Global megatrends, including demographic and climate changes, urbanisation and the limits to resources and energy are the drivers of future change [Strategic Foresight: Towards the 3rd Strategic Programme of Horizon 2020, 2015]. The unprecedented trend of population growth in a resource constrained world increasingly forces business and policy makers to integrate sustainability considerations into their decision making.
Non-energy and non-agricultural raw materials underpin the global economy and our quality of life. They are vital for the word’s economy and for the development of environmentally friendly technologies such as renewable energy systems. Especially the EU is highly dependent on imports, and securing supplies has therefore become crucial [ERA-MIN Research Agenda, 2013; Strategic Implementation plan for the European Innovation Partnership on Raw Materials, 2013].
The last 150 years of industrial evolution have been dominated by a one-way or linear model of production and consumption in which goods are manufactured from raw materials, sold, used and then discarded or incinerated as waste. In the face of sharp volatility increases across the global economy and proliferating signs of resource depletion, the call for a new economic model is getting louder. The quest for a substantial improvement in resource performance across the economy has led businesses to explore ways to reuse products or their components and restore more of their precious material, energy and labour inputs. A circular economy is an industrial system that is restorative or regenerative by intention and design. The economic benefit of transitioning to this new business model is estimated to be worth more than one trillion dollar in material savings [World Economic Forum & Ellen MacArthur Foundation, 2014].
In a linear economy the functionality is lost after a first use or in the best case after some down cycling phases. In a circular economy the goal is to keep the functionality and therefore value of a material as high as possible over a time period as long as possible. Materials will circle throughout the economy without being removed from it in the form of non-functional waste. The circular economy is the economic system in which resources are kept at the highest possible level of functionality at all times.
Figure 1: Material functionality in a linear and a circular economy [VITO, 2015].
Moving from the traditional, linear ‘make, use, dispose’ economy to a circular economy requires increased reuse, remanufacturing and recycling of products. This is an important aspect of the EU’s strategy to ensure the security of raw materials supply [EIP Raw Materials Scoreboard, 2016].
Advanced engineering materials and technologies present indispensable and exciting solutions for optimal resource use, substitution of critical materials, metal recovery, recycling of waste streams, and shorter loop closures.
In minimising the use of materials, advanced material technologies obviously contribute to lifetime extension and repair of products and especially to the use of ever less materials to provide a certain function to a product. To give just a few examples:
The substitution of critical raw materials (CRM) is another approach to mitigate the supply risks of raw materials. As substitution research takes many years to provide realistic solutions, it is a real insurance policy to develop timely research on substitution, to make available a set of options for possible preventive changes in the product design and the elaboration of contingency plans. Examples include:
Resource recovery and waste recycling
While showing vast potential also in this context, advanced material technologies remain particularly unexplored in resource recovery, cycle closure and the use of recycled materials in products. Here are some examples:
It must be stressed though that closing material cycles will not avoid the (sustainable) mining of primary raw materials that are necessary to sustain global population growth. Recycling can significantly contribute to though not to secure the supply of (critical) material resources in the raw materials constrained European economy.
Feeding & closing loops in the Circular Economy
As such, advanced material technologies are key to sustainable mining and recycling, to feed and close cascaded material and product cycles in a viable, growing circular economy.
Figure 2: Circular economy scheme, taken from [Circular economy, A new relationship with our goods and materials would save resources and energy and create local jobs, Walter R. Stahel, NATURE, Vol. 531, 24 March 2016].
Resource efficiency in manufacturing and processing, substitution of critical raw materials, resource recovery and recycling straightforwardly match the above concept of the Circular Economy. Non technological, new business concepts can also greatly support the effective use of raw materials. With the increased provision of services instead of products from economic production, product and material loops can be closed shorter than recycling. Advanced engineering materials can play a major role here as well, e.g. by extending the life time of products by improving the wear and corrosion resistance (shortest cycle), or providing controlled adhesion/release properties to facilitate remanufacturing. Moreover, material technologies may enable the building-in of sensors and communication systems in an Internet of Things approach, to monitor the status of products in a sharing community.
The role of EuMaT ETP
The newly established EuMaT WG8 on Raw Materials will act as the leading forum to contribute to the debate about the key role of advanced engineering materials and technologies in resource efficiency, substitution of critical materials, metal recovery, recycling or shorter cycle closure of products and waste, providing market and science based, realistic solutions for the EU manufacturing and processing industries.
The working group aims at triggering research and innovation ideas & activities in the H2020 Research Programme targeting the Societal Challenge of Resource Efficiency and Raw Materials, as well as Leadership in enabling and industrial technologies (LEIT).
As such, sustainable materials management touches upon all seven H2020 societal challenges, in particular the ones addressed in SC5 Climate Action, Environment, Resource Efficiency and Raw Materials.
EuMaT WG8 Raw Materials Chair
Research Manager Sustainable Materials Management Unit
VITO Vision on Technology
Boeretang 200, B-2400 MOL, Belgium
Finland has a real opportunity to create sustainable well-being and a successful carbon-neutral circular economy over the next 5 to 10 years. It maximises the conservation of materials and their value in circulation for as long as possible which, in turn, keeps the volume of emissions they produce to a minimum. The road map shows how to make the transition to a circular economy.
The change requires co-operation across sectoral and industrial boundaries. In many cases, the most attractive opportunities for new operating methods, for services made possible by digitisation and for extending the circulation of materials can be found somewhere in the middle.
