke investigated elasticity in the contexts of the bending of structural materials
and the stretching of springs, respectively.1 Young’s elastic modulus was introduced in the 19th
century and a variety of tensile test stands were invented in the 18th and 19th centuries.1
Extensometer devices were also developed in the 19th century.1 In 1871, Germany established its
“Bundesanstalt für Materialforschung und prüfung” (translated “Federal Institute for Materials
Research and Testing”) in Berlin, signaling the first institutional push towards standardization in
tensile testing and many other research areas.1 The first American tensile testing standard,
ASTM E8-24T, was issued in 1924.1
2.3 Uniaxial Tension Testing
An updated version of the first American tensile testing standard (ASTM E8/E8M-16a) is
the current standard followed by U.S. industrial and academic mechanical testing facilities.
These “Standard Test Methods for Tension Testing of Metallic Materials” can be used at room
temperature to determine various mechanical properties of metallic samples.2
ASTM E8/E8M-16a begins by outlining the tensile testing apparatus. Such devices
(Figure 1) have two crossheads, aligned on a central axis, attached to sample ends, that are
driven electromechanically or hydraulically away from one another until fracture occurs.3 One or
both crossheads can be driven during testing. Tests can be conducted using a constant load rate, a
5
constant strain rate, or a constant crosshead speed.3 Testing machines are outfitted with strain-
gauge load cells or pressure transducers so force can be measured or controlled.3 To track or
control strain rate, test stands must be paired with extensometers.3 Crosshead speed is
proportional to the strain rate and can be controlled by gear speed, screw speed, or hydraulic
pressure depending on the type of test stand used.3 During testing, wedge, split socket, screw,
self-adjusting, or pin grips are recommended depending on sample geometry in order to prevent
slip.2 Testing machines must be verified and calibrated to ensure that their force application and
crosshead movement speeds are consistent with what are displayed or outputted.4,5 Force
verification/calibration can be accomplished by applying a known load to the testing machine
and comparing to the machine readout.4 Speed verification/calibration is conducted using a linear
scale and stopwatches to calculate actual speed and compare to the indicated machine speed.5
6
Figure 1. A typical screw-driven electromechanical uniaxial tension machine. During testing, the moving
crosshead rises to produce tensile force on the sample. Figure from ref. [3].
A second important component of uniaxial tensile testing is the extensometer used to
measure sample strain. Extensometers are classified into three categories, defined as Types 1, 2,
and 3 by ASTM E83-16.6 A Type 1 extensometer defines its own gauge length and records
extension to calculate strain.6 An of example Type 1 extensometers are clip-on extensometers
that attach to samples and measure extension using either an linear variable differential
transducer or a strain-gauge.3 Type 2 extensometers also sense extension, but have their gauge
length defined by sample features or geometry.6 Type 3 extensometers intrinsically sense strain
through a ratiometric principle.6 An example Type 3 extensometer would be a virtual
7
extensometer applied using the Digital Image Correlation (DIC) technique, a method which will
be explained in more detail in Section 3.3.2. Extensometers, like test stands, must be verified,
calibrated, and classified based on accuracy of measurement according to the standards shown in
Table 1.6
Table 1. Extensometer accuracy classifications according to ASTM E83-16. Table from ref. [6].
The final component required for tensile testing is the tensile specimen. According to the
standard, specimen geometries are separated into three major categories: plate-type, sheet-type,
and round.2 Additional specialized specimen geometries are detailed for materials with
uncommon mechanical properties, geometries constrained by application, and specialized
processing such as cast irons, metal tubing, and powder metallurgy products. The proposed novel
test geometries in this work most closely resemble the rectangular (plate-type and sheet-type)
(Figure 2) and pin loaded (Figure 3) tension test specimens. Plate-type specimens are
recommended for the testing of plates with a minimum thickness of 5 mm.2 Sheet-type and
subsize specimens are recommended for sheets with maximum thicknesses of 19 and 6 mm,
respectively.2
8
Figure 2. Diagram and dimensions of standardized rectangular tensile test geometries per ASM E8/E8M – 16a. Sample
thicknesses should match those of the sheet or plate from which they were taken. Figure adapted from ref. [2].
Figure 3. Diagram and dimensions of standardized pin-loaded tension test specimen geometry per ASM E8/E8M – 16a.
Sample thicknesses should match those of the sheet or plate from which they were taken. Figure adapted from ref. [2].
In addition to standards for test fixtures, extensometers, and tensile sample geometries,
ASTM E8/E8M-16a contains specifications for the testing procedure. First, the speed of testing
must be monitored and kept within bounds determined by the ASTM standard or by industry
standards. Test speed can be defined as specimen strain rate, specimen stress rate, crosshead
speed, test time, or free-running crosshead speed.2 Monitoring and reporting of speed is
9
important, as the testing rate can affect the measured mechanical properties of materials. For
most metals, the measured uniaxial tensile strength and yields stress increase with increased
strain rates, although the magnitude of this effect varies with material.3,7
2.3.1 Calculation of Mechanical Properties
Once testing is completed, material properties, including yield strength, uniform
elongation, and UTS, can be calculated from recorded force and displacement data. Properties
are usually determined through analysis of an engineering stress vs. engineering strain curve
derived from force and displacement data (Figure 4). Two methods are recommended to
calculate yield strength for continuously yielding materials: offset and extension under load
(EUL). In the offset method, a line is constructed with an origin at 0 stress and a strain offset
value, conventionally 0.2 %, with a slope equal to that of the linear region of the experimental
stress/strain curve.2 The stress value of the intersection of this offset line and the experimental
curve is taken as the yield strength. In the EUL method, a stress at a specified extension value,
usually 0.5 % for steels with nominal yield strengths of less than 550 MPa, is recorded as the
yield strength.2 The yield strength of higher strength steels should be calculated using a higher
extension value or the offset method.2 In general, if values are found using both the offset and
EUL methods, the offset value should be favored.2 Elongation at fracture and fracture stress are
the last recorded values of engineering stress and strain, respectively, before fracture occurs as
indicated by a sudden, discontinuous decrease in force data.2 If fracture occurs outside of the
gauge region or is located less than 25 % of the elongated gauge length from the region’s edge,
the elongation at fracture may not be representative of the material.2 Uniform strain and UTS are
10
the values of engineering stress and strain corresponding with the maximum of the stress-strain
curve.2 A final useful parameter, reduction of area, is the difference of the initial cross-sectional
area of the sample and that of reassembled fracture specimen.2 Reduction of area allows for the
calculation of Poisson’s ratio and other related elastic constants. This value is only valid for
fracture in the gauge region.2
Figure 4. Typical stress-strain curve for a metallic tensile sample with labelled property values. Figure adapted
from ref. [3].
