Numerical simulation of labyrinth seals for pulsed compression reactors (PCR)

Resumo

A sealing system is proposed using labyrinth seals to minimize gas leaks, for which triangular,
rectangular, and trapezoidal geometric parameters are evaluated. For each one of the geometries a group of
parameters were optimized minimizing the gas leakage, using the multi-objective genetic algorithm (MOGA),
updating in each step the geometry employing a user defined Ansys SpaceClaim Python algorithm. For a new set
of parameters (height, width, and angle of the shape, space between cavities, and piston length) the script creates
the boundary geometry and meshing. The CFD analysis evaluates the gas leakage for the given geometry and
constant boundary conditions (10 MPa inlet pressure, 25 m/s piston speed, 40 m piston/cylinder gap, and ideal
methane gas), and uses this data as input to MOGA. The input set values were reduced to manufacturable
quantities, so finite or discrete values can be used across the iterations. The analysis of distributed properties such
as velocity, temperature, and pression, inside the cavity, showed a steady laminar regime with an energy loss due
to entropy increment. The most sensitive parameters are the piston length and height cavity for all shapes. The
trapezoidal shape presented the best performance in minimizing the mass flow leakage.