"The effectiveness of new solutions ultimately stems from the fact that they can be expanded and duplicated both elsewhere in Finland and around the world," says , Director of Sitra - a public fund aimed at building a successful Finland for tomorrow. She stresses that the work is just beginning. "Systematic change will require more fresh and even radical ideas in the future."
According to Sitra estimates, the circular economy would generate 2 to 3 billion euros in added value each year by 2030. The Club of Rome estimates that over 75,000 new jobs would be created.
The world's first national circular economy road map was drafted under the direction of Sitra in co-operation with the Ministry of the Environment, the Ministry of Agriculture and Forestry, the Ministry of Economic Affairs and Employment, the business sector and other key stakeholders.
"Large Finnish companies have already embraced the change they will be required to make in order to take on the new global challenges they will face," says Sitra Senior Lead "The world needs clean and smart solutions more urgently than ever before in order to ensure well-being that is based on the sustainable use of the environment."
The essence of the circular economy road map is to bring together a large number of outstanding pilots. The Technology Industries of Finland, University of Jyväskylä and a group of companies are developing a facility for recovery of precious metals from electronic waste. Currently 50 million tons of electronic waste is generated globally and only 15% of that is treated properly. At the same time, we are facing a shortage of several precious metals that are needed to manufacture electronic equipment and growing amounts are needed in new applications such as LED lighting.
The Minister of Agriculture and the Environment hopes that the circular economy will boost Finland's economic growth and competitiveness: "The road map published today challenges us all to make changes. I'm very happy to see so many committed to advancing the circular economy. We need small-scale, rapid trials and long-term policies of change," he says.
In June 2017, Sitra will hold the world's first international circular economy conference in Helsinki (www.wcef2017.com).
Animation of the recycling facility for electronic waste, WeeeFINer:
Over the past few years direct bonding of III-V semiconductors on Si has emerged as a promising alternative to hetero-epitaxy for the hybrid integration of active – amplification, emission in the direct gap III-V - and passive – guiding, switching in the indirect gap Si - optical functions in next-generation photonic integrated circuits (PICs) [1-9]. Such PICs offer a variety of advantages, foremost among which is the dense integration of advanced optical functions using sub-100nm patterns in the Si guiding layer. The processes used to pattern the Si may, however, deteriorate the quality of the hybrid bonded interface. Given the localized nature of such patterns it is highly desirable to have a technique to evaluate this quality on or in the immediate vicinity of these patterned regions.
The quality of a bonded interface, i.e. the strength of adhesion between the III-V and Si is expressed in terms of the surface bonding energy – the energy per unit area required to adiabatically and reversibly separate the two materials. This surface bonding energy can be measured in a variety of experiments during which the hybrid bonded stack is deformed until the interface yields. Among these experiments, the double-cantilever beam experiment (DCB) has been repeatedly shown to yield reliable measurements in the case of Si on Si and InP on Si bonding [4,5]. Nonetheless, given the centimetric samples required to carry out the measurement, it is better suited to wafer-scale measurements of the surface bonding energy.
The authors have developed a nano-scale analog to the DCB experiment. In this case, instrumented nano-indentation is used to locally deform the InP membrane. Within a certain range of applied indentation loads, the InP membrane is elastically debonded in the vicinity of the indent, forming a blister next to each facet of the indent. The geometry of the blister is recorded using atomic force microscopy and, using well established models for the DCB experiment, the surface bonding energy is measured. Scanning transmission electron microscopy is also used to understand the underlying mechanism responsible for the debonding and to correlate the geometrical features of the blister with those of the buried debonding crack that are required to measure the surface bonding energy [10-12]. The method is here applied to InP membranes bonded to Si using a variety of bonding methods. The application of the method on InP bonded on sub-100nm patterned Si is also discussed.
Figure 1: Low-magnification BF-STEM image of an InP/Si hybrid bonded stack. Direct oxide-free bonding was used in the present case. The image shows a cross-section of the blisters that forms during the nano-indentation experiment. The imprint of the indenter tip is clearly visible in the image, as is the dislocation-dense zone, located immediately below the imprint. The inset shows a magnification of the debonding crack that starts at the left edge of the dislocation-dense zone. From Reference .
Figure 2: AFM images (6 x 6µm) of debonding blisters in InP bonded (a) oxide-free on bare Si, (b) oxide-free on patterned Si, (c) thin oxide-mediated on bare Si, and (d) thin oxide-mediated on patterned Si. The height scale for all images is 55nm. After Ref. .
Using the AFM images and theory [13,14], the surface bonding energies for InP bonded to Si were measured and found are ~1J/m² in both direct oxide-free and oxide-mediated bonding on bare Si. A higher bonding energy is expected to be attained at higher annealing temperatures – such temperatures may, however, be undesirable in CMOS-compatible process flows for the targeted PICs. Nonetheless, a surface bonding energy equal or higher than 1J/m² indicates strong covalent bonding that result in a mechanically stable interface, suitable for most PIC applications.
In the case of InP membranes bonded to patterned Si, the debonding blisters where found to extend farther away from the indent for an equivalent height. This observation, a priori, indicates a lower debonding energy. This is not, however, true as the surface bonding energy has to be related to the actual surface available for bonding. When bonding to patterned Si – the patterns are square arrays of holes – the surface available for bonding is approximately 78.7% of the Si surface. The surface bonding energy measured in this case, however, is 57% lower than the one measured on bare Si for the same bonding process. A closer inspection of cross-sections of the samples in STEM revealed that the edges of the patterns where rounded off, and that actual surface available for bonding is indeed closer to this value.
