Curing-induced residual stresses affect the mechanical performance and failure of fibre-reinforced thermoset composites. Yet, their prediction remains challenging due to heterogeneous microstructures, non-uniform thermal and chemical fields, and pronounced strain localisation in the polymer matrix.This work presents a finite element–based microscale framework to simulate the coupled evolution of temperature, degree of cure, chemical shrinkage, and residual stresses during composite curing. A fully implicit integration scheme for the curing kinetics model is derived. As a key element of novelty, a non-local, strain-gradient plasticity model is incorporated and consistently coupled with the curing simulation to properly capture size effects and strain localisation in the matrix.Numerical results highlight the influence of curing parameters and thermal cycles on microscale residual stress development and premature failure of composites under transverse tension. Accounting for non-local effects is essential to prevent highly localised regions of plastic deformation. The numerical results show that, while the presence of a residual stress field does not justify the differences between the resin properties measured by in-situ nanoindentation and in the bulk, the data reduction methods for the fibre push-in tests need to be revisited. It was also demonstrated that differences in the residual stress field in under-cured resins explain the experimentally found larger strain associated to the onset of transverse matrix cracking in such resins.

Tungsten has been selected as the plasma-facing material for ITER's first wall and divertor. Plasma-facing components for fusion environments must withstand extreme thermal and mechanical loads, particularly during transient events such as edge localized modes. These events generate significant thermal gradients and large mechanical loads, subjecting the material to a wide range of strain rates. Failure of these components could interrupt the reactor operation and potentially damage the surrounding systems, including the test blanket modules. To evaluate the mechanical integrity of material under such conditions, tensile tests were conducted on two ITER-specification tungsten grades for divertor components at temperatures between 250 degrees C and 550 degrees C and strain rates ranging from 6.4 & times; 10_ 4 to 6.4 & times; 10_2 s_ 1. The goal of the study is to determine the viscoplastic behavior of tungsten and to assess the impact of temperature and strain rate on its mechanical performance. The fracture surfaces and initial microstructures were analyzed by scanning electron microscopy. The results reveal a clear strain rate dependency of the strength and ductility, with a ductile-to-brittle transition at lower temperatures and higher strain rates. Differences in performance between the two tungsten grades are linked to microstructural features and impurity content. These findings highlight the impact of the processing route and of initial microstructure on the mechanical performance of plasma facing components.

Accurate prediction of the lifetime of unidirectional fibre composites requires a model that captures matrix viscoplasticity across generic loading histories. We propose a compact, rate-dependent hardening law for highly crosslinked epoxies that unifies constant-strain-rate and creep behaviour. The yield stress depends exponentially on accumulated plastic strain and logarithmically on plastic strain rate, and the relationship is analytically invertible for direct use in a finite element code. Parameters are calibrated from compression tests at multiple strain rates and from hold-at-load creep tests; validation is performed on RTM-6 and new 736LT epoxy data. The model reproduces (i) the near-linear σ _y – logε̇ trend from pre-yield through softening and hardening, (ii) the time-dependent transition from pre- to post-yield during creep, including the rate surge near softening, (iii) long-term (14.5 h) creep more faithfully than stress–time power laws, and (iv) trends in cyclic, variable-rate, and tensile tests. The resulting, easily calibrated formulation enables robust simulation of matrix viscoplasticity in composite-scale models, improving durability predictions for load-bearing structures such as pressure vessels and wind-turbine blades.

Several tungsten grades were irradiated at nominal 400°C, 800°C, and 1100°C to a dose of 0.84 dpa using an artificially tailored neutron spectrum in the BR2 reactor. The irradiated materials includes ITER specification grades, one advanced pure tungsten grade, a W0.5ZrC particle-reinforced grade, and two potassium-doped grades. A 0.5 mm gadolinium shield was employed to suppress the thermal neutron component of the BR2 spectrum, producing a less lethargic neutron spectrum that reduces the tungsten transmutation rate to a nominal value of 0.5 at.% Re/dpa, significantly lower than in unshielded BR2 spectrum where it was over 2 at.% Re/dpa.The objective of this work is to investigate the influence of the reduced thermal neutron fraction on post-irradiation mechanical properties, assessed via room-temperature Vickers hardness and elevated-temperature tensile testing. The lower transmutation rate of tungsten into rhenium and osmium leads to reduced irradiation hardening compared to previous BR2 and HFIR campaigns conducted without thermal neutron shielding. One of the pure tungsten grades, demonstrates superior post-irradiation mechanical properties, probably because of its specific manufacturing process. A warm rolled potassium-doped grade performs better than all the other materials of this study, as a result of strong crystallographic texture and high initial dislocation density, which seems to promote property retention under neutron irradiation for the preferred orientation. Furthermore, fractographic analysis of tensile specimens reveals irradiation-enhanced recrystallization, a consequence of the reduced transmutation rate under this spectrum.These findings underscore the critical role of neutron spectrum tailoring for assessing and benchmarking the radiation response of tungsten-based materials in the context of fusion applications.

