Sub-Phasing

Sub-phasing capability, while essentially optional, represents structural change inside nanoFluidX more than a classic feature. Over the years, real-life application examples grew in size, both in terms of number of particles and number of phases. This has reflected on the performance of the code, so improvements have been made to address this.

This was particularly obvious, for example, in the case of chained transmissions, where each chain element is defined as a single phase, resulting often in total number of phases which approaches a hundred.

The speed up seen from using sub-phasing option can be significant. The largest gain that was seen in the testing for a realistic gearbox was 2.9x. The nature of the structural change implies that cases with more phases are likely to see larger performance benefits when using sub-phasing.

Finally, it is important to note that sub-phasing cannot be used when energy transport is enabled in the nanoFluidX solver.

Using sub-phasing has an impact on slip/no-slip, FCT and adhesion coefficient definitions. When sub-phasing is enabled, the slip/no-slip, FCT and adhesion parameters will be picked up from the first phase type definition and applied to the rest of the components (phases). For example if there are multiple MOVINGWALL phase types with makesubphase option set to true, the settings from the first phase definition will apply to all the other MOVINGWALL phases.

If the user would like to specify a unique value for slip/no-slip, FCT or adhesion parameters for a specific phase – the makesubphase command can be set to false and then the specific component will be treated as a separate phase inside the code.