Manufacturing Effects & Fatigue in Aerospace

Engineering Integrity Society

Tuesday, 16 October 2007

Messier Dowty, Gloucester

ABSTRACTS

"Characterisation of corrosion fatigue in steam turbine blade steels"
Dr Karen Perkins, Swansea University

Corrosion damage and its effect on subsequent fatigue performance will be illustrated for two common engineering materials – a high chromium steel employed for power generation applications and an advanced aluminium airframe alloy. In each case, the controlling influence of microstructure and materials processing routes on surface damage will be highlighted. Laboratory based fatigue experiments will be described that simulate the typical in-service environments experienced by these alloys and support LEFM based lifing predictions of cyclic behaviour, either subsequent to or coincident with corrosion damage

"Foreign object damage & subsequent fatigue response of titanium based areofoil alloys"
Prof. Martin Bache, Swansea University

Laboratory based experiments are described that introduce foreign object damage (FOD) into titanium alloy specimens which are subsequently assessed for fatigue response. Ballistic particle impacts of varying energies and at various locations (i.e. full face or edge) are characterised for micro-plasticity and localised cracking. Compared to pristine specimens, the fatigue strength of damaged specimens is significantly reduced. Analytical techniques to account for the influence of FOD on fatigue strength are explored.

"Accounting for doubt & uncertainty in materials fatigue test data"
Dr Peter Blackmore, Jaguar Landrover

It is widely known that fatigue is a stochastic process. Most of the time, though, engineers choose to ignore this fact and treat the phenomenon, in CAE models, for example, as being deterministic. Additionally, most of the laboratory tests used to define basic material fatigue properties use carefully machined and polished specimens that are rarely, if ever, representative of the surface condition of real-world components. By not properly understanding or accounting for the effects of statistical variations in material properties or incorporating the effects of as-manufactured surfaces serious errors can accrue in fatigue calculations.
Numerous methods to account for these effects have been proposed over the years; some more successful than others. This presentation will describe work at Jaguar-Land Rover that has resulted in a standard method for the determination of the strain-life fatigue properties of metallic materials that includes the determination of statistical lower bounds and simultaneously incorporates the effects of retained as-manufactured surface finish. This approach has enabled the development of CAE models with improved correlation with physical tests. This, in turn, enables virtual sign-off and significant reductions in product development times and costs.