Modeling and testing miniature torsion specimens for SiC joining development studies for fusion

doi:10.1016/j.jnucmat.2015.07.044

Journal of Nuclear Materials

Volume 466, November 2015, Pages 253-268

https://doi.org/10.1016/j.jnucmat.2015.07.044 Get rights and content

Highlights

•
Finite element damage models developed and applied to understand miniature torsion specimen.
•
Damage models correctly predict torsion joint failure locations for wide range of materials.
•
Tests of strong, stiff ceramic joints will not produce accurate shear strength values.
•
Miniature torsion specimen has diminished test utility but still valuable.

Abstract

The international fusion community has designed a miniature torsion specimen for neutron irradiation studies of joined SiC and SiC/SiC composite materials. Miniature torsion joints based on this specimen design were fabricated using displacement reactions between Si and TiC to produce Ti₃SiC₂ + SiC joints with SiC and tested in torsion-shear prior to and after neutron irradiation. However, many miniature torsion specimens fail out-of-plane within the SiC specimen body, which makes it problematic to assign a shear strength value to the joints and makes it difficult to compare unirradiated and irradiated strengths to determine irradiation effects. Finite element elastic damage and elastic–plastic damage models of miniature torsion joints are developed that indicate shear fracture is more likely to occur within the body of the joined sample and cause out-of-plane failures for miniature torsion specimens when a certain modulus and strength ratio between the joint material and the joined material exists. The model results are compared and discussed with regard to unirradiated and irradiated test data for a variety of joint materials. The unirradiated data includes Ti₃SiC₂ + SiC/CVD-SiC joints with tailored joint moduli, and includes steel/epoxy and CVD-SiC/epoxy joints. The implications for joint data based on this sample design are discussed.

Introduction

Joining of SiC and SiC-composites has been identified as a critical technology for the use of these materials in either future fusion reactors or in fission power reactors [1], [2], [3], [4], [5], [6], [7]. The international fusion materials community is currently irradiating and testing several joint types and compositions in the High Flux Isotope Reactor (HFIR) reactor at Oak Ridge National Laboratory (ORNL) [1]. Pacific Northwest National Laboratory (PNNL) is working with Politecnico di Torino (POLITO) and ORNL using miniature torsion specimens, also referred to as torsion hourglass samples, that have been specifically designed for joint shear strength testing using small irradiation volumes (see Fig. 1) [8]. Ceramic joints and joint shear testing have been studied for many years and several shear joint tests have been designed for ceramics and ceramic composites [4], [7], [9], [10], [11], [12], [13]. These include shear lap tests [14], asymmetric 4-point bend tests [11], [15], and double notch shear tests [16], [17]. Each of these tests has some disadvantages; including stress concentrations leading to non-uniform shear stresses [17], [18], [19] that create large uncertainties in shear strength values. In addition, these test specimens are all quite large, or conversely are not miniature-type tests, whereas irradiation volumes are small and demand miniature specimen designs. The miniature torsion geometry, therefore, was designed to provide a test specimen consistent with small irradiation test volumes associated with in-reactor irradiation testing [3], [8], [20], [21], [22], [23], [24] and to provide a more consistent shear strength test. The experimental data for some joint configurations revealed excellent data reliability and reduced data scatter [21], [24] for this specimen design. However, recent high-strength joints fabricated for SiC and SiC-composites have revealed that this test specimen design also has some problems [20], [22], [23], namely out-of-plane fractures that fail to provide simple shear strength values. Since the miniature torsion test appears to be reliable under certain conditions, but not others, and since this design is ideal for testing many joints in a small irradiation volume, a mechanical model of this joint was created to help understand the observed shear failures under a range of simulated conditions.

The PNNL joints, which are synthesized using displacement reactions between TiC and Si, are observed to fail out-of-plane, or in the base SiC material, during torsion testing [1], similar to what others had observed for high-strength joints [20], [22], [23]. Most of the types of joints reported in Ref. [1] exhibited out-of-plane or base material failures, including the glass ceramic joints from POLITO. This type of failure within the base material, while encouraging with regard to joint strength since it implies that the joint is as strong or stronger compared to the base material, does not allow accurate comparisons between types of joints, tailored joints, or failed joints. In particular, post-irradiation joint testing reported in Ref. [1] clearly showed some issues related to the shear strength measurement of several different strong joints. Some observations were consistent with irradiation-induced joint damage and, in the case of the PNNL displacement reaction joint and of glass-ceramic (CA) joints, a fracture mode change from base material to in-plane of the joint was observed that could be interpreted with an appropriate model [1].

