General Context General ContextThe general ambition of the project is to contribute to a better understanding and prediction of the multilevel interactions between interfaces present in materials and the deformation and failure mechanisms, in order to inspire the development of new materials with extreme structural performances. The field of research sits at the frontier between materials science and solids mechanics. One of the main driving forces for this project is the recognition that the significant progresses recently made in the development of nanostructured materials, especially towards ultra hard systems, must now be integrated into a more hierarchical approach of the material structure involving the control of multiple length scales. This is a necessity for the implementation of a multiproperty vision of materials, considering that several mechanical properties can only be improved if larger length scales are properly architectured. Among the features that can be used to modify and optimize the mechanical properties of materials, interfaces, such as grain, twin and phase boundaries, free or anodized surfaces, play a central role. It is indeed the combined effect of all these interfaces that interact with, trigger, or mitigate the deformation and fracture mechanisms that ultimately control the mechanical response of materials. As a matter of fact, the effect of the presence of interfaces is often larger than the effect of the chemical composition. The focus will be on mechanical properties involving the strength, ductility, fracture, creep, fatigue and wear resistance of metal based systems with a high density of interfaces organized at different scales, possibly involving organic layers. More precisely, the investigations will be performed on 3D bulk materials and on small dimensions systems in the form of thin films or nanowires with a high density of interfaces. Some key scientific issues concern the importance of rate dependent behaviour and back stress originating from the abundance of these interfaces, the stress/strain driven formation and mobility of these interfaces, the interactions with the elementary atomistic deformation mechanisms and with the cracking mechanisms, and the resulting size effects. Meeting this grand ambition will allow unravelling how interfaces can be best organized over a range of length scales in materials with hierarchical structures designed to enforce the adequate mechanisms. The practical outcomes consist in the development of predictive models and of improved processing routes as well as improved characterization methods towards the development of new materials with enhanced performances. ObjectivesThe field of research covered by this project rests at the frontier between materials science and solids mechanics. The general ambition of the project is to contribute to a better understanding and prediction of the multilevel interactions between interfaces and deformation and failure mechanisms in order to inspire the development of new materials and of new material architectures with extreme structural performances. The focus will be on mechanical properties such as strength, ductility, fracture, creep, fatigue and wear resistance of metal based systems, possibly involving organic layers, with the target to build predictive models and improved processing routes towards future technological developments. More than working on single properties, the target is to address combination of properties. Meeting this grand challenge should allow unravelling how interfaces can be best organized over several length scales in materials with hierarchical structures in order to enforce the adequate mechanisms. The characterisation of the complex interactions between defects and interfaces, which most often takes place at the atomic level, requires a panoply of advanced experimental techniques. New constitutive model formulations that account for interfaces have to be formulated and improved numerical methods have to be developed for properly representing the different scales and the presence of the, possibly evolving, interfaces. The investigations will be systematically performed on 3D bulk materials and on small dimensions systems in the form of thin films, both involving a high density of the interfaces of interest. Nanowires will also be produced from the films. Research GoalsThe specific objectives are organized in terms of categories of interfaces, setting also a natural workpackage (WP) structure for the project, in the following way. WP1 - 'Static and dynamic interfaces' WP2 - 'Localized bands and cracks' WP3 - 'Graded and architectured interfaces' MotivationsThe quest for structural materials with extreme properties has been intense over the last decade motivated by new progress in the control of the nanoscale structure of metallic materials. New strengthening mechanisms have been unravelled, leading to very hard materials, with a resistance near the theoretical limit for perfect crystals (e.g. Greer and De Hosson, 2011, Richter et al., 2009). The main route for triggering these strengthening mechanisms has been by producing very far-from-equilibrium nanocrystalline structures via processing methods such as large and rapid deformations, fast thermal treatments, powder-route manufacturing, or thin film deposition (e.g. Meyers et al., 2006, Zhu et al., 2010). These advances have