Subproject C5 Crack growth

Life time simulations of complex goods are based on phenomenological life time models that cannot entirely capture the problem-specific characteristics. To be able to predict the remaining life time of regenerated goods, more precisely methods are required that can take into account the properties of the engineering part under investigation.


Deformation in z-direction of a turbine blade

In sub-project C5, a quasi-static, multi-scale method for the simulation of cracks based on the XFEM has been developed that can take into account thermal loadings, heat conduction, non-local damage, as well as possible crack face contact to simulate regenerated parts under realistic circumstances. However, the quasi static model can only estimate the remaining life time due to an inaccurate reflection of the complex dynamic interactions. It is therefore necessary to extend the model to include inertia terms. To simulate dynamic crack growths for a large number of load cycles efficiently and accurately within reasonable computing times, the multi-timescale method WATMUS will be applied.


The developed thermal-mechanical multiscale projection method can be used to investigate the interaction between micro- and macrocracks in critical parts of the domain. Heat transfer through crack surfaces has been considered by employing a contact formulation. Non-local damage has been applied as a suitable criterion to determine whether crack propagation will occur, as well as the propagating direction and propagation length. These results can be directly used to develop a three-dimensional dynamic crack growth model, which will be applied to perform simulations of multiscale dynamic crack propagations of turbine blades and blisks.

Deformation field of a turbine blade for a multiscale calculation


The existing XFEM implementation is extented by inertia terms at one scale. After that, the equation of motion is cupled to transient heat conduction as well as to non-local damage. The coupled equations will be solved explicitly and simultaneously for each time step. Eventually, this implementation will be coupled to the multiscale projection method.

The temperature distribution on the surface of turbine blades/blisks from CFD simulations of sub-project C3 can be applied as boundary conditions. The maximum vibration amplitude and material parameters needed for the XFEM-analysis are taken from sub-project C3.

The transient simulation of the multiscale crack propagation under dynamic excitation allows for a mor precise prediction of the remaining life time of each specific engineering part, without the necessity of empirically gathered data. Thus, within the CRC a more precise evaluation of the regeneration paths is possible.

The developed methods can be readily employed for other capital goods in the industry, such as the thermal-mechanically highly loaded drives of diesel locomotive.


  • Holl, M.; Rogge, T.; Loehnert, S.; Wriggers, P.; Rolfes, R. (2014) 3D multiscale crack propagation using the XFEM applied to a gas turbine bladeComput Mech 53 (1), S. 173–188
    DOI: 10.1007/s00466-013-0900-5
  • Loehnert, S. (2014) A stabilization technique for the regularization of nearly singular extended finite elementsComput Mech 54 (2), S. 523–533
    DOI: 10.1007/s00466-014-1003-7
  • Holl, M.; Loehnert, S.; Wriggers, P. (2013) An adaptive multiscale method for crack propagation and crack coalescenceInt. J. Numer. Meth. Engng 93 (1), S. 23–51
    DOI: 10.1002/nme.4373
All publications of the Collaborative Research Centre  


Dr.-Ing. Stefan Löhnert
Prof. Dr.-Ing. habil. Dr. h.c. mult. Dr.-Ing. E. h. Peter Wriggers
Appelstraße 11 und 11a
30167 Hannover
Appelstraße 11 und 11a
30167 Hannover