2.4 Miniature Uniaxial Tensile Testing in Literature
Test methods like those outlined in the ASTM guidelines have proven to be useful in
industry as the standard for material certification and comparison. There are, however, certain
scenarios where the conventionally sized samples are prohibited by cost and volume of available
material. In such circumstances, novel, miniature tensile samples, test fixtures, and appropriate
extensometer techniques have been employed by researchers. Such research has been conducted
11
primarily on five groups of metals: metallic glasses, additively manufactured metals, welded
metals, in-service component samples, and irradiated metals from the nuclear industry. In this
section, rectangular samples and the corresponding measurement techniques are examined, as
they are most comparable to the reported work. Additionally, only samples greater than or equal
0.25 mm in thickness will be explored, as they are used to characterize the bulk properties of
sheets, plates, and larger components. Finally, this section will discuss known relationships
between sample geometry and reported property values.
2.4.1 Metallic Glasses
A first area of interest for miniature tensile testing is the characterization of metallic
glasses. These vitreous materials are created by rapidly quenching a melt composed of a variety
of metal and metalloid elements. Cooling rates on the order of 105–1012 °C/s are required to
avoid crystallization of the melt during quenching.8 To reach such cooling rates, fibers and sheets
are formed through rolling and melt spinning.8 Such processes create extremely small volumes of
material that cannot be machined into conventionally sized test samples. As metallic glasses
show promise in applications such as flexible magnetic shielding, power transformer core
laminations, and composite fibers, techniques for reliable testing are an important step in their
transition from research to industrial applications.8
One example study in this area was conducted by Wu et al. on a bulk metallic glass
composite.9 The studied material had a composition of Zr48Cu47.5Al4Nb0.5 and a microstructure
consisting of a continuous glassy phase surrounding crystalline precipitates. To test mechanical
properties, dog-bone samples with a gauge length of 10 mm, gauge width of 1 mm, and thickness
12
of 1 mm were extracted using wire electron discharge machining (EDM) from a cast ingot. Using
a constant strain rate of 1 x 10-4 s-1, researchers generated stress-strain data which they used to
show the effects of the precipitated phase on mechanical properties.
2.4.2 Additive Manufacturing
Another family of materials for which miniature tensile testing is advantageous is
additively manufactured metals. As such materials are being adopted into industry, extensive
work is underway to systematically understand relationships between processing parameters and
the resulting structures and properties. Such efforts often involve the survey of up to hundreds of
different samples in order to identify trends in material properties with respect to the powder and
processing variables. In order to save both time and material, the mechanical properties of such
samples are best studied with the use of miniature tensile specimens. The testing of functionally
graded additively manufactured materials, with compositions that vary throughout their
thickness, is an especially promising application of miniature tensile testing. Small sample sizes
allow researchers to test the properties of different composition regions as well as boundaries
between regions without the need to print regions separately.
Dongare et al. studied the mechanical properties of laser deposited Ti-6Al-4V in
comparison to wrought Ti-6Al-4V using novel miniature test specimen geometries.10 Specimens
had an overall length of 17.74 mm, thickness of 1 mm, gauge length of 3.3 mm, and width of 1
mm. In the grip regions of the specimens, 3 mm diameter holes were drilled so that specimens
could be held by loading pins in specially designed grips. Miniature testing produced yield
13
strength values matching those found using standard specimens and was proven to be reliable
and reproducible according to the authors.
Karnati et al. utilized miniature tensile specimens to characterize SS304L manufactured
through selective laser melting and through conventional hot rolling and annealing.11 The
investigators used three different sample geometries: the ASTM sub-size specimen mentioned in
Section 2.3.1, and two custom geometries (Figure 5). One sample was gripped with pins while
the other was gripped by its wedged ends in self-aligning grips. The authors noted that mean
tensile property values, such as yield strength and UTS, were higher for the bulk custom samples
than for the bulk ASTM sample due to the presence of fewer defects in the smaller volume. For
the additively manufactured samples, property agreement was not as strong, and this was
attributed to differences in solidification dynamics as the samples were directly fabricated.
Figure 5. Drawings of specimens tested by Karnati et al. Figure adapted from ref. [11].
Karnati et al. also tested Cu-Ni functionally graded materials fabricated through laser
metal deposition.12 Specimens were cut from deposited material with the gauge length along the
grading direction and with the same geometry as the MT2 specimen shown in Figure 5. Data
from testing revealed plastic deformation primarily occurred in Cu-rich sections of the gauge
region and failure occurred away from composition interfaces.
14
2.4.3 Weld Characterization
Welding engineering is a third area where miniature tensile testing is applicable.13
Intrinsic to the welding process is the creation of regions in and around welds with dissimilar
microstructures (Figure 6). The properties of these regions can significantly affect the
functionality of welded assemblies, and welding processes are designed to minimize the creation
of undesirable phases such as martensite in steels. Analysis of each of these regions individually
is not possible with conventional tensile testing techniques as too much material is required.
Miniature tensile testing could be utilized to test the mechanical properties of weld fusion and
heat affected zones, as well as the boundaries between zones, in a manner analogous to the
testing of functionally graded additively manufactured metals. Such analysis would aid in the
understanding of weld properties as well as the selection of new weld techniques to optimize the
performance of welded parts.
Figure 6. Steel weld schematic, exhibiting diverse microstructures as a function of weld zone. Figure adapted
from ref. [14]
Kartal et al. investigated the variation in mechanical properties in a 316H stainless steel
pipe multi-pass arc weld using both standard-sized and miniature “micro-tensile” specimens.14 A
standard specimen with a gauge length of 80 mm, width of 10 mm, and thickness of 3 mm was
machined normal to the weld line using EDM. Micro-tensile specimens with a gauge length of
15
3.75 mm, width of 0.7 mm, and thickness of 0.7 mm were EDM cut from the fusion zone, heat
affected zone, and parent material regions in directions normal to the weld line and in the pipe’s
radial direction. Strain was measured using DIC. The authors reported good agreement between
regional tensile properties measured through full field strain mapping of the standard specimen
and those found using the micro tensile specimens.
