Fatigue in material science is the structural damage of a material subjected to cyclic loading. Fatigue life is influenced by several factors in complicated ways.

We introduce a mathematical data-mining model using measured static material parameters and a database of known materials to predict unknown cyclic parameters describing the material fatigue behavior of new materials. This model supports the development of new materials with desired cyclic durability properties. It can be used to reduce the number of expensive practical cyclic load experiments.

Partners

• TU Darmstadt
• UC Berkeley

Funding

• DFG, Collaborative Research Center SFB-666.

Related talks

• A Data-Mining Linear Programming Model to Predict Material Fatigue Parameters, Seminar Nichtlineare Optimierung und Inverse Probleme, WIAS, Berlin, Germany, 2.11.2010.
• Mathematische Optimierung bei der Entwicklung spaltwalzfähiger Produkte im DFG Sonderforschungsbereich 666, 3rd Symposium of the Arbeitskreises Metallindustrie und Mathematik (MetMat#3), Bad Honnef, Germany, 18.11.2008.
• A Data-Mining Linear Programming Model to Predict Cyclic Metal Fatigue Parameters, IFORS 2008, Sandton, South Africa, 15.7.2008.

Related publications

• Chalid el Dsoki, Armin Fügenschuh, Holger Hanselka, Dorit Hochbaum, Irma Hernandez-Magallanes, Erick Moreno-Centeno, Andrea Peter, Das ANSLC-Programm und das SDM im Vergleich, Peter Groche (Ed.), Sonderforschungsbereich 666: Integrale Blechbauweisen höherer Verzweigungsordnung – Entwicklung, Fertigung, Bewertung, Meisenbach Verlag, Bamberg, pp. 97 – 106, 2008.