lead to a realm of remarkable scientific breakthroughs and paved the road for a new generation of structural materials, though with one major shortcoming: these new materials have a relatively low ductility and, often, a poor fracture toughness. Today, the materials science and mechanics of materials scientific communities strive both to develop a comprehensive understanding of the extreme strength attained in nanostructured materials and to invent new approaches for achieving a better balance of mechanical performances combining strength with good ductility, fracture toughness, interface toughness, fatigue resistance and/or wear resistance. Time has come to move to a multiproperty vision requiring the integration of the developments made at the nanoscale into a more hierarchical approach of the material structure where the control of multiple length scales allows additional degrees of freedom. The far-from-equilibrium processing methods used to produce nanocrystalline metals, offer a potential which has only been partially explored, going well beyond the “simple” refinement of the grain size. By proper control of temperature, rate, stress state and composition, combined with additional thermal treatments, various types of interfaces can be simultaneously generated at very small scales, such as grain, twin and phase boundaries, precipitates, dislocation cell walls, microshear bands, cracks or gradients of composition. It is the combined effect of all these interfaces that interact with, trigger, or mitigate the deformation and fracture mechanisms that ultimately controls the mechanical response of materials. As a matter of fact, the effect of the presence of interfaces is often larger than the effect of the chemical composition. For instance, the ductility in fine or nanograined materials can be restored owing to the presence of finely spaced nanotwin boundaries without loss of hardening (Lu et al., 2009). The optimum strength/ductility balance is attained for specific couples of grain size and twin boundary spacing through interacting elementary mechanisms (Zhu et al., 2011, Li et al., 2010) illustrating that interfaces engineered at the right length scales can profoundly influence the macroscopic properties. The physics related to combining different types of interfaces at the nano- and microscale is very rich with deformation and fracture mechanisms competing or cooperating with one another, leading to cascades of other phenomena or couplings with higher scales mechanisms. This richness of open and challenging scientific questions added to the potential to simultaneously address several mechanical properties provides the most important motivation for this project. Additional degrees of freedom can be offered when the nano- and microstructures and the associated interfaces are architectured at different scales by combining different materials (soft, hard, graded) and different topologies (inclusions, layers, fibres, porosity). For instance, it is now realized, inspired by biological systems (Munch et al., 2008, Espinosa et al., 2009), that the fracture toughness of seemingly brittle materials can be significantly improved through hierarchically organizing structures over several length scales. This effort of developing new material architectures, while taking into account the progress in nanostructured systems, is needed to address properties such as fracture toughness, fatigue resistance and wear which have received much less attention than the hardness. A rather unique example of this vision is illustrated by recent progresses in flexible electronics science. The technological motivation to produce flexible electronic devices has been instrumental in the search for more ductile metallic systems with small dimensions. To remain electrically conductive thin metallic lines, with thickness between 20 nm and 1 µm, must deform without cracking or at least without percolation of cracks. The thin films, which have a natural high strength owing to small grain sizes must resist fracture. Different intrinsic and extrinsic routes have been attempted to ductilize the films (e.g. Sun and Rogers, 2010, Lacour et al., 2003). The ductility is enhanced by controlling the elastic mismatch with the underlying elastomeric substrate, by patterning and pre-stretching the substrate, by controlling the roughness and adhesion as well as by playing with the intrinsic structure of the film such as its texture. Recent progress in materials processing, in characterization and testing techniques and in multiscale modelling of mechanical phenomena constitute also a motivation for addressing these complex issues and for expecting new discoveries. These techniques are briefly listed in section 3 “state of the art”, and described in more details next in form D. Most of these advanced techniques and tools are available in Belgium in the groups involved in the present project, taking also into account the contribution of the foreign partners and additional collaborations in Belgium. It proves the existence of a critical mass of expertises combined with long term joint collaborations. A final motivation for this research is that addressing the fundamental questions raised above can directly lead to the development of new materials with improved structural performances with a direct impact on applications. The long-term economical development