Therefore, this study was undertaken to determine 1) if some simple modifications to the miniature torsion specimen could be used to address this problem and 2) if a mechanics-based model could better quantify the joint failure response. The first step was to modify the specimen geometry to reduce the joined surface area of the torsion samples in order to understand the effects of fracture initiation on the outer annulus versus an inner annulus and the complex effect of total joint surface area on fracture. The second step was to create a finite element model of the miniature torsion specimen with parameters that could be varied over the wide range of tested materials from POLITO [3], [20], [21], [24]. This study will also use the model results to discuss the recent data set obtained at ORNL using the HFIR test reactor and the pre- and post-irradiation test results from several joints [1].

Section snippets

Miniature torsion specimen and modifications

The standard miniature torsion specimen (Fig. 1) was used for the majority of the tests reported here, designated as full-bonded joints and referred to as torsion hourglass samples (THGs). Reduced area annular joints were made by dimpling one of the surfaces with either a 2.3 or 3.1-mm diameter diamond slurry drill and are referred to as reduced-area torsion hourglass samples (RATHGs).¹ These joined samples were fabricated

Approach

This section summarizes the construction of elastic-damage and elastic–plastic damage models and computational method for predicting failure initiation and propagation [30], [31] in the joined specimens. The models consider the joined material, either CVD-SiC or steel, together with a thin joint region consisting of either Ti₃SiC₂+SiC of varying modulus and strength or an epoxy. The epoxy-bonded steel specimens simulate samples made and tested at POLITO to help validate the miniature torsion

Unirradiated torsion tests for Ti₃SiC₂+SiC joints

Table 2 lists the unirradiated dual-phase Ti₃SiC₂+SiC displacement reaction joints that were tested and summarizes the results. Fig. 5 is a graph of the results, including the unirradiated THG joints tested at ORNL using similar equipment and test parameters as POLITO. The strong 40 MPa THG joints all fail in the base material such that the entire THG sample is failed and, thus, any strength values are considered as torsional resistance values. For the CVD-SiC material and THG machining

Model predictions and comparisons

The damage models were created to help understand the fracture results from the THG specimens that exhibited non-planar fracture that was not truly reflective of joint properties. Rather, the literature refers to this data as “torsional shear resistance” of the THG [1], [3], [8], [20], [22], [23] when the THG specimen fails in the base material, or out-of-plane. The critical part of the damage models was to be able to simulate the stress-strain curves for the constitutive materials; otherwise

Conclusions

Miniature torsion specimens, referred to as THG specimens, often fail within the joint body or out-of-plane of the joint plane when the THG material is a ceramic, such as CVD-SiC, and the joint is strong. This study developed elastic and elastic–plastic damage models that demonstrated that for a wide range of joint moduli and strengths that out-of-plane failure is predicted in agreement with observations until the modulus of the joint material falls below about 1/3 of the CVD-SiC modulus. The

Acknowledgments

This research was supported by Office of Fusion Energy Sciences, U.S. Department of Energy (DOE) under Contract DE-AC05-76RL01830. PNNL is a multi-program national laboratory operated by Battelle Memorial Institute for the US Department of Energy under DE-AC06-76RLO 1830. The authors thank Prof. Jacques Lamon for helpful discussions.

References (50)

Y. Katoh et al.
J. Nucl. Mater.
(2014)
M. Ferraris et al.
J. Nucl. Mater.
(2011)
P. Colombo et al.
J. Nucl. Mater.
(2000)
R.H. Jones et al.
J. Nucl. Mater.
(1995)
M. Ferraris et al.
J. Nucl. Mater.
(1994)
H.-C. Jung et al.
J. Nucl. Mater.
(2011)
C.A. Lewinsohn et al.
J. Nucl. Mater.
(2000)
C.A. Lewinsohn et al.
J. Nucl. Mater.
(2002)
H. Serizawa et al.
J. Nucl. Mater.
(2007)
M. Li et al.
Carbon
(1999)

L.L. Snead et al.

J. Nucl. Mater.

(2007)

J. Renard et al.

Compos. Sci. Technol.

(1993)

T.S. Byun et al.

Acta Mater.

(2004)

V.S. Deshpande et al.

J. Mech. Phys. Solids

(2008)

C.K. Chen

Fatigue Fract. Eng. Mater. Struct.

(2004)

M. Ortiz et al.

Int. J. Solids Struct.

(1990)

J. Henager et al.

J. Nucl. Mater.

(2011)

P. Colombo et al.

J. Mater. Sci.

(1998)

H. Serizawa et al.

Mater. Trans.

(2002)

H. Serizawa et al.

Mater. Sci. Forum

(2007)

M. Salvo et al.

J. Am. Ceram. Soc.

(1997)

N.J.J. Fang et al.

J. Am. Ceram. Soc.

(1993)

P. Dadras et al.

Exp. Mech.

