Page 86 - IJAMD-1-3

International Journal of AI for
Materials and Design
Review of gas turbine blade failures by erosion

Ahsan et al. examined how erosion-induced pits and will occur. Kishore et al. used CFD to simulate particle
69
49
cracks become the starting points for fatigue fractures. collisions at various angles and velocities, identifying high-
Their research demonstrated that erosion damages the risk areas where erosion would likely lead to cracks and
protective surface layer of the turbine blade, making it structural damage.
more susceptible to oxidation and thermal fatigue. The
repeated removal of material leads to localized thinning, 3.4.4. Role of FEA in structural analysis
which reduces the blade’s structural integrity and load- While CFD focuses on fluid dynamics, FEA is used
bearing capacity. As cracks propagate, they lead to brittle to simulate how the blade material reacts to repeated
fracture or ductile tearing, depending on the material’s mechanical and thermal stresses caused by particle impacts.
temperature and properties at the time of failure. Khan and Sasikumar applied FEA models to study how
60
In addition, Błachnio et al. emphasized the role of erosion-induced material loss affects stress distribution
36
creep deformation in conjunction with erosion. At high across the blade. Using von Mises stress theory, they
temperatures, the blades are subjected to creep, a time- predicted the failure points where cracks would initiate
dependent deformation. Erosion accelerates this process by due to stress concentration caused by erosion pits.
removing material and exposing deeper layers to the same Moreover, FEA can simulate the fatigue life of a blade
high temperatures, leading to faster creep deformation. by predicting how cracks grow over time under cyclic
Over time, the combined effects of erosion, creep, and loads. Sabri et al. combined CFD and FEA to create
24
fatigue result in catastrophic blade failure. a fluid-structure interaction model, allowing them to
In summary, erosion creates surface defects that evolve simulate both the flow-induced erosion and the structural
into larger cracks under the combined influence of thermal deformation of the blades. Their integrated model
and mechanical stresses, leading to progressive fatigue provided a detailed view of how erosion progresses over
failure of the turbine blades over time. time, leading to blade failure.

3.4.3. Role of CFD in predicting failure points 3.4.5. Improving blade design
CFD and FEA are powerful tools used to simulate the These simulations are invaluable for optimizing blade
interactions between gas flows, solid particles, and turbine geometry and selecting materials to improve erosion
blade structures. 4,9,10,11 These simulations allow researchers resistance. Hoksbergen et al. demonstrated how CFD
38
to model the fluid dynamics of gas turbines and the and FEA simulations can be used to test different blade
structural response of blades to impacts, predicting failure designs, identifying configurations that reduce turbulence
points and optimizing blade design for better erosion and particle impact. For example, redesigning the leading
resistance. edge of a blade to minimize flow separation or applying
CFD simulations are used to model airflow patterns erosion-resistant coatings in high-impact areas can extend
and particle trajectories within GTEs, providing detailed the operational life of turbine blades. Surrogate models
insights into how particles behave in high-velocity gas and ML approaches, such as those explored by Yang
35
39
streams. Bera et al. employed CFD models based on et al., allow engineers to run simulations more efficiently
Navier–Stokes equations to simulate the velocity fields by approximating CFD and FEA results. This enables real-
and turbulence levels around turbine blades. They time predictions of erosion patterns and helps refine blade
identified erosion-prone zones, such as the leading edge designs without the need for time-consuming full-scale
and trailing edge, where particle impact is most severe. simulations.
CFD tracks particles through the gas flow, determining The leading edge of a turbine blade is one of the most
their trajectories, velocities, and impact angles with the erosion-prone areas due to its exposure to high-velocity
blade surface. This information is essential for predicting particle impacts. Modifying its geometry can significantly
erosion rates at different locations on the blades. Sun et al. reduce the intensity of these impacts and improve
62
also showed that particles striking the leading edge at high the airflow over the blade. A rounded leading edge is
velocity and near-normal angles cause significant material particularly effective because it minimizes stagnation
loss due to high-energy impacts. These simulations also pressure, reducing the energy transferred from particles
reveal flow separation and vortex formation, which to the blade surface on impact. This smoother curvature
contribute to localized erosion. also facilitates a more gradual transition of airflow over
By incorporating Eulerian–Lagrangian models, CFD the blade, decreasing the velocity gradient and reducing
allows for the simulation of solid particle impacts on the turbulence intensity. Another common design approach is
blade surface, predicting where the highest erosion rates the elliptical leading edge, which optimizes aerodynamic

Volume 1 Issue 3 (2024) 80 doi: 10.36922/ijamd.5188