The nano-/microscale stiffness and hardness of the crystalline structures developing in polyether ether ketone (PEEK), including individual spherulites and transcrystalline (TC) layers, were determined using nanoindentation and atomic force microscopy (AFM) in force spectroscopy mode in samples containing few carbon fibers. The crystalline morphologies were identified using polarized light microscopy (PLM) in transmission mode. The microstructure and weight percentage of crystallinity of each crystalline entity were also quantified using, respectively, tapping-mode (TM) AFM and Fourier transform infrared (FTIR) spectroscopy, to directly correlate the nano-/micromechanical behavior of PEEK to the crystalline structures. The crystalline structures of various dimensions involve different fraction of crystallinity as well as modulus and hardness values. The variation of the latter within individual crystalline entities can be correlated to a change of fibrils (i.e., lamellar stacks) density. These quantitative data are important in the context of optimizing the performance of PEEK based composites and of developing more predictive multiscale micromechanics-based models.

A new constitutive model for epoxy resins is proposed to accurately capture the micro-scale strain fields in fibre-reinforced composites. The model extends conventional plasticity by introducing strain-gradient effects through an implicit gradient formulation that accounts for pressure sensitivity and asymmetric tension/compression behaviour-critical features to represent the mechanical response of epoxies at small scales. The formulation is implemented in a commercial finite element platform through user-defined subroutines and solved via an analogy with coupled thermo-mechanical problems that include Helmholtz-type equations. The ability of the model to predict the size-dependent response and to prevent unrealistic overpredictions of strain magnitude is illustrated through the simulation of nanoindentation on neat epoxy and of the local strain field behaviour in a composite. The results are compared with experimental data, which are also used to calibrate the additional constitutive parameters through a composite Bayesian optimization strategy. The proposed framework significantly improves the predictive capabilities of micromechanical models for composite materials by incorporating scale-dependent plasticity mechanisms.

Abstract The assessment of irradiation effects in metallic glasses is important considering the renewed interest for this class of material for a variety of nuclear applications. The Zr 70 Ni 30 thin films metallic glass (MG) deposited on Si substrate by magnetron sputtering technique were exposed to 93.2MeV 129 Xe 23+ heavy-ion irradiation at room temperature, covering a range of ion fluences from 5×10 12 ionscm −2 to 8×10 13 ionscm −2 . The evolution of the irradiation-induced defects in Zr 70 Ni 30 MG has been investigated using Doppler broadening spectroscopy (DBS) and positron annihilation lifetime spectroscopy (PALS). Three lifetime components were distinguished, indicating the presence of different types of open-volume regions at the atomic scale in thin film. The combined results of both DBS and PALS demonstrated that ion irradiation initially increases the excess free-volume density with a homogeneous distribution up to a fluence of (≤2×10 13 ionscm −2 ). In contrast, with increasing fluence (>2×10 13 ionscm −2 ), a reduction in excess free-volume was found, which could be related to structural relaxation accompanied by modifications in atomic arrangement and defect distribution. Moreover, the correlation between the shape and wing parameters provides a basis to identify the nature of the defects, indicating that the type of defect changes at the higher fluence of 4×10 13 ionscm −2 and 8×10 13 ionscm −2 , which affects the performance of Zr 70 Ni 30 metallic glass after ion irradiation.

