Prediction of material failure

• Antragsteller: Professor Dr. Karl Maier, seit 7/2009
• Fachliche Zuordnung: Mechanische Eigenschaften von metallischen Werkstoffen und ihre mikrostrukturellen Ursachen ; Statistische Physik, Nichtlineare Dynamik, Komplexe Systeme, Weiche und fluide Materie, Biologische Physik
• Förderung: Förderung von 2008 bis 2011
• Projektkennung: Deutsche Forschungsgemeinschaft (DFG) - Projektnummer 64100804

Failure of construction parts due to fatigue is a phenomenon well known to the public. After severe accidents of airplanes, trains and cars, it is widely discussed. From our daily experience we know that in repeatedly loaded metals, alloys and polymers the lifetime is finite.Since almost 150 years the lifetime of construction parts is determined employing destructive test series like the Wöhler-test, where a huge number of identical samples has to be tested in a very time-consuming way. Up to the present these methods have not changed in principle. The general goal of this project is the development of a new reliable method for the prediction of the remaining useful lifetime of mechanically loaded components. On a microscopic scale material fatigue is based on accumulation of dislocations and other defects in the lattice. Already in the early stages of fatigue – within the first couple of load cycles – a significant increase of the defect density can be observed nondestructively by Positron Annihilation Spectroscopy (PAS). The transition from the state of normality (fatigued but stable properties) to failure is directly correlated to the local defect density, which is experimentally accessible by the S-parameter. Assuming that failure occurs, when the defect density exceeds a critical value locally, it can be employed as a precursor for failure. Our goal is a robust and broadly applicable method for an extrapolation of failure from an early state of fatigue.In some special cases the experimental results obtained with PAS are already sufficient for a prediction of fatigue failure. But a general technique of failure prediction is far from trivial and will only be possible by a combination of experimental methods and a model describing the accumulation of defects during fatigue. With such a model we do not intend a detailed description of dislocations on the microscopic scale. Instead, the existing knowledge of plasticity theory should be implemented as simple as possible but accurate enough to describe the evolution of the defect density during a huge number of fatigue cycles. As a first approach a model based on a cellular automaton was developed in a diploma thesis. In the present state the model could already reproduce some aspects of fatigue: the accumulation of defects and the local transition from elastic deformation to plastic flow. But in the present state idealized simple material properties and transition functions are implemented. For a sufficient modeling of fatigue the model has to improved essentially and validated by experimental data.

• DFG-Verfahren: Sachbeihilfen
• Beteiligte Person: Professor Dr. Gunter Markus Schütz
• Ehemaliger Antragsteller: Dr. Matz Haaks, bis 7/2009