2.4.4 In-Service Component Evaluation
A fourth area of application for miniature tensile testing is the study of the properties of
in-service industrial components.13,15 As components are used in any application, they are
inevitably affected by stresses, temperature changes, and other environmental factors that can
alter their properties, especially over long periods of time.13,15 In order to study these effects
efficiently by sampling from in-service components, only small volumes of material may be
removed in order to maintain the integrity and operability of these parts.13,15 Such data are useful
for the formulation of safety regulations, design and material selection for new applications, and
a better general understanding of the effects of long-term factors on material properties.13,15
Kumar et al. recommend the use of miniature tensile specimens to characterize material
extracted from operating nuclear reactor pressure equipment to estimate service life.16 The
authors propose this method specifically to minimize radiation exposure and demonstrated the
technique on stainless steel weldments. Proposed tensile specimens have a dog bone geometry
with a gauge length of 5.1 mm, width of 1.0 mm, and thickness of 0.25 mm. The authors also
provide more general guidelines for sample dimensions. They recommend that the gauge length
is greater than or equal to 5.65 times the square root of the area of the gauge region, the thickness
16
is greater than or equal to 10 times the material grain size, and the gauge width is less than or
equal to 8 times the gauge thickness. For testing parameters, crosshead control with a free
crosshead speed on the order of 10-3/s and thus a strain rate on the order of 10-4/s is
recommended.
2.4.5 Nuclear Materials
The area where miniature tensile testing has experienced the most innovation and use is
the area of irradiated metals. The nuclear industry has the most published work of any industry
regarding the fabrication and testing of miniature specimens due to many factors. First, research
is underway to develop new alloys more resilient to the extreme environments of nuclear power
systems, and, as in additive manufacturing and metallic glasses, miniature tensile samples are the
least wasteful method of characterizing the tensile properties of such materials. Second, samples
for close study of the effects of controlled radiation on properties must be low volume as test
reactors and particle accelerators do not have the capacity to contain large conventional samples.
Third, like other industrial components, reactor equipment can only be characterized with small
samples of material in order to preserve functionality and learn safety-relevant property
information. As a result of the strong economic and safety incentives for the nuclear industry,
many miniature tensile testing procedures, devices, and samples have been developed for such
research.
Kohno et al. investigated the effects of miniature specimen aspect ratio and thickness on
yield stress, UTS, and strain at UTS on both irradiated and non-irradiated samples.17 Samples
were comprised of two alloys: JPCA, a modified 316 austenitic steel, and JFMS, a
17
ferritic/martensitic dual phase steel. Samples of two types were fabricated. The first had gauge
dimensions of 1.2(w) x 5.0(l) x 0.5(t) mm3 and the second had twice the width and length.
Specimen thickness was altered by wire saw siding to vary the aspect ratio of specimen thickness
to width from 0.01 to 1. The authors reported that ultimate stress was aspect ratio dependent at
aspect ratios less than 0.4, changing by approximately 100 MPa from aspect ratios from 0.01 to
0.4, but was independent at higher values. Ultimate strain was also found to depend upon aspect
ratio, changing by about 20% from aspect ratios of 0.01 to 0.4.
Panayotou et al. investigated optimal specimen design for irradiated material testing.
Specifically, samples of 316 stainless steel, HT-9, and Alloy 3, a Brush Wellman copper-nickel-
beryllium alloy, were fabricated.18 The sheet specimens fabricated through punching had a gauge
length of 5.1 mm, width of 1.0 mm, and thickness of 0.25 mm. Samples were tested on an
electromechanical test frame. The authors provided a variety of insights from the experimental
data. Elongation data were shown to be a function of the length of the gauge region, and were
reported to be scalable to larger, geometrically similar specimens per Barba’s law. The range of
strength data for miniature samples was shown to be smaller than that for conventionally sized
samples due to the presence of fewer defects in smaller volumes of material.
Klueh conducted a study comparing tensile property values determined by testing of
various miniaturized sample geometries prominent in the nuclear industry.19 The sheet samples
tested were fabricated with three geometries. The first set of samples had a gauge region length
of 20.3 mm, width of 1.52 mm, and thickness of 0.76 mm. The second had a gauge region length
of 12.7 mm, width of 1.02 mm, and thickness of 0.25 mm. The third had a gauge region length of
7.62 mm, width of 1.52 mm, and thickness of 0.76 mm. Specimens were tested at room
temperature, 300°C, and 600°C in a vacuum chamber. Tensile properties measured were found
18
to be in good agreement between samples when factors such as grain size, cold work, and gauge
length were accounted for. Slight differences in properties between sample geometries remained
unaccounted for and further investigation was recommended.
2.4.6 Geometry-Property Relations
Studies involving miniature tensile specimens have discovered several trends connecting
sample geometry and measured properties. First, mechanical property values for miniature
specimens tend to be higher than conventionally measured due to the presence of fewer defects
in the metal, assuming uniform defect density.11 Second, gauge region thickness should be
greater than 10 times the grain size of the metal in order to account for inter-granular
deformation that represents bulk behavior.16 Third, ultimate stress and strain increase with the
aspect ratio between gauge region thickness and width for ratios less than 0.4 for several alloys.17
Another relationship between specimen geometry and measured properties is
described by Barba’s law. The extension at fracture of a tensile sample is the combination of
uniform extension prior to necking and local necking extension, as shown in Equation 2.1.3
𝐿𝑓 − 𝐿0 = 𝛼 + 𝑒𝑢𝐿0 (2.1)
In Equation 2.1, 𝐿𝑓 is the final gauge length, 𝐿0 is the initial gauge length, 𝛼 is local
necking extension, and 𝑒𝑢𝐿0 is uniform extension.3 When both sides are divided by 𝐿0, the total
elongation, 𝑒𝑓, is shown to be gauge-length dependent, as seen in Equation 2.2.3
𝑒𝑓 =𝐿𝑓−𝐿0
𝐿0=
𝛼
𝐿0+ 𝑒𝑢 (2.2)
19
Barba’s Law proposes that local necking extension is a function of a proportionality
constant, 𝛽, and the initial cross-sectional area of the gauge region, 𝐴0, according to Equation
2.3.3
𝛼 = 𝛽√𝐴0 (2.3)
Combining Barba’s Law with Equation 2.2 yields Equation 2.4.3
𝑒𝑓 = 𝛽√𝐴0
𝐿0+ 𝑒𝑢 (2.4)
Equation 2.4 shows that final elongation is dependent on the geometry of the gauge
region. Consequently, elongation values should only be compared if they are found using
specimens with the same ratio of √𝐴0
𝐿0.3 This is in agreement with ASTM guidelines which
indicate that elongation value comparison requires that the ratio between initial gauge length and
square root of initial gauge cross-sectional area be controlled.2
2.5 Structure, Properties, and Processing of SS304
To determine the accuracy and precision of data gathered from the proposed novel
sample geometries, samples of SS304 were fabricated and tested. This alloy is a wrought
austenitic stainless steel and contains the ferrite and austenite phases.20 The presence of
metastable austenite at room temperature is enabled by austenite-stabilizing elements, including
C, N, Ni, and Mn. If the alloy is rapidly quenched or extensively plastically deformed, martensite
form. The crystal structures of austenite, ferrite, and martensite are FCC, BCC, and BCT,
respectively.
