Page 83 - IJAMD-1-3

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

gas turbine blades respond to erosive forces. 42,43 This where the Eulerian frame is used to track the gas flow, and
integrated approach helps engineers predict failure the Lagrangian frame tracks individual particles. 75,76 The
points, optimize blade designs, and develop more effective particles’ velocities and trajectories are governed by the
maintenance strategies. In the context of CFD and FEA equation of motion given by Equation VI. The equation of
simulations, fluid dynamics equations govern the airflow motion for a particle under the influence of aerodynamic
and particle trajectories, while structural mechanics forces is given by Equation VI:
equations predict deformation and stress under erosion
impacts. CFD is critical in gas turbine analysis because it dv p
simulates the complex fluid flow and particle behavior in m p dt = F + F + F l (VI)
d
g
the high-temperature, high-velocity environment inside
turbines. 12,15,27,40 The purpose of CFD in erosion analysis where m is the mass of the particle; v is the velocity
p
p
is to understand how the gas flow and solid particles of the particle; F is the drag force; F is the gravitational
g
d
interact with the turbine blades, contributing to surface force; and F is the lift force. By solving Equation VI, CFD
l
degradation. simulations predict where and with what energy particles
The Navier–Stokes equations describe the motion impact the blade surface. This information is passed to
of fluid (air) around turbine blades, particularly for FEA models to analyze the structural response. FEA
predicting high-speed airflows and particle paths. The complements CFD by analyzing how the turbine blades
Navier–Stokes equations govern fluid flow, describing the respond to the mechanical and thermal stresses caused by
conservation of momentum, mass, and energy in the gas particle impacts. 23,24,31 FEA breaks down the turbine blade
phase. These equations are solved to compute the velocity into smaller elements, allowing for a detailed examination
71
field, pressure distribution, and temperature gradients of stress distributions and material deformations.
in the airflow around the turbine blades. By solving the FEA is used to predict how the blade structure responds
Navier–Stokes equations, CFD helps predict where the to the stresses induced by particle impacts. FEA solves for
flow separation and vortex formation occur, which are the displacement field of the structure under load, using
known to concentrate particle impacts. The velocity the relationship between stress and strain, as shown by
field around the blades, especially near the leading edges Equation VII, derived from Hooke’s Law for linear elastic
and trailing edges, indicates where the highest particle materials. 39,43,47 The governing equation in FEA for linear
velocities will lead to severe erosion. 69,73,74 The general form elastic materials is derived from Hooke’s law and the
of the Navier–Stokes equations for an incompressible fluid equilibrium of forces, as shown by Equation VII:
is given by Equation V:
σ + f = 0 (VII)
 ∂u  ij j, i
ρ  +⋅∇u u  =−∇+ µ∇ u + f (V)
2
p
 ∂t 
where σ is the stress tensor; and f is the body force.
i
ij
where ρ is the fluid density; u is the velocity field of the FEA divides the blade into a mesh of small elements
fluid; t is time; p is the pressure; μ is the dynamic viscosity; and solves this equation at each node of the mesh to
and f represents external forces (e.g., gravity). In CFD calculate deformations and stresses under operational
simulations, this equation is solved to predict velocity loads and particle impacts. Using von Mises stress theory,
78
fields and particle trajectories, which determine where the failure criterion can be written as Equation VIII:
particles will strike the turbine blades and the impact
velocities. In our theoretical framework, CFD simulations σ = 1  ( σ − ) +( σ − ) +( σ − ) 2  (VIII)
2
2
σ
σ
σ
are critical for identifying erosion-prone regions on v 2   1 2 2 3 3 1  
turbine blades. 72-74 By solving the Navier–Stokes equations
and tracking particle impacts, CFD helps predict surface where σ , σ , and σ are the principal stresses at a given
3
2
1
wear patterns and hotspots where erosion is most likely to point.
occur. This information is then passed to FEA for further To predict failure, the von Mises stress is used as a
structural analysis. criterion for yielding in ductile materials. The material is
In the Eulerian–Lagrangian framework, the gas phase considered to fail when the von Mises stress exceeds the
(airflow) is treated as a continuous medium (Eulerian yield strength of the material. FEA simulations use this
approach), while particles are tracked individually to predict when and where blade failure will occur as a
(Lagrangian approach). The particles carried by the gas result of erosion-induced stresses. By coupling CFD and
8,9
flow are modeled using the Eulerian–Lagrangian approach, FEA, engineers can predict erosion patterns and stress
Volume 1 Issue 3 (2024) 77 doi: 10.36922/ijamd.5188