The method combining instrumented nano-indentation in conjunction with AFM or STEM, is shown to provide reliable, precise, and localized measurements of the surface bonding energy of InP bonded to Si in a variety of bonding schemes. The localized nature of the measurement method render particularly useful in evaluating the adhesion of InP to nano-patterned Si – configurations that prevail in advanced PICs. The underlying mechanism for debonding relies on mechanical properties that InP shares with other III-V semiconductors. The method presented here can, therefore, be applied to other III-V/Si or III-V/III-V hybrid bonded stacks.
K. Pantzasa, E. Le Bourhisb, G. Patriarchea, G. Beaudoina, and A. Talneaua
1. G. Roelkensn, J. Broukaert, D. Van Thourhout, R. Baets, R. Notzel, M. Smit, J. Electrochem. Soc., 153, G1015 (2006).
2. M. Lamponi, S. Keyvaninia, C. Jany, F. Poingt, F. Lelarge, G. de Valcourt, G. Roelkens, D. Van Thourhout, S. Messaoudence, G. H. Duan, IEEE Photonics Technol. Lett., 24, 76 (2012)
3. P. Maszara, G. Goetz, A. Caviglia, J. B. McMitterick, J. Appl. Phys., 64, 4943 (1988)
4. F. Fournel, L. Continni, C. Morales, J. Da Fonseca, H. Moriceau, F. Rieutord, A. Barthelemy, I. Radu, J. Appl. Phys., 111, 104907 (2012)
5. D. Pasquariello, K. Hjort, J. Electrochem. Soc., 147, 2343 (2000).
6. Q. Y. Tong, Q. Gan, G. Hudson, G. Fountain, P. ENquist, Appl. Phys. Lett., 84, 732 (2004)
7. D. Liang, D. C. Chapman, Y. Li, D. C. Oakley, T. Napoleone, P. W. Joudawlkis, C. Brubaker, C. Mann, H. Bar, O. Raday, J. E. Bowers, Appl. Phys. A, 103, 213 (2011)
8. A. Talneau, C. Roblin, A. Itawi, O. Mauhuin, L. Largeau, G. Beaudoin, I. Sagnes, G. Patriarche, Appl. Phys. Lett., 102, 212101 (2013)
9. A. Itawi, K. Pantzas, I. Sagnes, G. Patriarche, A. Talneau, J. Vac. Sci. Technol. B, 32, 0221201 (2014)
10. K. Pantzas, G. Patriarche, E. Le Bourhis, D. Troadec, A. Itawi, G. Beaudoin, I. Sagnes, A. Talneau, Appl. Phys. Lett., 103, 081901 (2013)
11. K. Pantzas, E. Le Bourhis, G. Patriarche, A. Itawi, G. Beaudoin, I. Sagnes, A. Talneau, Eur. J. Appl. Phys., 65, 20702 (2014)
12. K. Pantzas, E. Le Bourhis, G. Patriarche, D. Troadec, G. Beaudoin, A. Itawi, I. Sagnes, A. Talneau, Nanotechnology, 27, 115707 (2016)
13. D. Maugis, Contact, Adhesion and Rupture of Elastic Solids, Springer, Berlin (1997).
14. P. Gillis and J. Gilman, J. Appl. Phys., 35, 647-658 (1964)
Acknowledgments: The authors would like to gratefully acknowledge funding from the CNRS RENATECH network and the Agence Nationale de la Recherche projects COHEDIO and ANTIPODE.
When the education in physics started in the newly founded University of Joensuu in the beginning of the 1970’s, optics was considered as the technology for future to be focused. First research topics were holography and optical materials followed by spectral color research. Similar conclusions were also made in several other universities and research institutes in Finland: leading-edge experimental research in photonics is financially possible even in a small country, like Finland, not requiring millions or tens of millions and above Euros investments in infrastructure. Finland also holds strong position in computing giving valuable support for theoretical and applied research.
Without any historical background in conventional optics, photonics research in Finland was quickly focused on modern optics. In the Institute of Photonics at the University of Eastern Finland (UEF), previously known as the University of Joensuu, research topics today cover 3D printed photonics and free-form optics, carbon nanomaterials, fabrication of micro- and nanostructures for photonics, functional surfaces with femtosecond laser ablation, integrated optics, optical coherence of non-stationary fields, optical measurements of liquids, spectral color research, imaging plants and environment, spectroscopy in materials research, and Surface-Enhanced Raman Spectroscopy (SERS). The institute has about 100 scientists including 12 professors. Multidisciplinary research is conducted in the Department of Physics and Mathematics, the Department of Environmental and Biological Sciences, and School of Computing in close collaboration with the Department of Chemistry. UEF offers photonics education in English in its Master’s degree in photonics and Postgraduate education (PhD) in Photonics programs, which both attract ca. 20 foreign students from all over the world to move to Joensuu every year.