20
In addition to stabilizing phases, alloying elements present in SS304 improve mechanical
properties, corrosion resistance, and ease of processing.21 C participates in interstitial solid
solution strengthening, increasing alloy strength and hardness. Mn is present to improve hot
working properties, and to increase toughness, strength, and hardenability. To improve
machinability at the cost of decreased corrosion resistance, P and S are added in small quantities.
P also increases alloy strength. Corrosion resistance is provided by Cr, which forms an oxide
passivation layer on the surface of the steel. Ni also assists in improving corrosion resistance,
strength, and toughness, even at high temperatures. To prevent crevice and pitting corrosion, Mo
is added. Cu is present in small amounts to increase corrosion resistance in sulphuric acid and
seawater. Finally, N increases yield strength and resistance to pitting corrosion.
SS304’s properties vary depending on alloying element amounts and processing steps.
The UTS for a sheet of SS304 less than 8 mm thick ranges from 540-750 MPa.22 Minimum proof
stress and percent elongation for SS304 in the same form are 230 MPa and 45%, respectively.
SS304 has a density of approximately 8.00 g/cm3, a melting point of about 1450 °C, and an
elastic modulus of around 193 GPa. Finally, the alloy is highly resistant to various types of
corrosion including pitting, crevice, and stress-corrosion cracking in acidic, chloride, and basic
environments, with the exception of HCl.23 SS304 is also resistant to attack by most organic
compounds.
SS304 can be processed through a variety of conventional methods. Its high formability
allows for deep drawing, as well as tube drawing and sheet rolling.22,23 The alloy has a high
hardenability and can be cold worked with intermediate annealing steps and a final full anneal.
Annealing is important to reduce residual stresses and defects that could serve as corrosion sites.
21
SS304 is also easily hot worked and has intermediate closed die forgeability. Finally, the alloy
can be machined and fusion welded, although large welds require post-weld annealing.
Given its mechanical, thermal, and corrosion resistant properties, as well as a wide
variety of applicable processing techniques, SS304 is ubiquitous in many applications.23 These
include but are not limited to foil, wire, tubing, food processing equipment, architecture, valves,
decorative automobile parts, and chemical tanks. SS304 can also be employed in elevated
temperature systems including heat exchangers and chemical processing equipment. The
maximum service temperatures are 870 °C for intermittent heating and 925 °C for continuous
heating.23 Finally, SS304 can be used in low temperature applications such as cryogen transfer.
2.6 Engineering Considerations
If adopted, the proposed miniature tensile test methods will have economic,
environmental, sustainability, ethical, health, safety, social, and political ramifications. Many of
these effects will be secondary or tertiary and will depend on the degree of adoption by the
scientific and industrial communities.
Miniature tensile testing will provide a variety of economic benefits to those who use it.
As the method requires smaller samples, more data can be gathered from a given volume of
metal. This will decrease costs and waste production. As force required to test samples is directly
proportional to sample cross-sectional area, smaller samples will require less applied force.
Smaller applied forces can be achieved with miniature test stands which require less power to
operate, further decreasing costs.
22
Environmentally, miniature tensile testing will be beneficial. As mentioned previously,
smaller samples require less material, produce less waste, and can be tested with machines
requiring less power. Less power will in turn generate fewer greenhouse gas emissions and thus
will aid the minimization of humanity’s carbon footprint. Technologies that will be benefited by
the proposed method will also benefit the environment. Certain additive manufacturing
techniques, for example, generates less waste than conventional subtractive manufacturing.
Miniature tensile testing is a sustainable technology, as it is more affordable than
conventional testing, requires less space, and can be applied to a wide range of materials. This
thesis focuses on application to metals, but the technology can be adapted to be used on ceramics
and polymers. A challenge for sustainability is the lack of industry standards for miniaturized
testing. Additionally, miniature test stands like the one used to gather data for this thesis, are not
mass-produced and are custom-made. This lack of supply will limit the spread of the technique.
Several possible ethical issues exist with miniature tensile testing. As this technique is
still being proven through experiments, data gathered using the method must be viewed with
caution. Many more studies must be conducted to develop industry standards for miniature
tensile testing and to ensure reliability. If standardization is not completed and questionable
material data are used in engineering designs, resulting devices will have an increased risk of
failure. In order to prevent this, the method of material data collection must be publicized and
qualified in any works involving miniature tensile testing.
Regarding health and safety, miniature tensile testing has beneficial implications. As
miniature tensile stands apply lower loads on smaller samples, machine malfunctions and
catastrophic failure will result in less damage and danger to operators. However, machines still
23
pose a threat to operators if they are not handled according to their design, so safety standards
need development before widespread use.
Socially and politically, miniature tensile testing will have the effect of decentralizing the
generation of material property data. Due to its decreased financial and spatial requirements,
smaller companies and labs will have the means to test materials and publish their findings.
Individual inventors may be able to afford and operate miniature tensile stands due to the same
factors. This will accelerate material development and the maturation of related technologies
worldwide. Widespread adoption of the technology will undoubtedly be affected by various
government safety regulations that must be developed to protect workers and researchers.
24
Chapter 3
Materials and Procedure
3.1 Sample Composition and Processing
Samples originated from a 0.0351 inch thick, 12 inch by 12 inch sheet of SS304 (ATI
Flat Rolled Products, Vandergrift, PA) that was cold rolled, cut, and annealed. The exact alloy
composition is given in Table 1. Samples were cut from the sheet using a water jet machine
(Omax Corporation, Kent, WA) at the Penn State Bernard M. Gordon Learning Factory. After
cutting, samples remained attached to the SS304 sheet with tabs. These tabs were broken using a
Dremel 3000 rotary tool (Dremel, Racine, WI) with an abrasive disk attachment. Residual tab
material was ground off the samples using a TwinPrep 5 grinding station (Allied High Tech
Products, Inc., Rancho Dominguez, CA) loaded with 120 grit SiC grinding paper.