(1990)

D. Fujita et al.

Ceram. Eng. Sci. Proc.

(2005)

Cited by (16)

Progress in development of SiC-based joints resistant to neutron irradiation
2020, Journal of the European Ceramic Society
Citation Excerpt :
Studying neutron irradiation effects on SiC joints requires a reliable mechanical test method applicable to a miniature test specimen because of the limited volume of irradiation capsules (e.g. 50 × 6 × 6 mm3 in this study). Previous efforts in the development of miniature specimen testing techniques established the validity of torsional shear testing of hourglass ceramic joints [15,16]. The advantages of torsion testing over existing test methods such as single-lap offset tests (ASTM D905–08) and asymmetrical four-point bending tests (ASTM C1469–10) include small test specimen size (6 × 6×3 mm3), pure shear failure of specimens achieved for certain joint types, and a uniform stress state at the outer surface of the bonding layer in the torsion specimen.
This study fills a knowledge gap regarding neutron-irradiation resistance of SiC joints for nuclear applications, by investigating high-dose neutron irradiation effects on the strength of selected joints and low-dose neutron irradiation effects on recently developed joints fabricated by state of the art processing methods. The joining methods used for the high-dose radiation study included pressure-assisted liquid-phase sintering (LPS) of SiC nanopowder, pressureless calcia-alumina glass ceramics joining, and reaction sintering of Ti-Si-C powders with hot-pressing. The joints were neutron-irradiated at 530 °C to 20 displacements per atom (dpa). Other joining methods included low-pressure LPS of cold-pressed SiC green body, pressureless reaction sintered Ti-Si-C powder joint, spark plasma–sintered Ti diffusion bond, and hot-pressed Ti diffusion bond, which were irradiated at ∼500 °C to ∼2 dpa. There was no notable degradation of torsional strengths of the joints following the high-dose irradiation. The irradiation-induced degradation at low neutron dose was highly dependent on joint type.
Finite element optimization of sample geometry for measuring the torsional shear strength of glass/metal joints
2020, Ceramics International
Citation Excerpt :
The results revealed that there is a dependency of measured nominal shear strength (τnominal) on specimen's configuration, for which the results of recent finite element analyses [32–34] could not present a reasonable explanation. Indeed, the simulations were done using axisymmetric models assuming that deviation from axial symmetry in the square flanges does not affect the central portion of specimen [32–34]. Also, it was recently observed that, contrary to flexural strength, the shear strength of steel/sealant joint decreases with increasing temperature [16,20].
Assessment of mechanical properties of glass/metal joints is a challenging process, especially when the application relevant conditions of the joints have to be considered in the test design. In this study, a finite element method (FEM) is implemented to analyze a torsional shear strength test designed for glass-ceramic/steel joints aiming towards solid oxide fuel/electrolysis cells application. Deviations from axial symmetry of the square flanges (ends) of respective hourglass-shaped specimens and also supporting and loading sockets of the test set-up are included in the model to simulate conditions close to reality. Undesirable tensile stress and also shear stress concentration appear at the outer edge of glass-ceramic layers, which are less for the hollow-full specimen. The simulation results show that for a specimen with either 9 mm thick square- or 6 mm thick triangular-flanges, locally enhanced tensile stresses almost disappear, resulting in a symmetric shear stress distribution. The difference between the analytically derived nominal shear strength and the real critical shear stress derived via simulation reduces with decreasing the fracture torque.
Recent progress in the development of SiC composites for nuclear fusion applications
2018, Journal of Nuclear Materials
Citation Excerpt :
The irradiation-induced cracking seen in SiC joints was numerically studied to understand the material degradation mechanism and optimize joint microstructures based on modeling. Henager et al. found that modulus of the bonding layer largely affects the stress state of the hourglass joint in torsion and consequently affects the failure location [86]. In the case of a stiff bonding layer, stress is effectively applied not only within the bonding layer but also within the substrate, which explains the observed experimental substrate failures.
Silicon carbide (SiC) fiber reinforced SiC matrix composites continue to undergo development for fusion applications worldwide because of inherent advantages of the material including low activation, high temperature capability, relatively low neutron absorption, and radiation resistance. This paper presents an international overview of recent achievements in SiC-based composites for fusion applications. Key subjects include applications in fusion reactors, high-dose radiation effects, transmutation effects, material lifetime assessment, and development of joining technology (processing, test method development, irradiation resistance, and modeling capability). This paper also discusses synergy among research for fusion materials and non-fusion materials (for fission and aerospace applications). Finally, future research directions and opportunities are proposed.