Other focus areas in photonics research in Finland include, e.g., wafer and fiber lasers, and non-linear and applied optics at Optoelectronics Research Center (ORC) and the Department of Physics, respectively, at Tampere University of Technology; optical measurement in bio economy, mining industry, process industry, and environment monitoring at the University of Oulu, micro and nano photonics at Aalto University; laser and infrared spectroscopy at the University of Helsinki; time variable lasers, and optical confocal and near-field microscopy at the University of Jyväskylä; quantum optics, interferometers, and fluorescence based molecules at the University of Turku; optical measurement, and laser welding and machining at Lappeenranta University of Technology; silicon photonics, MEMS technologies, SPR technology, optics instrumentation, bio photonics, lighting research, and printable optoelectronics at VTT Technical Research Centre of Finland, optical metrology at MIKES (nowadays part of VTT); optronics and acoustics at the Finnish Defense Research Agency (FDRA); and photonics based atmosphere and weather observing at the Finnish Meteorological Institute.
Since 2014 photonics industry and photonics-enabled industry in Finland has worked closely with academia under the cluster organization Photonics Finland, headquartered in Joensuu. Photonics Finland has almost 300 private members and 40 corporate members. Photonics offers almost 10 000 jobs in Finland with the photonics related revenue totaling ca. 0.5 B Euros (source: Photonics Finland, 2014). Per capita, Finland has the highest number of photonics companies, that is over 170 (source: EPIC, 2013). However, most of them are small or medium size companies. But also large enterprises, like Metso, Nokia, Kone, Wärtsilä, Valmet, and Vaisala exploit photonics in their main products and services. Key photonics competences in Finland are optical sensing and imaging (machine vision, spectral imaging etc.), micro and nano photonics (R2R, solar cells, 3D printed optics, MOEMS, silicon photonics etc.), and laser and fiber optics (fiber lasers, semiconductor lasers etc.). Entities in Finland will also play important role in the key innovations for future, which are related with optics, such as spectroscopy and spectral imaging (e.g., Specim, Senop, and Spectral Engines), augmented reality (e.g., Nokia, Microsoft Finland, Magic Leap Finland, Nanocomp, and Dispelix); 3D printed optics (e.g., UEF); and nano carbon (e.g., Nokia, Aalto University, and UEF).
European Optical Society (EOS) is an umbrella organization for national optical societies in Europe extending from Portugal to Russia and Norway to Italy. Since 2013 the society has been headquartered in Joensuu. The Joensuu Region has named photonics as one of its spear heads of economic growth besides forest bio economy, and eLearning.
Figure 1. Finnish pavilion at Photonics West 2016 in San Francisco, CA, USA
Professor at the Institute of Photonics, UEF
Vice President at Photonics Finland
Executive Director at EOS
The UCM- Research Group “Physics of Electronic Nanomaterials” (www.finegroup.es) is active since more than 20 years. The current research of the FINE group refers to the relationship between the structural features of electronic nanomaterials and their optical and electronic local properties. Most of the investigated materials are semiconductor nanomaterials, mainly oxides, synthesized by thermal evaporation methods. Nanostructures with a large variety of morphologies, with predominance of elongated structures such as nanowires, nanoneedles or nanobelts, are grown and characterized. The Group has a large experience in application of different micro- and nanocharacterization techniques to the study of structural, optical and electronic properties, including problems related to energy levels, dopant incorporation, electronic recombination, surface effects, crystal defects, growth mechanisms, etc.
A recent example of our research, is the study on “Light guiding and optical resonances in ZnS microstructures doped with Ga or In ” which has been published in Journal of Materials Chemistry C (http://doi.org/10.1039/c5tc02383a) and selected as hot paper by the Editorial Board (2015 Journal of Materials Chemistry C Hot Papers). This research is part of the PhD Thesis of Belén Sotillo recognized with an Excellence award in the PhD program of Physic at the University Complutense in Madrid.
In this work, the high refractive index of ZnS (about 2.4 in the visible range) is exploited to fabricate micro- and nano- resonant cavities. Light guiding behaviour has been studied in several ZnS structures. In particular, in this work the behavior of rods and plates has been investigated and it has been shown that the morphology of the studied structures plays a determinant role in the supported resonant modes. In ZnS:Ga and ZnS:In structures investigated in the present work Fabry–Pérot as well as whispering gallery resonant modes are supported. In figure 1 some of the structures obtained are shown. The leftmost column shows the SEM images of a hexagonal and a triangular plates (in fact a truncated hexagonal plate). The thickness of the plates is around 100nm. In the second column the µ-PL images are shown. The big brilliant spots correspond to the laser incident point. The optical resonances are clearly observed at the plate edges, the patterns are shown in the scheme. The rightmost column shows an example of both light guiding and Fabry-Pérot modes in a sword-like structure.
Figure 1. Indium doped ZnS microcavities with different morphologies support different WGM modes. From left to right columns show SEM images of hexagonal plates, the corresponding µ-PL images, an scheme of the WGM resonaces obtained. The rightmost column shows lightguiding (along the axis) and FP resonances (across) supported by a nanosword.
The effect of different dopants on the refractive index has also been studied. A good agreement has been found between the experimental results and the theoretical model used to describe both types of resonances, which have been used to estimate the refractive index of the doped material (figure 2).
Figure 2. Influence of the dopant (In or Ga) on the refractive index of ZnS.
In ZnS:Ga wires with a triangular cross-section a lasing effect has been observed in the blue-green range. To our knowledge, it is the first time that resonant modes in the visible region have been studied in doped ZnS. can be mentioned.
Figure 3. Lasing effect observed in ZnS:Ga nanorods emitting in the blue-green region of the spectrum.
