A probabilistic approach for the integral inlet separator particle trajectory simulations by fitting probability functions with five points Gauss-Hermite methods was presented

and its impact on particle tracking accuracy was performed as compared with deterministic particle rebound model and elastic bounce model. Results show that

the above three bounce models perform well in the prediction of separation efficiency for AC coarse sand; for Cspec sand

results based on elastic bounce model deviates from reality; the deterministic particle rebound model overpredicts the separation efficiency of Cspec sand

especially for the swirling inlet particle separator

and the probabilistic bounce model achieves an intermediate value in good agreement with experimental data.