We live in times of massive, digital transformation. The challenge of this for research and development is to find the right digital methods for our specific needs from the many possibilities. This is because data analysis, modeling and simulations do not present an end in themselves. They have to make it possible to produce new and more efficient products in shorter development times.
There is huge potential for using digital tools in development of new lubricants. They make it possible to predict previously unknown processes and properties. They thereby help not only to forecast critical operating states in devices to be lubricated, but they also make a significant contribution to achieving a deeper understanding of our lubricants. For example, in a validating laboratory experiment there is no or limited access to extreme conditions such as those that occur in real machinery. However, they can now be mapped or extrapolated using simulation methods. This enables us to adjust our products to achieve an even closer match to the requirements of our customers, which currently means above all further reducing wear and energy consumption in machinery, thereby increasing efficiency.
Another major advantage of computer simulation is reflected in the “virtual” assessment of alternative, previously unavailable raw materials, the use of which may hold promise. Expensive experiments based on the trial-and-error principle can be dispensed with and development times can be shortened.
NEW LUBRICANTS FOR E-MOBILITY
Our customers in the automotive industry are also facing various new challenges. The transition to e-mobility is central here. The electric drivetrain will play a leading role in future transport concepts, as will the combination of electric motor and transmission. Such systems call for different lubricants than those required by a combustion engine. For example, it is possible to combine the transmission lubricant and the motor and power electronics coolant in a single product. The compatibilities with materials such as copper, but also with high temperatures that are prevalent in electric motors – due to high speeds and strict performance requirements – are particularly important here. It is vital to take into account all these conditions during the development of new lubricants.
THE DIGITAL TOOLBOX
Suitable digital methods can be used extremely profitably to complete these development tasks. They provide a better understanding of our lubricants and sometimes even predict the properties of new formulas. The impacts of these properties on the overall system can also be calculated.
Various digital tools are used in research and development at FUCHS. These include data analysis, chemoinformatics, which is used to calculate the properties of molecules, and the design of experiments. However, alongside rapidly growing processing power, modeling and simulation methods are becoming increasingly important. This enables us to look, as though with a magnifying glass, further inside the virtual lubricant – so far until eventually we reach nano-scale, a range of a few nanometers in which molecules and atoms become visible. We are applying this approach to three different levels of variables, right down to nano-scale: on the macro-scale (larger than 10–3 m) we are simulating the technical unit, on the micro-scale (approximately 10–6 m) the lubricant gap, and at the lower end, on the nano-scale (smaller than 10–9 m) the molecules.
Modeling approaches of today and tomorrow
We are conducting extensive research on lubricants, across all variables and timescales: from the behavior of individual molecules within fractions of a second to predictions of properties for technical applications over longer periods. In the long term, we aim to recognize the decisive influencing variables across all these dimensions, understand the significant correlations between them and combine the associated modeling approaches in an integrated method.
NANO-SCALE: THE CALCULATED ADDITIVE
The nano-scale tests conducted by the research and development department at FUCHS are reflected in a current development project from the area of electric machinery: to protect against the premature wear of electric motors, what is called an “anti-wear additive” was “virtually” added to the oil-based lubricant. This additive was digitally simulated under “operating conditions.” We were particularly interested here in the temperature range in which the molecule takes effect – in other words, the moment in which it reacts. However, the behavior in relation to copper was also important.
For this purpose, the additive was released in a mathematical model via ab initio methods. Equations were solved with state-of-the-art computers, thereby predicting the additive’s properties. This simulation not only helped us to understand the additive’s reaction mode, but the digital tools also pointed out molecules with improved properties that were then synthesized using conventional methods and are now in the test phase.
MICRO-SCALE: THE SIMULATED LUBRICATION GAP
Despite all the progress made in terms of high-performance computers, such calculations at a molecular level cannot be directly extended to our empirical world. Here, we would be talking about over 1023 molecules, an unimaginably large number.
Through simplified approaches to molecular behavior, digital simulations can nevertheless be completed “one level higher” – to the micrometer scale (10–6 m). This will enable us to describe processes in the lubrication gap at the micro-scale level. It is therefore possible to make predictions regarding how our lubricants and additives affect wear and the surface behavior. And measurement variables that are difficult to ascertain in experiments, such as the lubricants’ flow properties under extreme pressures and at high speeds, are also “digitally accessible.”
MACRO-SCALE: SIMULATION AT THE PRODUCT LEVEL
The use of application-specific optimized lubricants is crucial to ensure the reliability and efficiency of lubricated systems. Simulation at the macro-level comprises the virtual mapping of individual machine elements, from gears and roller bearings to consideration of the overall system.
Modern simulation solutions and a close working relationship with research partners make it possible to identify the relevant contact and operating conditions of various applications. With this knowledge, we can perform targeted investigations of the necessary product-specific properties using model test benches, thereby ensuring that our products meet the specified requirements.
By implication, this means that we are developing digital methods to make it possible to map and predict, for example, the wear-related behavior of our products as realistically as possible using calculation approaches. As a result, we aim to make it possible for our customers to recognize relevant correlations between fluid and application at a very early phase of their own system development in order to make sound and reliable assessments of these with which to ensure optimal coordination. These possibilities enable virtual pre-screening – and thus significant reductions and optimizations in terms of both our development processes and those of our customers.
AI AS A FORMULATION AID
Alongside simulations, data analysis is becoming increasingly important in research and development. Assessments of completed research projects play no less of a role here than the design of experiments. With methods that combine both approaches, models can be built for new lubricant formulations and customer requirements satisfied in a shorter time.
With clustering algorithms, we can also recognize various patterns in data arising from the use of our lubricants – and access valuable information regarding cause and effect mechanisms. Once these mechanisms have been understood – and there is an understanding of the formulations used as well as the chemical structures of the raw materials – neural networks can be trained to predict certain oil properties. This is how we are implementing artificial intelligence (AI) in lubricant development and optimization.
Digitalization is a megatrend. It will continue to pick up in speed. New generations of computers as well as better and quicker algorithms will make it possible to produce lubricants that are constantly improving in quality and efficiency. Anyone wishing to set the pace in development has to make use of it.