Prof. Dr. Paloma Fernández
Dept. Física de Materiales, Faculty of Physics
28’40 Madrid, Spain
Authors: Tomi Suhonen1,*, Anssi Laukkanen1, Tom Andersson1, Tatu Pinomaa1, Kenneth Holmberg1
1 VTT Technical Research Centre of Finland, Kemistintie 3, 02044 VTT, Espoo, Finland
Development of new materials and understanding of material and process behaviour is always a complex equation of crossing interactions. Physical and chemical phenomena are affected from the nano- and/or molecular level up to macroscopic level. Interactions between the material processing, structure, properties and performance (PSPP) need to be understood more deeply. For this purpose, modelling skills have developed rapidly in recent decades, with the support of increased numerical calculation capacity and commercial, open source and in-house multi-level and multi-physics software development. In this publication we are presenting some highlights from our current modelling activities obtained wit VTT ProperTune concept related to powder metallurgical (PM) and additively manufactured (AM) materials. We hope they will inspire new ideas on what could be done and obtained via digital approach to design.
Computation-driven design of hard and PM materials
The mesoscalemodelling concept of current work exploits the generally acknowledged PSPP approach in utilising modelling and simulation for material design problems, adapted for PM materials and presented in more detail in Holmberg et al.1,2. At the core are the linkages between material characteristics that can be systematically built and investigated by developing models between the different steps of the PSPP paradigm. The resulting correlations can be studied to quantitatively grasp the significance of nano-microstructural material features and physical phenomena giving rise to material properties and product performances. These approaches enable the effective exploitation of computation and multiscale materials modelling in design of composite materials. The implementation of these means in practice falls within the realm of integrated computational materials engineering (ICME), where computation merges with experimental, characterisation and material informatics to introduce a general and practically exploitable concept for simulation-driven material design. For hard and PM materials in the implementation of PSPP we have identified three common steps, or specific problems typically tackled and found to be of interest, primarily due to the goal of exploiting modelling as a part of a performance-driven material design chain. These can be viewed as common use case problems on how to exploit multiscale modelling e.g. in design of PM materials, material selection or solving specific material-related problems. First, in Fig. 1, the structure–property problem is presented and case examples briefly outlined for various hard material systems and affiliated microstructures. In a structure–property problem the modelling problem is centred around the digital representation of material nano-microstructure and phenomena responsible for basic material properties, such as strength, viscoplastic strain rate, hardness, etc. As such, the key element is in possessing the capability to generate realistic enough representation of material structure, such as reinforcing constituents and defect structure, along with e.g. the underlying metallic microstructure.
Figure 1.Modelling in solving the structure to property problem for PM materials
The most common problem types are affiliated with deformation response and following properties such as compressive strength, deciphering what features in material structure are critical for strengthening of the system and how to systematically work towards desired material properties. As such, the computation takes the form of simulating and carrying out virtual testing of common laboratory experiments such as hardness, compression and scratch testing, and the models are applied to investigate the correlation between structural morphologies, mechanisms of deformation and the resulting material properties. The structure–property–performance problem can be viewed as the next step in exploiting computation for design of hard and PM materials, and such a concept and case example is presented in Fig. 2. Performance as a term is typically affiliated with more elaborate properties than those related to material properties, such as resistance to fatigue, wear or fracture toughness (see e.g. Holmberg et al.1,2 for a more thorough discussion and further references). Subsequently, performance by the convention adopted by current authors pertains to behaviour also including component design, the use of the material in its operating conditions and environments. As such, the structure–property–performance can also be identified as an extension of the structure–property problem to more generic conditions. In Fig. 2, a case example of fatigue evaluation of selective laser melting precipitation hardened microstructure of steel is given. The differences to make note of in comparison to the structure–property problem are that material defect structure, criteria for failure micromechanisms or explicit modelling of failure processes and the conditions under which the product operates are introduced in modelling the problem. In Fig. 2, the analysis proceeds by merging a microstructural analysis of how a defect structure responds under deformation during fatigue cycling, and the subsequent stress–strain response is utilised to carry out a microstructure informed fatigue analysis. The merits of such an analysis lie in the fact that individual microstructural defects can be quantitatively linked to fatigue initiation life, enabling one to for example evaluate the significance of specific microstructural features on fatigue life, or whether certain specific defect types are at all of relevance for operational life of a component.
Figure 2. Modelling in solving the structure, property to performance problem for PM materials
The links to e.g. certification and developing PM parts for extreme performance environments are imminent in addition to the general notions presented for performance-driven material design. As the third and last use case type for modelling-driven material design, the PSPP analysis type for PM materials is presented in Fig. 3. The use case can be viewed from the perspective of the structure–property–performance problem by introducing simulation of the material processing and manufacturing step, i.e. the analysis can now treat for example solidification and sintering structures directly, rather than use characterisation or like information in generating the computer-generated representation of material structure. The analysis complexity increases, but the potential impact does so as well. For example, in Fig. 3, the case depicts a PSPP analysis for additive manufacturing (AM). The process model comprises of a thermomechanical FE solver, which in current case is an example run on eight different process parameter sets to yield differing thermal histories.
Figure 3. Modelling in solving the process, structure and property to performance problem for PM materials
The output of such a thermomechanical analysis can be used in a PF analysis for alloy solidification, producing the resulting nano-microstructure. The structure– property–performance problem henceforth is carried out similarly as in other analysis scenarios. The merit of the PSPP analysis chain is in enabling the linking of material processing parameters and variables all the way to component performance. This enables the systematic investigation of causal relations accounting for all critical stages from material and component manufacturing to performance of the product, and these relations can then be systematically exploited in design of products with the required performances and also in evaluating the cost-effective means of delivering those performances and developing improved materials.
