In the present paper, a steady subsonic gas flow either in a circular micropipe or in a planar microchannel driven by pressure within the slip flow regime is studied theoretically by using a perturbation expansion method to solve compressible Navier-Stokes equations. The isothermal flow assumption used in previous theoretical studies is given up. High-order boundary conditions of velocity slip and temperature jump are adopted at the wall. The set of dimensionless governing equations with two small similarity parameters, namely, the ratio of height to length ε, and the Knudsen number Kn, is approximated successively by using the perturbation expansions. The various cases such as ε≪Kn2, ε∼Kn2, and ε∼Kn1.5 are studied in detail. Explicit analytical solutions for pressure, density, velocity, temperature, and mass flow rate are obtained up to order of O(Kn2). It is shown that the solution formulas for long channels (ε≪Kn2) in lower order are in exact agreement with previous theoretical results. In particular, it is proved that the isothermal flow assumption is indeed reliable for relatively lower-order expansions. However, for higher-order expansions, the flow cannot be considered as isothermal, and the higher-order temperature correction is also given. The present high-order perturbation solutions can be applied even to a relatively shorter channel, and the results agree very well with those by the direct simulation Monte Carlo approach.

Scattering kernel models for gas–solid interaction are crucial for rarefied gas flows and microscale flows. However, most existing models depend on certain accommodation coefficients (ACs). We propose here to construct a data-based model using molecular dynamics (MD) simulation and machine learning. The gas–solid interaction is first modelled by 100 000 MD simulations of a single gas molecule reflecting on the wall surface, which is fulfilled by GPU parallel technology. The results showed a correlation of the reflection velocity with the incidence velocity in the same direction, and also revealed correlations that may exist in different directions, which are neglected by the traditional gas–solid interaction model. Inspired by the sophisticated Cercignani–Lampis–Lord (CLL) model, two improved scattering kernels were constructed to better reproduce the probability density of velocity determined from MD simulation. The first one adopts variable ACs which depend on the incidence velocity and the second one combines three CLL-like kernels. All the parameters in the improved kernels are automatically chosen by the machine learning method. Compared with the numerical experiments of a molecular beam, the reconstructed scattering kernels are basically consistent with the MD results.