Understanding and modelling the mixed lubrication regime: summary of achievements and perspectives of industrial applications

The purpose of modelling is generally to be predictive, and these works on friction and lubrication in cold strip rolling were no exception. Predictiv ity is e.g. necessary when trying to design new schedules for new products, or even when designing new rolling mills. To this end, a tribological mode l needs be founded on a careful analysis of fundamental tribological phenom ena. Our vision of cold rolling tribology was deeply transformed and renewed by the works presented(*) issue, starting from the behaviours of superficial m etal and oxide layers: brittle alumina undergoes a complete transformation upon cold rolling, together with the underlying metal which may recrystalli ze, more or less deeply depending on rolling conditions. More ductile iron oxides offer a better protection of the metal. Polar additives form aluminium soaps on naked metal surface only; on iron, on both oxide and naked metal. Alcohols are oxidized into aldehydes and aci ds and hydroxylate the oxide surfaces. Sulfur EP additives form FeS on iron , phosphate esters form organo-metallic phosphates. Reaction products are f ound much more abundant after durable contact (UHV Tribometer, LTDS) than a fter single, short contacts (rolled strip, CEMEF): their action should be m ore through reaction with rolls than with strips. Among the new tools and m ethods, ToF-SIMS and gas phase lubrication in the Ultra-High Vacuum Tribome ter have proven quite fruitful. The mechanisms by which polar additives start to concentrate near surfaces before reacting with them have been studied. Even the base oil may undergo demixtion at surfaces if composed of molecules of different polarities. Fur thermore, these adsorbed layers promote sliding of the lubricant film, depe nding on their arrangement ("chemical roughness"), with consequences on fil m formation and load bearing capacity. By the way, the conditions of load bearing capacity creation have been stud ied, in particular its decrease due to micro-cavitation in divergent areas, strongly depending on roughness and its orientation. Continuously decreasi ng gaps are therefore a must for the high load bearing capacities necessary in metal forming. A model of the mixed lubrication regime describes the evolution of tribolog ical internal variables and friction. Coupled with a rolling process model, it allows the investigation of the effect of a large number of process or material parameters, to draw e.g, sets of Stribeck curves. A few applications have already been effected to understand or correct prac tical problems. Thus, the strip tension - forward slip characteristics can be studied (fig. I); it is suggested to have some impact on mill vibrations ("chatter"), a most detrimental phenomenon. The model also shows an increa se of friction coefficient when the roll diameter is decreased, which hinde rs the positive effects expected from such a change (smaller rolling force and torque). It allows to discriminate, among the many factors influencing friction in thin foil rolling, those with a major impact, which may most pr ofitably be used as actuators (speed, oil viscosity, strip and roll roughne ss). Potential applications to lubricant formulations are also evoked; roug hness morphology control may be another way of better controlling friction. Certainly, predictivity of the model is jeopardised by the poor knowledge o n the boundary friction (on "plateaux"). If the nature of these boundary fi lms is better and better known, their tribological properties are hardly un derstood, even in studies carried out in very academic situations like in t he Surface Force Apparatus (SFA) or Friction Force Microscopes (FFM). They include the viscosity of thick glassy films formed by extreme pressure addi tives (EP); or the molecular-scale friction of grafted or self-assembled fi lms of polar molecules. Thus by now, a friction coefficient has to be chose n arbitrarily for these boundary areas. No doubt that the rapid progress of nano-techniques based research will allow application to our complex indus trial systems. The final challenge is to apply the developed tools and techniques to emuls ion lubrication, which introduces two new components, emulsifiers and water , and their interactions with metal or oxide surfaces. Such phenomena as "p late out" (demixtion of oil from water forming a fat film on the surfaces) or "dynamical concentration" (preferential entrainment in the bite of oil d roplets as compared to less viscous water) then become dominant. Models of these mechanisms exist and could be operated with profit in a strip rolling context.