1. K. Holmberg, A. Laukkanen, A. Ghabchi, M. Rombouts, E. Turunen, R. Waudby, T. Suhonen, K. Valtonen and E. Sarlin: ‘Computational modeling based wear resistance analysis of thick composite coatings’, Tribol. Int., 2014, 72, 13–30.
2. K. Holmberg, A. Laukkanen, E. Turunen and T. Laitinen: ‘Wear resistance optimization of composite coatings by computational microstructural modeling’, Surf. Coat. Tech., 2014, 247, 1–13.
The radial braiding technology enables to manufacture near net-shaped textile preforms. Because of their crimped textile architecture and the evenly distributed load adsorption, braided reinforced fiber composite materials are characterized by an outstanding failure behavior . These features make braiding to one of the most promising technologies for the production of fiber-reinforced hollow profiles such as braided A-pillars which already get produced in large series .
Manufacturing simulation allows faster prototyping, by avoiding long and expensive experiments. This is the reason why the Institut für Textiltechnik is taking efforts to close the digital simulation process chain (see Fig. 1). The digital process simulation chain extends from the simulation of the braiding process, over the homogenization and validation till to the simulation of entire components.
Fig. 1. Digital process simulation chain
The large number of elements and contacts in the process simulation of the braiding leads to very long and not industrial-suited computing times. Even with newest processors several days are usually needed for calculation. Under consideration of these facts the current research aim is to develop a computationally efficient braiding simulation in ABAQUS Explicit which contains the mapping of complex states of friction, over complex mandrel geometries.
Since a composite consists of matrix and rovings that consists out of filaments, there are several levels of observation and simulation. The individual levels (see Fig. 2) differ in terms of the model scale and the consideration of the physical phenomena. The most detailed level is the micro model level, here the individual filaments are considered. However, because of the order of several thousand filaments and their contacts, it is not possible to achieve industrial-suited simulation times. Therefore, the modeling takes place on the meso modeling level, so that the interaction between the individual filaments and also the change in the roving cross section cannot be considered and the roving is represented as a single filament. The next level is the macro modeling level, here textile reinforcement structures are considered and these models treat surface elements and their material behavior which is described on material laws. Another possibility to model the braiding simulation is the analytical modeling. This approach focuses only on the kinematic movement of the yarns and not on the kinetic forces which cause them.
Fig. 2. Simulation time and accuracy relation of existing models
In order to reduce the simulation time of meso scale models till the calculation times of macro models or even analytical models an iterative optimization methodology of model reduction, model property improvement and multiplication is used. On the basis of a reduced two-thread model a combination morphological box and the design of experiments is used to identify the most significant factors influencing the processing time. The model is divided into sub models such as roving, bobbin, contact, braiding ring and mandrel. For every sub model factors of influence, for example the mass scaling or the element size, are manipulated with respect to computing time, realistic thread and contact behavior.
Fig. 3. Development of two-thread model (a) for multi-thread model (b)
Once the improvements have been performed, a multi-thread model is developed with the by multiplying the two-thread model. The simulation results are analyzed in terms of their calculation time, their thread behavior and contact behavior. Simultaneously a decrease in the accuracy of the simulation has to be prevented.
By the mentioned approach it is possible to identify the significant factors influencing the computing time. Those factors can be optimized by reaching a minimum in the calculation time and keeping simultaneously a realistic fibre behavior. As a result, even complex geometries, such as pressure vessels with complex friction definition can be simulated in minutes with high prediction accuracy. Overall a decrease of computing time from several days down to minutes could be demonstrated.
1) Hufenbach W., Gude M., Czulak A., Gasior P., Kretschmann M.; „Manufacturing and pressure tests of braided vessels with integrated optical fiber sensors”;Diagnostyka Materiaów Polimerowych 2011
2) Gardiner G.; „Wet compression molding“, Blog Composites World, http://www.compositesworld.com/blog/post/wet-compression-molding, 02.01.2016
M. Kolloch, M. Lengersdorf, T. Gries
Institut für Textiltechnik (ITA) of RWTH Aachen University
Marc Bernacki1, Benjamin Scholtes1,2, Amico Settefrati2, Nathalie Bozzolo1, C. Moussa1, D. Pino Muñoz1, Y. Zhan1,3, Emmanuel Rigal3, Christian Dumont4, Remy Besnard5, Isabelle Poitrault6, Joëlle Demurger7, Aurore Montouchet8, Isabelle Bobin8, Jean-Michel Franchet9
1Mines ParisTech, PSL Research University, CEMEF, Sophia Antipolis, France
2Transvalor S.A., 694 avenue Maurice Donat, Mougins, France
3CEA Liten, Grenoble, France
4Aubert & Duval, Les Ancizes-Comps, France
5CEA Valduc, Is-Sur-Tille, France
6Industeel, ArcelorMittal, Le Creusot, France
7Ascometal, CREAS, Hagondange, France
8Areva Creusot Forge, Le Creusot, France
9Safran, SafranTech, Magny-Les-Hameaux, France
Metal forming modeling can be predictive only if the strain rate, strain and temperature dependency of the flow behaviour are correctly described. The mechanical properties and behaviour of metallic materials mainly depends on the content and structure of dislocation network, this points out the need to incorporate microstructure concepts into the numerical models. The goal is to correctly describe the main physical mechanisms occurring in metals during thermomechanical processes i.e. work-hardening, recovery, grain boundary migration, nucleation and grain growth related to dynamic, static or metadynamic recrystallization. Macroscopic and homogenized models are widely used in the industry, mainly due to their low computational cost [1, 2]. If this mean field framework is quite convenient, it can be synonymous, for a given material, with a large amount of experiments with advanced laboratory devices. Moreover, the homogenization of the microstructure does not permit to capture some very local phenomena.