A multiscale microstructural approach to ductile-phase toughened tungsten for plasma-facing materials
2018, Journal of Nuclear Materials
Increasing fracture toughness and modifying the ductile-brittle transition temperature of a tungsten-alloy relative to pure tungsten has been shown to be feasible by ductile-phase toughening (DPT) of tungsten for future plasma-facing materials for fusion energy. In DPT, a ductile phase is included in a brittle tungsten matrix to increase the overall work of fracture for the material. This research models the deformation behavior of DPT tungsten materials, such as tungsten-copper composites, using a multiscale modeling approach that involves a microstructural dual-phase (copper-tungsten) region of interest where the constituent phases are finely discretized and are described by a continuum damage mechanics model. Large deformation, damage, and fracture are allowed to occur and are modeled in this region that is connected to adjacent homogenized elastic regions to form a macroscopic structure, such as a test specimen. The present paper illustrates this multiscale modeling approach to analyze unnotched and single-edge notched (SENB) tungsten-copper composite specimens subjected to three-point bending. The predicted load-displacement responses and crack propagation patterns are compared to the corresponding experimental results to validate the model. Such models may help design future DPT composite configurations for fusion materials, including volume fractions of ductile phase and microstructural optimization.
Joining and machining of CMCs
2018, Comprehensive Composite Materials II
Ultrasonic inspection provides a noninvasive means of probing the interior of a composite structure. The primary purpose is for the verification of integrity, quality, and reproducibility of the final product during manufacture; as a diagnostic tool for in-service evaluation during maintenance of a component; and as a final characterization in a retirement-for-cause setting.
The following discussion first introduces the basis of the ultrasonic method as it relates to composites and then presents highlights of a few selected advanced methods of ultrasonic inspection.
Modeling thermal and irradiation-induced swelling effects on the integrity of Ti<inf>3</inf>SiC<inf>2</inf>/SiC joints
2017, Journal of Nuclear Materials
Citation Excerpt :
It is hoped that this modeling method can be used to help design improved joints for future irradiation tests by indicating material incompatibilities in advance of expensive in-reactor exposure tests. A typical miniature torsion hourglass (THG) specimen with a fully bonded joint is shown in Fig. 1 [1,4] that was used in the HFIR irradiation study tested at 800 °C and 5 dpa. This paper studies the same THG specimen made of CVD-SiC substrates joined by a 15-μm Ti3SiC2/SiC dual-phase bonding layer or joint.
Previously, results for CVD-SiC joined by a solid state displacement reaction to form a dual-phase SiC/MAX phase joint subsequently irradiated at 800 °C to 5 dpa indicated some cracking in the joint. This paper elucidates the cracking origin by developing a model that accounts for differential thermal expansion and irradiation-induced swelling between the substrate and joint materials by using a continuum damage mechanics approach with support from micromechanical modeling. Damage accumulation in joined specimens irradiated at four temperatures (300 °C, 400 °C, 500 °C and 800 °C) is analyzed. We assume the experimental irradiation dose is sufficient to cause saturation swelling in SiC. The analyses indicate that the SiC/MAX joint survives irradiation-induced swelling at all the irradiation temperatures considered. The joint experiences only minor damage when heated to and irradiated at 800 °C as well as cooling to room temperature. The prediction agrees with the experimental findings available for this case. However, the joint heated to 300 °C suffers severe damage during irradiation-induced swelling at this temperature, and additional damage after cooling to room temperature. Irradiation at 400 °C and subsequent cooling to room temperature produced similar damage to the irradiation 300 °C case, but to a lesser extent. The joint heated to 500 °C and irradiated at this temperature suffered only very minor damage, but further moderate damage occurred after cooling to room temperature.

View all citing articles on Scopus

View full text

Modeling and testing miniature torsion specimens for SiC joining development studies for fusion

Highlights

Abstract

Introduction

Section snippets

Miniature torsion specimen and modifications

Approach

Unirradiated torsion tests for Ti3SiC2+SiC joints

Model predictions and comparisons

Conclusions

Acknowledgments

J. Nucl. Mater.

J. Nucl. Mater.

J. Nucl. Mater.

J. Nucl. Mater.

J. Nucl. Mater.

J. Nucl. Mater.

J. Nucl. Mater.

J. Nucl. Mater.

J. Nucl. Mater.

Carbon

J. Nucl. Mater.

Compos. Sci. Technol.

Acta Mater.

J. Mech. Phys. Solids

Fatigue Fract. Eng. Mater. Struct.

Int. J. Solids Struct.

J. Nucl. Mater.

J. Mater. Sci.

Mater. Trans.

Mater. Sci. Forum

J. Am. Ceram. Soc.

J. Am. Ceram. Soc.

Exp. Mech.

Ceram. Eng. Sci. Proc.

Unirradiated torsion tests for Ti₃SiC₂+SiC joints