Table 2. Composition of tested SS304 (wt%).
Fe C Mn P S Si Cr Ni Mo Cu N
Bal. 0.06 0.93 0.032 0.0001 0.40 18.11 8.20 0.46 0.48 0.06
Prior to testing, samples were painted a speckle pattern to enable digital image
correlation. First, samples were taped to index cards in the grip regions and sprayed with a base
coat of American Accents Spray Paint 2X Ultra Cover Flat Spray (Rust-Oleum, Vernon Hills,
IL) which served as a paint and primer. Following the base coat, a black paint was prepared by
mixing five parts Airbrush Medium (Golden Artist Colors, Inc., New Berlin, NY) to four parts
25
Carbon Black Fluid Acrylic Paint (Golden Artist Colors, Inc., New Berlin, NY) and filtering
through a 190 Micron Blue Nylon Mesh Paper (TCP Global, San Diego, CA). The black paint
was lightly applied to the samples using an Iwata Custom Micron CM-B Gravity Feed Dual
Action Airbrush (Anest Iwata-Medea, Inc., Portland, OR) pressurized to 30 psi.
3.2 Sample Geometries
A total of 7 geometries were fabricated, with 5 copies of each of these geometries, termed
Geometries A-G. All samples had the same thickness as the sheet from which they were cut,
which was between 0.0348” and 0.0354” according to tolerances provided by the manufacturer.
The curves connected to the reduced parallel region in Geometry G could not be cut to the
desired radius as the minimum cutting radius for the water jet was about 0.381 mm.
Figure 7. Geometry A-type specimens. All dimensions shown are in millimeters.
26
Figure 8. Geometry B-type specimens. All dimensions shown are in millimeters.
Figure 9. Geometry C-type specimens. All dimensions shown are in millimeters.
27
Figure 10. Geometry D-type specimens. All dimensions shown are in millimeters.
Figure 11. Geometry E-type specimens. All dimensions shown are in millimeters.
28
Figure 12. Geometry F-type specimens. All dimensions shown are in millimeters.
Figure 13. Geometry G-type specimens. All dimensions shown are in millimeters.
29
Due to the untimely shutdown of the laboratory space used for testing during the COVID-
19 pandemic, only samples of Geometries C-F were successfully tested and analyzed. The
measured gauge region widths and thickness of the tested samples are shown in Table 3. These
measurements were averages of three measurements in the gauge region for both width and
thickness.
Table 3. Measured gauge region widths and thicknesses of tensile samples prior to testing.
Sample Gauge Region Width (mm) Gauge Region Thickness (mm)
C1 3.07 0.93
C2 3.01 0.90
D1 3.07 0.94
D2 3.01 0.90
E2 1.57 0.96
F1 1.56 0.95
F4 1.51 0.90
3.3 Tensile Testing Procedure
Once samples were prepped for testing through speckle patterning, they were loaded into
the custom miniature load stage (Sylvan Engineering LLC) (Figure 14).
30
Figure 14. Custom miniature load stage used for tensile tests.
Samples were all gripped using pin grips (Figure 15a). First, spacers were inserted into
the grips to hold sample pins (Figure 15b). Next, pins were inserted into these spacers (Figure
15c). Samples were then placed between the pins once they were measured, and painted (Figure
15d). In order to fit samples between pins, the previously mentioned Dremel tool was used to
thin the region of the samples between the pins, not the reduced parallel region. Spacers were
added after the sample to reach the full height of the grips (Figure 15e-15g). Finally, the spacers
were tightened with a left-handed and right-handed screw on each grip (Figure 15h). Once
samples were successfully installed, they were loaded using the stand and unloaded until the
force and displacement were zeroed.
Displacement Transducer
Sample Grips
Motor
10 mm
31
Figure 15. Two-pin sample grip assembly, proceeding from (a) to (h).
After the samples were installed in the test machine, a GRAS-50S5M-C camera (Point
Grey Research, Inc., Richmond, BC) for DIC was mounted on an aluminum scaffold above the
test stand (Figure 16). Both the test stand and camera scaffold were mounted on an auto-leveling
hydraulic table to prevent any environmental vibrations from affecting data. The camera was
screwed into place and connected to a computer loaded with the VicSnap (Correlated Solutions,
Irmo, SC) image capturing program. Lenses and extenders (Fujifilm, Tokyo, Japan) as well as
camera height were chosen based on sample size so that the reduced parallel region could be
visible and in focus throughout testing. A 10 mm extender was used for C and D-type samples
while a 30 mm extender was used for E and F-type samples. Samples After the camera was set
(
(a)
(
(b)
(
(c)
(
(d)
(
(e)
(
(f)
(
(g)
(
(h)
10 mm
32
up, fiberoptic illuminator lights (Cole-Parmer, Vernon Hills, IL) were positioned around the
sample to ensure clear imaging without over or under exposure.
Figure 16. Experimental setup, including lighting, scaffolding, camera, mini load stage, and control box.
Once setup was complete, the VicSnap program was configured to record images during
the test. First, the frame rate for VicSnap was set to record one image every second. Next, the
camera display was checked to ensure no overexposure was present. As a final step before
Mini Test Stage
Control Switches
Camera Lens
20 mm Lens Extender
Camera
Fiberoptic Lights
Force/Displacement
Control Box
Camera Scaffold
Tensile Sample
10 cm
33
testing, the load rate was selected to fulfill the quasi-static criterion of a strain rate on the order
of 10-4 s-1 and the test mode was set to tension. This rate was unique for each sample type as each
had its own unique gauge length (Table 4). Sample F1 was tested before a regular test load rate
calculation was adopted and thus had a slightly higher load rate than Sample F4. The
corresponding strain rate was still within the quasi-static regime.
Table 4. Sample reduced parallel and gauge region lengths and load rates for analyzed samples.
Sample Type
Reduced Parallel Region
Length (mm)
Extensometer Length
(mm)
Strain Rate
(mm/s)
C1 12 10 0.00072
C2 12 10 0.00072
D1 13 11 0.00072
D2 13 11 0.00072
E2 6 5 0.00072
F1 7 6 0.00081
F4 7 6 0.00073
To begin testing, image and force data acquisition was started in VicSnap. Immediately
afterward, the test stand was turned on to begin force application. The sample was loaded until
fractured occurred. Upon fracture, the test stand and VicSnap were stopped, in that order. Next,
the grip screws were undone, and the samples and spacers were removed. Broken samples were
saved in labeled bags for further analysis. Finally, VicSnap was used to take a picture of a ruler
in the grip for scale.