Over the last decades, lower-scale models (called full field models) have been developed in order to simulate explicitly the microstructural evolution [3-5]. The idea behind these mesoscale simulations is that the morphology and the topology of the grain boundary network play a non-negligible role in the evolution of the microstructure. Recently a new full field approach, based on a Level Set (LS) description of the interfaces in a finite element (FE) context has been introduced to model 2D and 3D primary recrystallization, including the nucleation stage, and has been extended to take into account the grain growth stage [6-8]. Moreover in this LS context, Smith-Zener pinning (SZP) phenomenon can be taken into account in a natural way . These full field approaches are generally associated with an elevated computational cost making them hardly usable for 3D computations. Moreover they require many numerical parameters whose calibration is not straightforward. Recent major developments and improvements addressed these issues [10-12] making possible the use of these approaches in an industrial context.
Numerical model description
Simulations are performed on a Representative Volume Element (RVE) at the mesoscopic scale where the microstructural features are explicitly represented. The polycrystal is constructed in a statistical way by respecting the topological characteristics of the grains and the metallurgical properties. Efficient algorithms have been developed to respect a given grain size distribution and attributes [13,8]. The microstructural evolution is given by the displacement of interfaces (grain boundaries for example). The model considered here works around a LS description of the interfaces in a FE framework. This front capturing approach has the advantage of avoiding the difficult problem of tracking interfaces and allows to naturally handle complex topological events occurring during grain boundary motion.
Theoretically, each grain must be represented by its own LS function. In practice, non-neighbouring grains in the initial microstructure (separated by a certain number of grains) can be gathered to form global LS (GLS) functions. This approach allows using a small number of functions Np compared to the total number of grains constituting the microstructure Ng and thus limiting the numerical cost. The initial separation between grains belonging to the same GLS function must be chosen small enough to limit the computation time and sufficiently high to avoid a numerical coalescence. To address this issue, an efficient grain recoloring algorithm has been recently developed [11,12]. The interface of each GLS function is then displaced by solving a set of convective-diffusive equations  and a reinitialization procedure in order to keep the metric properties of the distance function. An efficient and parallel reinitialization algorithm based on a direct approach and using optimized searching procedures has been recently developed  leading to significant computational cost reductions in comparison to the classical reinitialization approach used in [7-9] consisting in solving an Hamilton-Jacobi equation for each GLS function.
As illustrated in Figure 1, the model works in 2D and in 3D. Realistic predictions necessitate a sharp description of the interfaces. This issue is achieved in 2D thanks to an anisotropic mesh adaptation around the interfaces allowing to reach this objective while conserving a high numerical efficiency. As the interface moves, periodic remeshing is performed such that the refinement zone always coincides with the interface position. This technique allows improving precision and reducing computation times. If this meshing strategy is also usable in 3D (see Figure 1a), current developments concern the improvement of the associated remeshing numerical cost. Boundary conditions applied to the RVE are representative of what experiences a material point at the macroscopic scale (thermomechanical cycle of the point considered).
Figure 1. (a) 3D description of a polycrystal and corresponding finite element mesh (anisotropic meshing near grain boundaries), (b) Initial 304L 2D polycrystal (50000 grains) and (c) microstructural evolution of this microstructure during an isothermal heat treatment (1050°C): comparison with the Burke and Turnbull model [11, 13].
3D large-scale representative simulations
Figure 2 represents, in context of static recrystallization, the microstructural evolution of a 304L austenitic stainless steel subjected to an isothermal heat treatment at 1050°C after deformation. The microstructural evolution is driven by the reduction of the total grain boundaries length/area and of the stored energy.
Figure 2. 3D microstructural evolution of a 304L steel during static recrystallization . The color codes correspond to the stored energy (on the left column) and to the grain size (on the right column).
In order to limit the final grain size which can be detrimental for the mechanical properties, a classical method consists in precipitating second phase particles (SPP) that can hinder the grain boundaries motion. If this approach is usually efficient, under specific conditions abnormal grain growth (AGG) may occur. AGG can be described as the selective growth of only a few grains while other grains do not grow in the microstructure. It may occur as a result of a heterogeneous stored energy field leading to a driving force for some grain boundaries that overcomes the pinning force, or as a result of grain boundary energy anisotropy, or as a result of a heterogeneous SPP distribution. Mean field models cannot predict such a local phenomenon (because of the microstructure homogenization); therefore the development of efficient and accurate 3D modeling tools able to account for the pinning phenomenon is necessary. Based on the work described in , the recent numerical developments [10-12] allowed performing 3D simulations as shown in Figure 3 . SPPs are considered inert and are represented as holes in the FE mesh (green surfaces in Figure 3). In such a way, incoherent or coherent particle/grain interfaces can be considered through appropriate boundary conditions and the dragging effect is naturally modeled by the modification of the local curvature when the grain boundary passes through the particles.
Figure 3. A 3D grain growth simulation for Inconelä 718. The green spheres represent the precipitates.
