Residual-stress measurements in laser-clad-repaired low-pressure turbine blades for the power industry

Phil Bendeich, David Carr, Ken Short, Richard Blevins, Caroline Curfs, Oliver Kirstein and Gerard Atkinson (ANSTO), Tom Holden (Northern Stress Technologies), and Ron Rogge (NRC-Chalk River)

 
Left:  Trial turbine blade cladding at TRUenergy's plant at Torrens Island, South Australia;  Right: Brian Dempster, Jim Harris and Milan Brandt (Swinburne University)

Steam turbines in thermal power stations are a crucial piece of infrastructure essential to modern life. At the turbine-blade tip, the steam is traveling at supersonic speeds and metal erosion occurs. Replacement of a complete set of blades costs is in the multi-million-dollar range.

We have been using neutron strain scanning, in conjunction with the Cooperative Research Centre for Welded Structures, CSIRO and Swinburne University, to study a new in-situ repair method: laser cladding. A thin layer of Stellite® is welded onto the blade with a robot-controlled diode laser.

We have studied both small coupons of Stellite® welded onto stainless steel (below left) and a real blade with a repair weld on its leading edge (below right). Neutron diffraction was used to study the spatial distribution of stresses below the Stellite® in both cases, and the experiments were performed at the NRU reactor at Chalk River, Canada, and at our own HIFAR reactor.

In the case of the blade (see below left), the upper surface of the aerofoil leading edge is the free surface while the Stellite® repair is on the underside.  The yellow and red diamonds represent the gauge volume (in the neutron experiment) which is scanned through the blade in a vertical direction, by translating the whole blade in the neutron beam.

The final result, at one position from the leading edge is shown in right-hand panel above, with the 3 principal stresses mapped out on a half-millimetre scale, as function of depth underneath the repaired surface.

The net result of our study is that:

• Laser cladding generates considerable tensile stresses on both the clad surface and the rear face of the parent metal.
• There is a complex heat-affected zone stress profile due to differences in thermal expansion of the 2 materials.
• Grit blasting imparts compressive stress to ~100 mm depth on the blade surface.
• Post welding heat treatment is effective in minimising the stress profile.

The measurements on the HIFAR strain scanner used a gauge volume of 1 x 1 x 5 mm3, with each strain point taking ~25 minutes to collect.  On the new KOWARI residual-stress diffractometer at the OPAL reactor both data-recording times and gauge volumes will be significantly reduced.