Stress-induced amorphization has recently attracted attention as a potential deformation mechanism in ceramics, semi-conductors or minerals. Its activation is promoted when conventional plasticity, e.g. driven by dislocations, is inhibited. However, the mechanisms underlying this phenomenon are still unclear. Based on quantitative in situ TEM tensile testing of small-sized olivine bi-crystals, we demonstrate that stress-induced amorphization and grain boundary sliding can be activated under high stresses at room temperature in specimens with high angle grain boundaries. Low angle grain boundaries are less prone to this phenomenon. Varying the iron content in olivine demonstrates that iron inhibits amorphization and, consequently, promotes brittle failure. This contrast with the accepted view that iron promotes ductility but at high temperatures. These findings coming from natural minerals provide a novel approach regarding the control of the mechanical properties of hard materials at low temperatures.

Proper functioning and reliability of future fusion tokamaks, including ITER, will heavily depend on the capacity to exhaust extreme heat fluxes. Tungsten (W)-based monoblocks constitute a critical core structural component. These are made of copper-chromium-zirconium (CuCrZr) cooling ducts which are joined to W using a copper (Cu) interlayer. These monoblocks must ensure efficient heat exhaustion while preserving the structural integrity under extreme operating conditions. The characterization and modelling of the interface cracking resistance of W/Cu joints is key. A single parameter interface toughness is not rich enough to encompass a wide range of stress states, especially when plasticity develops in one of the adjoining materials. The goal of this work is to develop a robust methodology to determine interface toughness and an approach to transfer data from small scale specimens to real components. A combined methodology using miniaturized three-point bending specimens and cohesive zone modelling was developed and validated, first, at room temperature. Bimaterial specimens with a notch are tested, exhibiting a ductile mode fracture with dimples on the Cu side, which suggests good delamination resistance of W/Cu joints. Cracking is accompanied by plastic dissipation. The peak strength and interface energy cohesive zone parameters are identified providing excellent correspondence between simulated and experimental load-displacement curves. Validation is made by comparing plastic zone sizes and by connecting the values to a micromechanical analysis.

The origin of plasticity in solids can often be attributed to local collective rearrangement of atoms or molecules in regions named shear transformation zones to which an eigenstrain is associated. This approach is currently receiving a recrudescence of interest, owing, among others, to its versatility for application to metallic glasses, polymers, ionic glasses as well as to crystals deforming by dislocations or twinning. The Kinetic Monte Carlo method has been widely adopted in the literature to simulate this class of mechanisms at the mesoscale, supplemented with a finite element description in order to solve boundary value problems. In this method, the time step is computed explicitly based on the magnitude of the current stress, which may sometimes lead to significant inaccuracies. Several studies in the literature have proposed alternative approaches to address this issue. The problem is serious as the choice of the numerical algorithm may sometimes create contradictions with the physical time. In this work, different formalisms are proposed and numerical experiments are performed to study the effect of these modifications for the one and the two-dimensional cases. Different geometries, and material parameters that correspond to metallic glasses are considered, and the results are compared with those obtained from the statistical mean-field model. Finally, an implicit approach for computing the time step size is described, enabling accurate predictions of both the time step and the corresponding external load.

Several structural components of technological relevance include elastoplastic interlayers, such as in adhesive joints, laminated composites, or multilayered coatings. The effective toughness strongly depends on the plastic dissipation within the interlayer and on the crack path, being both coupled. Predicting where the crack propagates within the layer is therefore essential for determining the cracking resistance of the assembly. An approach to simultaneously predict plastic dissipation inside an elastoplastic adhesive joint and the corresponding crack path is explored using a steady-state formalism. A cohesive-zone approach is adopted, with the crack path selected based on the phase angle, quantifying the mode-mixity at the crack tip extracted from the mode I and II works of fracture in the first cohesive element. This approach is validated against LEFM-based criteria and the J-vector criterion from static simulations. For an elastic layer, all criteria lead to the same crack-path selection. For an elastoplastic layer, the cohesive-element-based method and the J-vector criterion predict similar paths and reveal that plastic deformation and crack trajectory are coupled. This leads to crack deflection towards the interfaces when the zone of intense plasticity is larger than about one-tenth of the adhesive thickness. Consequently, plastic dissipation alongside elastic properties, joint dimensions, and T-stress influence the crack path in adhesive joints, which is important not only in the context of structural integrity assessment but also towards the design of efficient dissipative interlayers. In addition, the crack selects the path corresponding to the lowest fracture toughness, highlighting the importance of predicting the crack path.