Finally, the last figure (Fig. 4) illustrates the use of these new numerical tools in order to model the evolution of the microstructure during the last steps of a HIP-bonding of 316L austenitic stainless steel . Hot Isostatic Pressing-diffusion bonding of grooved plates is a potentially attractive technique for manufacturing compact heat exchangers. During the process, microstructural changes must be controlled and groove deformation must be minimised. First, plastic deformation processes result in collapse of the surface asperities, the interfacial void closure is then achieved by diffusion controlled mechanisms such as creep and grain boundary diffusion, bonding is finally achieved by grain growth. The process parameters must be accurately calibrated in order to ensure the disappearance of the initial diffusion welding plane while controling the final grain size distribution. Modeling at the microscale is then a precious tool to reach these goals. Figure 4 illustrates the 3D modeling of the last step of a HIP-bonding process (grain growth phenomenon) for 316L thanks to the LS approach, the color code corresponds to the grain size and more than 20,000 grains are considered in the simulation.
Figure 4. 3D modeling of the last step of a HIP-bonding process (grain growth phenomenon) for 316L thanks to the LS approach, the color code corresponds to the grain size (mm). In the left side, the white surface corresponds to the grain boundary network. On the right side, a cutting view of the grain boundary network is described with the bonding plane (defined by the middle Z-plane). More than 20,000 initial grains are considered in the simulation. (Top) Representative microstructure at the beginning of the grain growth mechanism (where T-junctions are then present in the bonding plane). (Bottom) Evolution of the microstructure after 600s at 1100°C.
A full field approach using the LS method in a FE context has been developed to simulate the microstructural evolution during forming processes. Modeling at the mesoscopic scale can give insight into the understanding of complex microstructural phenomena but it can also be used to optimize/calibrate higher scale models (like mean field models). These simulations allow describing in a natural way the materials in terms of microstructural features. The recent improvements done to reduce high computation times generally associated with these models make possible now their use for industrial applications.
The objective of the industrial partners of these developments is to provide the best possible quality for the final product of their customers within the shortest delivery time. With that aim, the development of numerical models able to predict microstructure evolution, at the mesoscopic scale is nowadays of prime importance. These digital models will allow being very reactive to new markets with high enough confidence in the proposed manufacturing sequences and parameters. The final aim is the capitalization of this knowledge and of these tools in a multiscale software package framework usable in an industrial context. This is exactly what it is proposed through the DIGIMUâ software package dedicated to the industrialization of these developments . Furthermore, scientific perspectives in terms of understanding and modeling of metallurgical phenomena and development of the LS methodology are then as ambitious as the industrial perspectives in terms of applications. Figure 5 illustrates the consortium involved in the development of the DIGIMUâ software package.
Figure 5. Consortium involved in the development of the DIGIMUâ software package.
1. P. Bernard, S. Bag, K. Huang, R.E. Logé, Mat. Sci. Eng. A 528 (2011) 7357-7367.
2. M. Mukherjee, U. Prahl, W. Bleck, Steel Res. Int. 81 (2010) 1102-1116.
3. A.D. Rollett, D. Raabe, Comp. Mater. Sci. 21 (2001) 69-78.
4. K. Piekos, J. Tarasiuk, K. Wierzbanowski, B. Bacroix, Recrystallization and Grain Growth III Proc. (Jeju Island, Korea) (2007) 1157-1162.
5. L.Q. Chen, Ann. Rev. Mater. Res. 32 (2002) 113-140.
6. M. Bernacki, H. Resk, T. Coupez, R.E. Logé, Mod. Sim. Mat. Sci. Eng. 17 (2009).
7. M. Bernacki, R.E. Logé, T. Coupez, Scr. Mat. 64 (2011) 525-528.
8. A.L. Cruz-Fabiano, R. Logé, M. Bernacki. Comp. Mat. Sci. 92 (2014) 305–312.
9. A. Agnoli, N. Bozzolo, R.E. Logé, J.-M. Franchet, J. Laigo, M. Bernacki, Comp. Mat. Sci. 89 (2014) 233-241.
10. M. Shakoor, B. Scholtes, P.-O. Bouchard, M. Bernacki, App. Math. Mod. 39 (2015) 7291-7302.
11. B. Scholtes, M. Shakoor, A. Settefrati, P.-O. Bouchard, N. Bozzolo, M. Bernacki, Comp. Mat. Sci. 109 (2015) 388-398.
12. B. Scholtes, R. Boulais-Sinou, A. Settefrati, D. Pino Muñoz, I. Poitrault, A. Montouchet, N. Bozzolo, and M. Bernacki. 3D level set modeling of static recrystallization considering stored energy fields. Comp. Mat. Sci., 2016, In press.
13. K. Hitti, P. Laure, T. Coupez, L. Silva, M. Bernacki, Comp. Mat. Sci., 61 (2012) 224–238.
14. J.E. Burke, D. Turnbull, Prog. in Met. Phys., 3 (1952) 220-292.
15. B. Scholtes, M. Shakoor, N. Bozzolo, P.-O. Bouchard, A. Settefrati, and M. Bernacki. Advances in level-set modeling of recrystallization at the polycrystal scale - development of the digi-μ software. Key Eng. Mat., 651–653 (2015) 617–623.
16. E. Rigal, N. Bouquet, M. Bernacki, and F. Bernard. Etablissement et évolution des interfaces lors du soudage diffusion. Journées annuelles de la SF2M, Matériaux et conversion d’énergie, 2015.