of our society is indeed facing critical and urgent challenges in the management of energy and raw materials resources, and of the environment questions, such as those concerning solid wastes and end-of-life goods. High performance materials with better properties go along with more efficient use of resources, safer and more reliable applications, and longer life of the systems made of these materials. Transportation pushes for lighter and safer materials, coatings must be more wear resistant to allow longer protection times, energy production, harvesting or recovery requires materials with versatile properties often at high temperature, microelectronic and micro-electromechanical devices must become more reliable. State of the Art (General)A general state of the art is presented in this section first in terms of the current understanding of the mechanisms of interactions between different types of interfaces and the deformation and fracture mechanisms, with a focus on nanostructured materials or materials dominated by a high density of these interfaces. Finally, general trends on modelling, processing, and characterization are outlined. In Form D, more detailed coverage of the state of the art will be provided, directly connected to the investigation proposed in different research tasks. Interface dominated deformation and failure mechanismsThe traditional view regarding interfaces such as grain boundaries (GBs), phase and twin boundaries is that they constitute obstacles to dislocation motion leading to pile-ups, local stress concentration, geometrically necessary dislocations and back stress. The most common microstructure-based size effect is that the strength depends inversely on the interface spacing (typically to the power one-half), i.e. the famous Hall-Petch effect (Hall, 1951, Petch, 1953). It is only recently that rigorous constitutive models have been capable to incorporate these size effects, by accounting for plastic strain gradients in a rigorous elastoplasticity framework (Fleck, 1997). This simple view of dislocation barriers has been challenged with the advent of nanostructured materials, in which the interface spacing is reduced down to 100 nm or even smaller. Dislocation mechanisms still control the deformation process but without the classical forest type hardening. The stress level can reach very high values due to the lack of mobile dislocations (Meyers et al., 2006). The result is that mechanisms other than dislocation blocking start being activated at the interfaces, such as dislocation nucleation and annihilation, and dislocation transmission. These mechanisms have been evidenced using advanced TEM methods (Kumar et al., 2003) and by MD simulations (Li et al., 2010, Van Swygenhoven, 2002). These mechanisms imply short range interactions and are thus thermally activated. Hence, the interface behaviour becomes more complex than a simple rate independent impenetrable barrier to dislocations. In terms of properties, extreme hardness is observed sometimes near the theoretical limit for perfect crystals, and this has attracted most of the attention of the community up to now (e.g. Greer and De Hosson, 2011, Richter et al., 2009). Unfortunately, these nanostructured materials often present a poor ductility due to a lack of strain hardening. It is only during the last decade that remedies have been proposed to restore the ductility either by combining different types of interfaces such as nanotwins with nanograin sizes (Lu et al., 2009) or by playing with other length scales such as with bimodal grain size distributions (Wang et al., 2002). Investigations on other properties such as fracture toughness, fatigue and wear in these systems have been quite limited up to now, see (Meyers et al., 2006, Dao et al., 2007). The effects of the intrinsic rate dependency and of the back stress controlled hardening on the strength, ductility, toughness, wear constitute a widely open field of investigation. Interfaces are not always static but can evolve which offers additional degrees of freedom and additional complexity in the mechanics of materials. Deformation twins and strain or stress induced phase transformation involve dynamic interfaces that form, move and multiply during deformation leading to an evolution of the microstructure characteristic lengths. The interface spacing becomes an evolving quantity, which significantly impacts the strain hardening capacity (see e.g. Bouaziz et al., 2008, Gil Sevillano, 2008, Lani et al., 2007). The potential of the TRIP (Transformation Induced Plasticity) effect (Lani et al., 2007, Fischer et al., 2000, Herrera et al., 2011) and of the TWIP (Twinning Induced Plasticity) effect (Bouaziz et al., 2008, Idrissi et al., 2010) has been demonstrated in a variety of single- and multiphase steels exhibiting strength/ductility balance far outside the usual envelope. The controlled generation of hard interfaces during deformation provides a continuous source of isotropic and kinematic hardening. The combination of these effects within a nanostructured microstructure is still an open field of investigation. Grain subdivision during plastic deformation constitutes another example of evolving interfaces. If the global average texture evolution can be predicted reasonably well (Beyerlein and Tóth, 2009), in particular with models incorporating orientation fragmentation along GBs