The development of fusion and generation IV fission reactors poses huge challenges on the development of structural materials capable to withstand high operating temperatures and severe neutron irradiation. A recently launched European project INNUMAT advances the development of new alloys by assessing ageing and damage in fission/fusion operational conditions e.g. including neutron irradiation, liquid metal corrosion tests and longterm annealing. A systematic comparison has been conducted on seven candidate metallic alloys, including well-characterized alloys such as Eurofer97 and 15-15Ti as well as advanced materials such as high entropy alloys (HEA), oxide dispersion strengthened (ODS) steel and alumina forming austenitic (AFA) steel. The goal is to investigate the microstructure and mechanical properties via hardness, tensile tests, Small Punch Tests (SPTs) and then address the microstructure-property relationships. The ODS steels demonstrate enhanced strength and ductility as compared to Eurofer97, and HEA and AFA steel exhibit a significant improvement of ductility when compared with 15-15Ti. In addition, a good correlation between SPT and tensile results confirms the reliability of SPT techniques which is essential for the application of SPT for future post-irradiation analysis. These results form a solid baseline for the next step of determining the effect of neutron irradiation for these metals.

The mechanical properties of amorphous olivine (a-olivine) deformed at room temperature are investigated in situ in a TEM under uniaxial tension using a Push-to-Pull (PTP) device. Thin films of a-olivine were produced by pulsed laser deposition (PLD). With or without electron irradiation, a-olivine films deform plastically, with a gradual transition that makes impossible the determination of a precise threshold. The strength attains values up to 2.5 GPa. The increasing strain-rate in load control results in an apparent softening with stress drop. The fracture strain reaches values close to 30 % without e-beam irradiation. Under electron illumination at 200 kV, the strength is lower, around 1.7 GPa, while higher elongations close to 36 % are obtained. Alternating beam-off and beam-on sequences lead to exceptionally large fracture strains equal to 68 % at 200 kV and 139 % at 80 kV. EELS measurements were performed to characterize the interaction between the electron beam and a-olivine. At a voltage of 80 kV, radiolysis accompanied by oxygen release dominates whereas at high voltage (300 kV) the interaction is dominated by knock-on type defects. Radiolysis is also the main interaction mechanism at 200 kV with low exposition which corresponds to most of our in situ TEM deformation experiments. To interpret the mechanical data, a simple 1D model has been developed to rationalize the load transfer between the PTP device and the specimen. The strain-rate sensitivity is 6 to 10 times higher when a-olivine is deformed under electron irradiation.

Neutron irradiation of structural materials in nuclear applications is accompanied with a variety of microstructural changes including the generation of helium nano-voids at high doses. When operating under stress, irradiation creep is taking place, leading to change of dimensions of components, as well as relaxation of the springs or screw torque. These distortions may have important implications on the performance of the nuclear reactor. Irradiation creep is particularly complicated to investigate and the link between creep rate and microstructure remains partly elusive. Ultra-miniaturized testing of thin films makes possible the use of ion irradiation and implantation methods to emulate the effect of in-reactor neutron irradiation. Here, on-chip tests on freestanding copper films reveal much slower irradiation creep of helium implanted layers compared with non-implanted copper, with or without ion irradiation hardening prior to testing. This enhanced resistance to creep is due to the presence of the high density of helium bubbles, as evidenced by transmission electron microscopy, that impede dislocation glide and climb. A unified elementary model for moderate to high stress creep under irradiation is proposed showing that the creep law for the studied range of stress primarily depends on the yield strength which increases with helium bubbles.