34
3.4 DIC Strain Measurement
Following mechanical testing of specimens, image data acquired with VicSnap were
analyzed using the Vic2D (Correlated Solutions, Irmo, SC) software. During analysis,
deformation of the sample was calculated using DIC algorithms, which track the movement of
the speckle reference pattern on the sample surface.24 These algorithms divide the images into
square subset regions which include distinct speckle features. Starting from the initial image, the
algorithms create new subsets for each image that are matched to corresponding subsets from the
previous image. After the subsets are matched, the displacement of the geometric center of each
subset from one image to the next is recorded. These displacements are then used to calculate a
strain field for the sample. For DIC analysis, a subset size of 29 pixels and a step size of 7 pixels
were used for images with dimensions of 2448 x 2048 pixels. Subset size refers to the width and
height of the subset square and step size refers to the distance between subset centers. To further
improve accuracy, a seed point was placed in the center of the gauge region as a stationary
reference point.
Once strains were calculated, a virtual extensometer was applied within the program to
track strain between two points within the gauge region (Figure 17). This, along with point
extensometers placed on both ends of the virtual extensometer, was used to generate strain data
for the sample. Gathered data was analyzed along with force data from the VicSnap program in
Microsoft Excel to find stress-strain data for the samples. Values of importance which were
gathered include 0.2 % offset yield strength, ultimate tensile strength, elongation at failure,
elastic modulus, and elongation at ultimate tensile strength. Finally, force-displacement and
stress-strain data were plotted to garner a full picture of tensile behavior.
35
Figure 17. Example DIC strain field and virtual extensometer of C1 sample halfway through testing.
5 mm
36
Chapter 4
Experimental Results and Discussion
4.1 Stress-Strain Curves and Mechanical Property Data
Mechanical property data gathered through testing reveals inconsistent correspondence
with predicted trends. First, all samples exhibited ductile fracture behavior that is typical for
SS304 at room temperature (Figure 18). Initial elastic behavior, strain hardening to a plateau, and
a decrease in strain corresponding to necking are clearly visible in every curve.
Figure 18. Engineering stress vs. engineering strain curves for tested samples.
37
As seen in Figure 18, samples fell into one of three groups with superficially similar
behavior. Samples C2, D2, and F4 exhibited similar yield and ultimate tensile behavior, with
more variance in their elongation (EL) at fracture. Samples C1 and D1 also exhibited similar
yield and ultimate tensile behavior, with C1 having a slightly higher UTS and lower elongation.
Finally, E2 and F1 had the lowest yield strength and UTS values of the group although they had
significant elongation. Table 5 contains several mechanical property values for each of the tested
specimens as well as values provided by the sheet manufacturer.
Table 5. Yield stress, UTS, elongation (EL), and elastic modulus values for samples and from the manufacturer/literature.
Sample 0.2% Yield Stress
(MPa)
UTS
(MPa) EL at UTS EL at Failure
Young's Modulus
(GPa)
C1 341 705 0.51 0.68 198
C2 361 740 0.52 0.68 126
D1 338 693 0.56 0.71 119
D2 352 733 0.52 0.60 140
E2 312 668 0.54 0.69 142
F1 319 652 0.51 0.75 82.7
F4 353 741 0.56 0.66 159
Supplier 290 645 0.66 190-20325
38
To find Young’s modulus values for the specimens, lines of best fit were computed for
the linear elastic regime of the engineering stress vs engineering strain curves. These modulus
values were used to create a line beginning at a stress of 0 MPa and a strain of 0.2%. The stress
at the intersection of this line and the experimental data was taken to be the 0.2% yield stress.
UTS was the maximum stress recorded for each sample, and EL at UTS was the corresponding
strain. EL at failure was taken to be the elongation value after which strain decreased by a value
of 10 MPa or more between data points, an indication of imminent fracture.
All samples measured had higher 0.2% yield stress, and UTS values than those provided
by the manufacturer. This supports the hypothesis that smaller samples exhibit higher strength
than the conventional samples most likely used by the sheet manufacturer. As mentioned in the
background, this would be due to the presence of fewer strength-reducing defects. However, as
seen in Figure 19, UTS and 0.2%YS values do not decrease with increasing gauge region volume
(GRV). Gauge region volume is calculated as the product of gauge region length, width, and
thickness. Values do appear to decrease with increasing gauge region volume within each sample
geometry, but this would need to be further investigated with additional experiments due to data
spread.
39
Figure 19. UTS and 0.2% YS vs. gauge region volume of samples. For a given sample type, both values appear to
decrease with increasing gauge region volume.
Another trend predicted by literature is an increase in UTS and elongation at UTS with
gauge region thickness (GRT) to gauge region width (GRW) aspect ratio values from 0 to 0.4,
followed by a plateau. When plotted, these data do not seem to have any clear relationship. This
is due to a large spread in experimental data as well as a lack of intermediate GRT/GRW values
between 0 and 0.3 and between 0.3 and 0.6 (Figure 20).
Figure 20. UTS and EL vs. gauge region volume of samples. No trends are apparent.
40
The final prediction of literature tested by this set of experiments was that EL at failure
should increase with the ratio of √𝐴0
𝐿0. This trend is shown when EL values are plotted, both
within a sample type and between sample types (except Type C), unlike the other expected
trends (Figure 21). With more tests and data, it is expected that this trend would be more clearly
defined.
Figure 21. EL at fracture vs √(A_0 )/L_0 for tested samples. The two variables are positively correlated as
expected from literature.
4.2 Possible Explanations for Deviation from Predicted Behavior
In addition to a failure to follow predicted trends, several other questions are raised
regarding the relationships between sample geometry and measured material properties. The
most glaring problem with the data gathered in this project is the significant spread of UTS, 0.2%
offset YS, and EL at failure between samples of the same type. Several possible explanations for
exist, particularly machine malfunction and measurement error.
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.00 0.05 0.10 0.15 0.20 0.25
EL a
t Fa
ilure
√(𝐴_0 )/𝐿_0
EL at Fracture vs √(𝐴_0 )/𝐿_0
Type C
Type D
Type E
Type F
41
Samples E2 and F1 exhibited abnormally low strength values compared to those of the
other samples. was the testing procedure. These specimens may have shifted into an improper
orientation during testing and experienced a higher stress concentration. This may be the case as
a Dremel tool was used to help fit the sample in between the pins and may have resulted in non-
uniformities in that area of the sample that could lead to sticking. Also, as seen in Figure 22,
sample E2 failed in the gauge region but not directly in the center. This may indicate that there
was a locally weak region which led to a lower UTS.
