Skip to main content
European Commission logo print header

Enhanced reliability and lifetime of ceramic components through multiscale modelling of degradation and damage

Final Report Summary - ROLICER (Enhanced reliability and lifetime of ceramic components through multiscale modelling of degradation and damage)

Executive Summary:
The macroscopic mechanical properties of silicon nitride depend on its microstructure. As it is possible to prepare Si3N4 with various microstructures, so it is also possible to obtain materials with different mechanical and/or damage properties. The complex relation between the microstructural properties and the macroscopic thermoelastic, fracture, and wear properties is neither well known nor well understood. This challenge was tackled in RoLiCer by employing computer simulation following a multi-scale paradigm, in which the simulation of the properties of silicon nitride is broken down in simulations on three different interconnected levels: the atomic scale, the micro and meso-scale, and the macro-scale. The simulations were backed-up by targeted model experiments.
The results of RoliCer enabled constructing a robust subroutine that models mechanical as well as thermomechanical wear in ceramic components by means of finite element method. Moreover, lifetimes predictions of ceramics were accomplished as fatigue life predictions.
Using cohesive zone modelling, crack nucleation and propagation for short crack in the range 10-100 µm was predicted. Fracture behaviour was successfully modelled for different silicon nitride grades by optimising traction separation parameters. Additionally, XFEM method was successfully applied and provided good results when bridging effects were not strong (i.e. for longer cracks).

The 3D microstructure of silicon nitride was modelled using EBSD as input for numerical models. For instance, the effective thermoelastic properties of silicon nitride were computed based on mean field and full field homogenization techniques. The fracture behaviour of silicon nitride was studied using finite element analysis and it was shown that damage initiates across the intergranular interface, thus, preceding transgranular fracture. Moreover, two macroscale damage models for silicon nitride were implemented in RoliCer; an analysis has been made under static and cyclic loadings.
The atomistic simulations in RoliCer focused on describing the properties of the grains, grain interfaces, and amorphous bulk and IGF materials. Ab initio simulations and molecular dynamics simulations were conducted. Ab initio simulations were used to study the cohesive strength of Si3N4 grains with and without defects and impurities. The results were further used to assess the accuracy and reliability of the empirical potentials. Furthermore, the adhesion between grains was studied by means of IGFs.
Two silicon nitride grades were tailored for bearing and wire rolling applications. The materials were fully characterised and compared to state-of-the-art silicon nitride grades.
In the prototype testing and demonstration of functionality phase of the project, ceramic rolls were used in hot wire-rolling experiments. The rolls were successfully used to roll approximately 150,000 kg of different steel qualities including an unprecedented quantity of nickel-base superalloys wires.

Project Context and Objectives:
The objective of RoliCer is to boost the reliability and longevity of advanced ceramics by using extensive data from engineering applications, numerical simulations and targeted model experiments.
Studies on advanced ceramics have highlighted their exceptional superior mechanical, thermal and tribological properties for many engineering applications. Ceramics are especially well suited for modern rolling and sliding bearings, as well as for metalworking and cutting tools.
However, there are still concerns over the reliability and lifetime of ceramic materials. RoliCer as an EU-funded collaboration of companies that develop and manufacture ceramic materials and components and research institutes that conduct research by means of numerical simulations and experiments to assess materials on the different scales – works to bridge the gap in knowledge between damage, wear, reliability and lifetime of ceramics. The project investigates the damage and degradation of silicon nitride (Si3N4) by means of multiscale numerical simulations (atomistic, microscale, mesoscale and macroscale) as well as targeted model experiments. The outcomes of the project are validated through industrial testing.
The macroscopic mechanical properties of silicon nitride (Si3N4) depend on its microstructure: on the size and orientation of the Si3N4 grains, on the chemical composition of the glassy phase and on the inter-granular film (IGF) glassy phase which mediates the interaction and structure between nearby grains. As it is possible to prepare Si3N4 with different microstructures, so it is also possible to obtain Si3N4 materials with different mechanical or damage resistance properties. In principle, this enables the engineering of Si3N4 with specific properties by tailoring its microstructure. Unfortunately, the complex relation between the microstructural properties and the macroscopic thermoelastic, fracture, and wear properties is neither well known nor well understood. Knowing this relation would be of great practical importance in the development of Si3N4 based components for many advanced applications, including aerospace, energy production industries, metal rolling industries and much more.
This challenge was tackled in the RoLiCer project employing computer simulation following a multi-scale paradigm in which the simulation of Si3N4 is broken down in simulations on three different interconnected levels: on the micro-scale, the meso-scale, and the macro-scale.

One of the most challenging aspects of the methodology is that evolving from small-scale models to large-scale models involves several changes, including those related to time. There are many aspects involved in transferring information from the over-simplified, small-scale model of a very complex phenomenon into the larger-scale model and vice versa. For instance, calculations and simulations at the atomic level provide insight into phenomena and may explain what happens on larger scales. A clear example is fracture: the occurrence of fracture on the atomistic level shows the orientations and planes at which fracture is most likely to occur on larger scales. On the other hand the macro-scale fracture mechanics rely on concepts such as the stress intensity factor, which is an indicator of the stress state at the crack tip in ceramic materials. It is used to predict fracture, and hence failure, in ceramics. Stress gradients within the material may influence the stress intensity factor, directing towards certain design or loading recommendations.
Since ceramics are applied in various fields, no one material combination can be expected to portray a magic recipe. RoliCer is targeting two groups of applications: bearings and manufacturing tools. With bearings, the ceramic rolling components are subjected to severe mechanical contact stresses. With manufacturing tools, the ceramic tools undergo high mechanical and tribological stresses at high temperatures. The project aims at designing two distinct materials capable of sustaining the loading conditions pertaining to each type of application. This being said, both material combinations must be economically feasible for production as well.
Parallel to the numerical simulations, the research in RoliCer focuses on both material design and implementation in the industry. The testing phase will involve using material grades developed within the project to manufacture rolling components that will be examined in full-scale hybrid bearings tests and rolls that will be tested in a nickel-base superalloy wire rolling mill.

Project Results:
See attachement

Potential Impact:
Apart from the important research findings reached in RoliCer, the project resulted in developing two silicon nitride ceramic grades to satisfy the requirements of two completely different sets of loading regimes; by itself this is considered as a commercial advancement within the portfolio of a European ceramic manufacturer. The industrial testing of silicon nitride rolls in hot wire-rolling of superalloys confirmed that ceramics are a promising class of materials for metalworking applications. Further improvements and industrial tests succeeding the project have been committed by the project partners.
The work conducted by the RoliCer team is on track to be translated for application in industrial and technical settings. The outcomes of RoliCer will help to broaden the range of technical applications significantly, because there will be opportunities for fast and efficient design of robust high performance systems using ceramic components. In addition to providing new insights into ceramics, the RoliCer project has also provoked new questions that will guide future research.
From the dissemination activities within RoliCer, many students and young researchers within the European Union have had the opportunity to get into contact with professionals within the field of engineering and materials science through the organization of several workshops and many seminars and through participation in numerous conferences, including having a RoliCer-Symposium within the proceedings of the fourth MSE “Materials Science and Engineering” Congress, which took place between the 23rd and the 25th of September 2014 in Darmstadt, Germany. Additionally, the involvement of many young researchers within the R&D activities throughout the project has helped them to broaden their horizon and deepen their knowledge within their respective fields. In addition to several small research topics and master theses, six PhD candidates were involved in the project. As regards the consortium of RoliCer, the project has enhanced and strengthened the networking, share of knowledge and know-how amongst professionals within the European Union and amongst research institutes, higher education establishments and companies.