Tensile properties such as strength and ductility are essential for structural integrity assessment of critical components. In the context of nuclear applications, the flat tensile geometry as compared to the reference standard cylindrical geometry offers a number of advantages in terms of material use efficiency, ease of machining, best packing under irradiation and simplified remote handling of active samples. Accordingly, the interchangeability of data extracted from flat and cylindrical specimens is a key issue. Furthermore, this interchangeability must be demonstrated for irradiated samples. Many metallic materials show significant reduction or even a full lack of uniform elongation after neutron irradiation, with most of the strain hardening regime taking place during the post-necking stage. As the necking development depends on geometry, this raises questions on the validity of changing the test specimen geometry. Here, the interchangeability of mini-flat and cylindrical tensile samples after neutron irradiation is investigated by combining experimental and computational analysis. The investigated material is EUROFER97 steel irradiated at 300 degrees C in the conditions relevant for the ITER fusion reactor. Finite element (FE) simulations are performed with a Gurson-type ductile fracture model parameterized based on the experimental tensile response. The hardening law extracted from mini-flat samples and applied to predict the stress-strain response of a cylindrical sample with 90 % accuracy or better in terms of total elongation, reduction of area, and fracture strength as compared to the direct experimental data obtained with cylindrical tensile specimen geometry.

Nanoindentation (NI) and atomic force microscopy (AFM) nanoindentation, coupled with polarized light microscopy (PLM), were used to determine the nano-/micromechanical behavior of the amorphous regions and individual crystalline structures, both spherulites and transcrystalline (TC) layers, in PEEK samples containing few carbon fibers. To this aim, thin model samples with a controlled thickness were manufactured to allow both microstructure characterization in transmission mode and indentation tests without substrate effects. Surface roughness of the model samples was carefully minimized to get reliable and low dispersion from indentation experiments. The artefacts and sources of uncertainty of performing indentation experiments on thin polymer films containing some fibers are also discussed. The AFM nanoindentation added value is the possibility of evaluating the mechanical behavior of crystalline structures at the nanoscale, for the determination of mechanical behavior heterogeneities at the intra-spherulitic and intra-transcrystalline scale.

The effect of strain-hardening on ductile crack growth is explored based on a small scale yielding finite element approach using an advanced nonlocal Gurson model. A focus is put on considering high strain hardening exponent n up to 0.5, while classical literature is often limited to n = 0.2, in order to encompass materials like stainless steels as well as several modern TRIP-TWIP alloys and high entropy alloys. First, J2 plasticity-based simulations are performed to set the static crack reference. These simulations provide a hint about the origin of the increase of fracture toughness with increasing n, connected to much smaller finite strain zones at a given loading level quantified by the value of the J integral. In addition, it is found that above n similar to 0.3, the opening stress does not attain a maximum value at a distance equal to one to two crack openings but keeps increasing towards the surface of the blunted crack tip. Then, Gurson-based simulations are used to determine the JR curve for different n and initial porosity, and associated quantities related to crack initiation such as JIc, critical crack tip opening displacement Sc, and fracture process zone length. As already found in earlier studies, both JIc and Sc increase with increasing n, although the effect is much more marked on JIc. The origin of this first-order effect is unraveled by looking at the stress triaxiality, damage, and plastic strain fields. Even though the near crack tip stress triaxiality increases with n, the associated lower plastic strain at a fixed distance to the crack front leads to much lower void growth rates and delays void coalescence. As a important side result, the simulations appear very sensitive to an accurate fine-tuning of the adjustment factors entering the Gurson model at high strain hardening, pointing towards the intrinsic limitations of the model when n is large. This study confirms the interest in developing alloys with large strain hardening capacity, not only with respect to tensile properties but also in view of enhancing the ductile fracture toughness.

The nucleation and interaction mechanisms of hexagonal close packed (epsilon-HCP) martensite in a Fe50Mn30Co10Cr10 high entropy alloy were investigated using atomic-resolution scanning transmission electron microscopy and molecular dynamic simulations. We demonstrate that the nucleation of epsilon-HCP is controlled by the activation of different mechanisms, including the six-plane and stair-rod cross-slip mechanisms. The nature of the intricate interaction mechanisms between two conjugated epsilon lamellae are highly dependent on the level of deformation and on the difference of thickness between the two crossing epsilon-HCP. Various absorption and transmission mechanisms of the incoming epsilon-HCP across the primary epsilon-HCP are revealed, involving local lattice rotation, the formation of new boundaries as well as stress induced twinning and phase transformation at the intersection sites. The synergy/competition between these mechanisms is discussed and compared with recent literature. These findings shed new light on the elementary mechanisms at the origin the remarkable strain hardening capacity of this category of high entropy alloys.