Another possible cause of unexpected properties could be the water jet cutting process.
The fabrication method is not especially precise, and local variations in gauge region area may
account for abnormal strength values. This was not accounted for in the experiment as gauge
region width and thickness values used were averages of three measurements and may be larger
than actual minimum cross-sectional area. The initial image of sample E2 prior to loading shows
variation of sample width, especially in the lower gauge region where fracture occurred (Figure
23). Such variation is present in all samples tested. Like E2, F1 exhibits abnormally low yield
stress and UTS, as well as the highest elongation of the samples. Its behavior is most likely due
to machine malfunction which was occurring during its testing. Data for sample F4 may also
have been affected by machine malfunction. Fractographic and microstructural analysis would
indicate if impurities influenced failure in addition to the other error sources mentioned.
42
Figure 22. Sample E2 post-fracture. Fracture occurred off-center and the remaining pieces appear misaligned.
Figure 23. Close-up of sample E2 prior to loading. Roughness can be seen in the lower left-hand side of the
gauge region.
5 mm
5 mm
Roughness
43
Chapter 5
Summary and Conclusions
In this work, the effects of sample geometry on measured mechanical property values were
investigated to further develop a new miniature tensile testing technique for metallic materials. Seven
novel miniature test geometries were fabricated from a sheet of SS304. Five of these geometries were
successfully tested. The resulting data reflected several trends in material behavior with sample geometry
predicted from literature despite large variation in measured properties within sample types. Some but not
all variation may be the result of material defects. Considering these findings, as well as similar force-
displacement data for samples of the same dimensions, the accuracy of gathered data, specifically
measured gauge region dimensions, is called into question.
Overall, miniature samples exhibited higher yield strength, UTS, and elongation values than those
provided by the manufacturer. These higher values agree with literature predictions that attribute
differences to a smaller number of defects contained in miniature sample geometries. However, this does
not appear to be a linear relation as these values did not increase with gauge region volume. Another
prediction from literature, that UTS and elongation at UTS would increase with a ratio of gauge region
thickness to width from 0 to 0.4 and then plateau was not supported by data. This could be due to the low
range of ratios covered by the tested sample geometries. The final literature prediction, that EL at failure
should increase with the ratio of √𝐴𝑜
𝐿𝑜 was reflected in gathered data.
Several additional influences on measured properties were posited. Sample E2’s low strength
values could also have been affected by machine malfunction during testing. Sample F4’s abnormally low
strength values and high elongation value may have resulted from machine malfunction. Gathered data
may also have been affected by errors in measurement of sample gauge region thickness and width, which
were irregular due to the water jet cutting process.
44
This work provided experimental support for trends in material behavior with sample geometry.
If these trends can be modeled with equations based on more experimental data, perhaps new, more
generalizable tensile testing standards may be developed. This would allow for a diverse range of new
sample geometries and testing scales, broadening options for industrial, academic, and experimentalists.
Specifically, this would allow for the further development of miniature tensile testing techniques for low
volume metals in pursuit of further understanding of material behavior. Such testing would assist in the
development of new materials such as metallic glasses, new processes such as additive manufacturing,
and countless other unknown innovations of the future.
45
Chapter 6
Future Work
First, sample geometry tolerances should be altered to improve fit within the miniature test stage
and remove the necessity of Dremel machining. If possible, future samples should be fabricated from a
sheet with a more precise process such as wire electrical discharge machining as the water-jet machining
resulted in irregular sample edge dimensions which may have contributed to fracture. Additionally, more
sample geometries should be designed so that each dimension can be tested as the controlled variable and
so that a broader dimensional range can be covered.
Before testing occurs, camera distance from samples, extender type, and other camera parameters
should be documented. To account for dimensional variation, samples should be measured in their
entirety using an imaging tool. Important dimensions to note would be minimum and maximum thickness
and width in the samples. Ideally, conventional samples from the same lot of material should be tested in
the same conditions in order to make a more direct comparison. The number of samples tested should be
increased in order to increase the statistical significance of gathered data and to account for variation in
properties.
Future work should be ideally used to develop mathematical relations between sample geometry
parameters and measured material properties. It should be conducted for a broader range of metallic
samples, and possibly be expanded to polymeric or ceramic samples where applicable. If such testing
could be accomplished, it should be used to develop novel testing standards so that the benefits of gained
knowledge can be practically applied. Such standard development will require significant time and
monetary investments but would be undoubtedly beneficial to many fields, including novel material
characterization, failure analysis, and quality control.
46
BIBLIOGRAPHY
(1) Aegerter, J.; Gray, T.; Loveday, M. S. Tensile Testing of Metallic Materials: A Review;
Teddington, UK, 2004.
(2) ASTM International. E8/E8M-16a Standard Test Methods for Tension Testing of Metallic
Materials. ASTM International: West Conshohocken, PA 2016. https://doi.org/https://doi-
org.ezaccess.libraries.psu.edu/10.1520/E0008_E0008M-16A.
(3) Tensile Testing, 2nd ed.; Davis, J. R., Ed.; ASM International: Materials Park, OH, 2004.
(4) ASTM International. E4-16 Standard Practices for Force Verification of Testing
Machines. ASTM International: West Conshohocken, PA 2016. https://doi.org/https://doi-
org.ezaccess.libraries.psu.edu/10.1520/E0004-16.
(5) ASTM International. E2658-15 Standard Practices for Verification of Speed for Material
Testing Machines. ASTM International: West Conshohocken, PA 2015.
https://doi.org/https://doi-org.ezaccess.libraries.psu.edu/10.1520/E2658-15.
(6) ASTM International. ASTM E83-16 Standard Practice for Vertifiation and Classification
of Extensometer Systems. ASTM International: West Conshohocken, PA 2016.
https://doi.org/10.1520/E0083-16.
(7) Davies, R. G.; Magee, C. L. The Effect of Strain-Rate upon the Tensile Deformation of
Materials. J. Eng. Mater. Technol. Trans. ASME 1975, 97 (2), 151–155.
https://doi.org/10.1115/1.3443275.
(8) Varshneya, A. K.; Mauro, J. C. Introduction. In Fundamentals of Inorganic Glasses;
Elsevier: Cambridge, MA, 2019; pp 1–18.
