Abstract
The various equations representing special cases of boundary lubrication are examined and grouped as “models”, The study indicates that there are no serious inconsistencies, but that each model is so seriously limited in its coverage that large gaps exist in coverage. Methods for closure of these gaps are suggested and numerically explored to a limited extent. Hope is extended that in the near future it will be possible to compute wear rates from basic physical properties without actual experimentation on the specific lubricant and surfaces. This would constitute a useful new system for machine element design.