Back to search

FRINATEK-Fri prosj.st. mat.,naturv.,tek

Towards the next generation of lubricants by artificial intelligence

Alternative title: Towards the next generation of lubricants by artificial intelligence

Awarded: NOK 10.9 mill.

Lubricant formulation and production is a century-old industry in which the development of formulations is still based on existing recipes adding selected additives. Lubricants are used for achieving the required level of friction and wear rates in moving engineering systems. They can be either liquid, solid or semi-solid (greases). Their formulation consists of a base fluid in the highest proportion (up to 95%) and an additive package (chemical substances providing functionality and improving its properties). Current state of the art for producing lubricants is based on empirical experience of substances that have been functioning as lubricants for decades. In addition, the additives giving functionality to the lubricant are typically found by trial and error development. In this project, an artificial intelligence (AI) approach has been taken towards producing new lubricant formulations with focus on better performance, but also on sustainability and recyclability. The definition of the framework has been performed and the focus will be onto mining short chain PAO base lubricants and synthetic esters due to their environmental acceptability. Also, a machine learning approach staring with ionic liquids as additives has been started. Data bases are being gathered at the moment and the first algorithms are being produced. Data bases form the European ecolabel and substances used in the food industry are being sought.

Lubricant formulation and production is a century-old industry in which the development of formulations is still based on existing recipes adding selected additives. The new products need to be tested in a step-wise procedure, starting in lab scale (quick and cheap method), and scaling up until real tests (which are long term, expensive, difficult and complex processes) can be performed. This long, expensive and often frustrating process can be boosted with the help of Artificial Intelligence (AI). Formulating lubricants requires gathering information of available chemicals and substances that have the physical and chemical characteristics required for providing a system with the desired friction and wear. Mining in the available data of chemicals that could potentially become the next generation of lubricants requires new tools, methods, and new approaches, radically different from those used until now. Lubricants are used for achieving the required level of friction and wear rates in moving systems. They can be either liquid, solid or semi-solid (greases). Their formulation is rather simple, i.e. consists of a base fluid in the highest proportion (up to 95%) and an additive package (chemical substances providing functionality and improving its properties). Current state of the art for producing lubricants is based on empirical experience of substances that have been functioning as lubricants for decades. In addition, the additives giving functionality to the lubricant are typically found by trial and error development. An example of this is the additives based on Zn that were originally developed as antioxidants and finally turned out to be the best additives for friction and wear control . However, this approach brings only small incremental improvements in lubricant formulations. Therefore, the development for covering the demands of the new generation of lubricants is happening at lower speed than required for applications and for environmental legislations.

Funding scheme:

FRINATEK-Fri prosj.st. mat.,naturv.,tek