(9) Wu, F.-F.; Chan, K. C.; Jiang, S.-S.; Chen, S.-H.; Wang, G. Bulk Metallic Glass
Composite with Good Tensile Ductility, High Strength and Large Elastic Strain Limit.
47
2014. https://doi.org/10.1038/srep05302.
(10) Dongare, S.; Sparks, T. E.; Newkirk, J.; Liou, F. A Mechanical Testing Methodology for
Metal Additive Manufacturing Processes Manufacturing Engineering + Mechanical
Engineering.
(11) Karnati, S.; Axelsen, I.; Liou, F. F.; Newkirk, J. W. INVESTIGATION OF TENSILE
PROPERTIES OF BULK AND SLM FABRICATED 304L STAINLESS STEEL USING
VARIOUS GAGE LENGTH SPECIMENS.
(12) Karnati, S.; Zhang, Y.; Liou, F. F.; Newkirk, J. W. On the Feasibility of Tailoring
Copper–Nickel Functionally Graded Materials Fabricated through Laser Metal
Deposition. Metals (Basel). 2019, 9 (3), 287. https://doi.org/10.3390/met9030287.
(13) Hyde, T. H.; Sun, W.; Williams, J. A. Requirements for and Use of Miniature Test
Specimens to Provide Mechanical and Creep Properties of Materials: A Review. Int.
Mater. Rev. 2007, 52 (4), 213–255. https://doi.org/10.1179/174328007X160317.
(14) Kartal, M.; Molak, R.; Turski, M.; Gungor, S.; Fitzpatrick, M. E.; Edwards, L.
Determination of Weld Metal Mechanical Properties Utilising Novel Tensile Testing
Methods. In Applied Mechanics and Materials; Trans Tech Publications Ltd, 2007; Vol.
7–8, pp 127–132. https://doi.org/10.4028/www.scientific.net/AMM.7-8.127.
(15) Lord, J. D.; Roebuck, B.; Morrell, R.; Lube, T. 25 Year Perspective Aspects of Strain and
Strength Measurement in Miniaturised Testing for Engineering Metals and Ceramics.
Mater. Sci. Technol. 2010, 26 (2), 127–148.
https://doi.org/10.1179/026708309X12584564052012.
(16) Kumar, K.; Madhusoodanan, K.; Rupani, B. Miniature Specimen Technique as an NDT
Tool for Estimation of Service Life of Operating Pressure Equipment. BARC Newsl. 2007.
48
(17) Kohno, Y.; Kohyama, A.; Hamilton, M. L.; Hirose, T.; Katoh, Y.; Garner, F. A. Specimen
Size Effects on the Tensile Properties of JPCA and JFMS. J. Nucl. Mater. 2000, 283–287,
1014–1017. https://doi.org/https://doi.org/10.1016/S0022-3115(00)00245-2.
(18) Panayotou, N.; Atkin, S.; Puigh, R.; Chin, B. Design and Use of Nonstandard Tensile
Specimens for Irradiated Materials Testing. In The Use of Small-Scale Specimens for
Testing Irradiated Material; Corwin, W., Lucas, G., Eds.; ASTM International: West
Conshohocken, PA, 1986; pp 201–218. https://doi.org/10.1520/STP33004S.
(19) Klueh, R. L. Miniature Tensile Test Specimens for Fusion Reactor Irradiation Studies.
Nucl. Eng. Des. Fusion 1985, 2 (3), 407–416. https://doi.org/https://doi.org/10.1016/0167-
899X(85)90028-X.
(20) Manilova, E. P.; Lucas, G. M.; Vander Voort, G. F. Metallography and Microstructures of
Stainless Steels and Maraging Steels. In Metallography and Microstructures; Vander
Voort, G. F., Ed.; ASM International, 2004; pp 670–700.
https://doi.org/10.31399/asm.hb.v09.a0003767.
(21) Aalco - Ferrous and Non-Ferrous Metals Stocklist. Stainless Steels Alloying Elements
https://www.azom.com/article.aspx?ArticleID=13089 (accessed Jan 6, 2020).
(22) Aalco - Ferrous and Non-Ferrous Metals Stocklist. Stainless Steels - Stainless 304
Properties, Fabrication and Applications
https://www.azom.com/article.aspx?ArticleID=2867 (accessed Jan 6, 2020).
(23) Aggen, G.; Washko, S. D. Wrought Stainless Steels. In Properties and Selection: Irons,
Steels, and High-Performance Alloys; ASM International, 1990; pp 841–907.
https://doi.org/10.31399/asm.hb.v01.a0001046.
(24) LePage, W. digitalimagecorrelation.org https://digitalimagecorrelation.org/ (accessed Apr
49
9, 2020).
(25) Properties: Stainless Steel - Grade 304 (UNS S30400)
https://www.azom.com/properties.aspx?ArticleID=965 (accessed Apr 10, 2020).
ACADEMIC VITA
Andrew Johnson abj5175@psu.edu
Education
The Pennsylvania State University, Schreyer Honors College May 2020
College of Earth and Mineral Sciences University Park, PA
Bachelor of Science in Materials Science and Engineering
Work Experience
Intern Summer 2019
GE Aviation Lynn, MA
• Conducted failure analysis investigations for various engine components
• Developed heat treatment schedule for novel Cu-based alloy
• Documented effects of surface finish techniques on additively manufactured parts
Intern Summer 2018
GE Aviation - Unison Industries Norwich, NY
• Facilitated defective part root cause analysis meetings, communicated with inspectors, and
addressed quality issues for sensors department
• Calibrated automated spring test stand, wrote SOP, and created Excel files for data analysis
• Calculated insulative powder usage rate and created a tracking method to regulate powder ordering schedule and decrease inventory
Intern Summer 2017
GE Aviation Evendale, OH
• Investigated microstructure-property relations in nickel-based superalloy samples
• Photographed failed high-pressure turbine blades and compiled images in to investigate field
corrosion patterns
Research Experience
Undergraduate Researcher Spring 2018 – May 2020
Materials Characterization Research Group University Park, PA
• Analyzed additively manufactured Al-10Si-Mg samples fabricated using powder bed fusion in
multiple orientations to investigate microstructure and mechanical property anisotropy.
• Tested Ti64-V-SS304 functionally graded materials to observe trends in mechanical properties
and fracture behavior
Accomplishments
Eagle Scout 2017
• Fulfilled all requirements to achieve the highest rank in the Boy Scouting program of the Boy
Scouts of America.
Skills
• Java
• Engineering Drawing
• Microscopic Analysis
• ImageJ
• MATLAB
• Microsoft Office
top related