due to grain-grain interaction (Tóth et al., 2010), predictive models for band-type grain subdivision are still in an early stage. Among the problems that this type of modelling is facing are the strong directionality of the evolving substructures that requires an anisotropic treatment of work-hardening and the changing number of texture components which is difficult to implement in polycrystal plasticity models. Recently, the attention of the community has been attracted by observations of different types of GB migration or sliding mechanisms occurring at room temperature in nano or ultra fine-grained materials, as observed by in-situ TEM (Rupert et al., 2009) or in MD simulations (Cahn et al., 2006). The large amount of energy stored in the high density of GBs combined to the lack of mobile dislocations to relax the stress provide the driving force for GB migration. All these mechanisms of interface evolution are thermally activated leading to rate dependent behaviours, with an effect on the strength and ductility (Gianola et al., 2006). Plastic localization phenomena from micro to macro shear bands are known from the early stages of plastic deformation analysis, see e.g. review by Rice (Rice, 1976). Still, the questions of the microstructural origin of the thickness of the band, of the deformation mechanisms taking place inside the band with strong plastic strain gradients and back stress, and of the interaction of the band with other interfaces, remain open. Clear bands constitute one example of localization involving soft elongated regions free of radiation defects inside an irradiated hard matrix (Byun and Farrell, 2004). It is commonly accepted that they appear as a consequence of the local removal by dislocations of radiation-induced dislocation loops, but the details of the formation mechanisms of such bands remain unclear. Plastic localization in nanocrystalline materials occurs early due to the low strain hardening capacity (Meyers et al., 2006). The potential of postponing plastic localization owing to a large strain rate sensitivity controlled by the addition of twin boundaries in copper has been recently discussed in the literature (Lu et al., 2009) calling for more investigations. Macro-shear bands have dimensions exceeding those of single grains, i.e. the bands are polycrystalline in nature. In the past, texture softening and/ or damage accumulation have been found to be possible causes for macro-shear banding (Anand and Kalidindi, 1994) but the prediction of time and place where such defect would appear in a metal forming process has so far not become accessible because FE softwares for forming processes do not allow for an accurate prediction of the joint evolution of the crystallographic texture and of the damage. Adiabatic shear bands (ASBs) constitute another type of macro-shear bands where thermal softening is at the origin of the strain localisation, which may produce a strongly deformed nanocrystalline area (e.g. Meyers et al., 2003). A better understanding and modelling of ASBs requires the knowledge of the micro- and nanostructure at different stages of the deformation. Roping and ridging is a surface localization phenomenon observed on metal sheets in forming operations. Surface damage becomes visible after plastic deformation imposed next to the thermo-mechanical treatments which is undesirable for the reflectivity, lubricant transport, weldability and adhesion (Hirsch, 2006). Several crystal plasticity finite element models (CPFEM) are proposed in the literature to predict the surface roughness that typically develops during forming but none of the models is capable to accurately predict the experimentally observed roughness profiles (Kusters et al., 2010). It is also known that deformation processes can lead to a complex near surface layer (Zhou et al., 2011). Plastic localization and fracture by accumulation of damage in small dimension metallic systems has not yet been investigated in depth, compared to the attention dedicated on strength. A reason is that most conventional test methods involve compressive loadings (nanoindentation, nanopillar compression) preventing the occurrence of fracture. The combined effect of the small thickness and fine microstructure on plastic localization is still quite open, see (Pardoen et al., 2010) for a review. Finally, the competition between ductile failure by plastic localization and void coalescence under shear dominated loading is currently an important, if not the main, question of interest in the ductile fracture community (Barsoum and Faleskog, 2007, Tvergaard and Nielsen, 2010). This question is essential for the control of forming operations working under large shear deformations. The state of the art related to graded and architectured interfaces is rather wide as it touches many sub-fields of materials science. A recent tendency in the quest for new performances is to integrate the (often neglected) length scales intermediate between the nano- or microstructure and the macroscopic level to produce architectured or hybrid materials, see (Ashby and Bréchet, 2003, Chehab et al., 2009, Embury and Bouaziz, (2010), as inspired by biological systems (Munch et al., 2008, Espinosa et al., 2009). This effort towards developing new material architectures encompasses the efforts towards multi-property enhancement owing to the combination of several materials or phases presenting the proper contrast of mechanical properties among a given microstructure or assembly, e.g. see recent review (Mortensen and Llorca, 2010). Graded interfaces, in a broad sense, constitute a key ingredient in this quest towards new architectures as it allows progressive change of properties from one material to another, from one scale to another, and as it sometimes delivers intrinsically new properties. For instance, severe plastic deformation (SPD) can lead to partial mixing of constitutive elements when two materials are co-deformed. Several authors have shown (Sauvage et al., 2005, Raabe et al., 2010) that SPD can cause over-passing of the solubility limit, creating out-of-equilibrium structures that present interesting mechanical properties. The insertion of soft layers in between hard layers can be seen as a sort of graded interface to control the load transfer, cracking behaviour, energy absorption and integrity of the interfaces. Examples of combinations of polymer and metallic layers leading to enhanced performances, either thin or thick, involve blast resistant structures (Rathbun et al., 2006) and recent developments in flexible electronics applications (Sun and Rogers, 2010, Lacour et al., 2003). The impact of surface passivation also directly affects the plasticity mechanisms of the underlying metallic film by creating barriers to dislocations which would otherwise leave the crystal, as shown in (Nicola et al., 2006, Xiang and Vlassak, 2006). Playing with various sorts of passivated layers is one possible way of architecturing 2D systems. Multiscale modelling of interface dominated materialsCurrent higher gradient continuum micromechanics based constitutive models provide a framework to treat size effects and (simple) interface effects, while allowing the treatment of large, representative boundary value problems (Fleck and Hutchinson, 1997, Fleck and Willis, 2009, Gurtin and Anand, 2008, Abu Al-Rub and Voyiadjis, 2006). These higher order theories and/or classical crystal plasticity theories, informed by lower scale molecular dynamics (MD) or discrete dislocation dynamics (DDD), constitute the current medium to transfer atomistic effects to upper scales. The first attempts to enrich the description of interfaces into strain gradient based constitutive models based on a physical description of the mechanisms of deformation have been made in the last few years (Gudmundson, 2004, Aifantis et al., 2006, Massart and Pardoen, 2010), but the road to physically justified interface laws is still long. Interface fracture and sliding are often modelled using cohesive zone models (Needleman, 1987). The highly successful ALAMEL model for deformation texture prediction is in fact an interface-dominated version of the old Taylor model. New progresses in finite element formulations and in homogenization schemes permit numerically efficient scale transitions up to the macro-level under static or dynamic conditions, with thermal couplings and update of the underlying microstructure (Ozdemir et al., 2008). New approaches to treat moving interfaces are offered by applying XFEM and level sets (Moës et al., 2002). MD and 3D DDD numerical simulations, see books (Raabe, 1998, Yip and Diaz Rubia, 2009), deliver extremely useful information, difficult to extract experimentally but on simplified or small enough model systems. The main limitation of MD remains the difficulty to treat thermally activated mechanisms due to the limitation in time increments. ProcessingModern processing techniques offer new avenues to produce ideal and/or new materials with 2D or 3D architectures. The processing of 2D systems involve thin films and multilayers deposited in accurately controlled conditions by CVD, PVD, ALD, laser cladding and melting, electrochemistry, with possible anodized layers, involving precipitates, twins or solid solutions, as well as soft polymeric and nanoporous layers, with patterns made by FIB or lithography (Ohring and Gall., Materials Science of Thin Films).The processing of new 3D systems involves ultra fine grained alloyed materials produced by large and fast deformation methods such as ECAP, dynamic torsion, friction stir processing, explosive forming, cumulative roll bonding with additional macrolithography, de-alloying and powder metallurgy techniques (Zhilyaev and Langdon, 2008). New material architectures as discussed above require adding in the toolkit of processing methods assembly techniques such as welding, gluing, interlocking, or molecular assembly and bonding (Ashby and Bréchet., 2003). Characterization and testingRegarding characterization and testing, new advances in in-situ mechanical tests coupled to TEM or synchrotron and neutron diffraction, to microdiffraction or tomography, to surface analysis, or to film deposition allow determining the mechanical responses while modifying the microstructure under constant monitoring, see (Robertson et al., 2011) for a recent review. Additional progress in X-ray CT scan, crystallographic mapping in SEM and TEM, with on-chip mechanical testing of thin films (see (Gianola and Eberl, 2009) for a recent review on nanomechanical testing), with nanoindentation and nanoscratching, with nanoDMA, interface cracking and wear tests contribute to increase the panoply of techniques to probe mechanical effects starting from the atomic scale. |