Optimization of Lubricant Design through Use of Design of Experiments Methodology By: Jason Galary Applied Science and Tribology, Nye Lubricants, Inc. Department of Mechanical Engineering, University of Massachusetts Dartmouth A new automotive lubricating grease was formulated using polytetrafluoroethylene (PTFE) to thicken polyalphaolefin (PAO) base oil. This grease was developed to meet targets for rheology and protection against wear for automotive applications. The formulation was developed and the manufacturing process was optimized using computational data analysis instead of a more typical trial-and-error approach. The computer data analysis methodologies were design of experiment (DOE), regression analysis (RA), and evolutionary algorithm (EA). In this study, a DOE methodology was employed to optimize the process of designing a lubricating grease. Typically, a lubricant is designed through use of a combination of technical knowledge, prior experience, and trial-and-error experimentation to develop a starting point for formulation and processing conditions. Then, modifications are made along the way to improve the formulation and processing procedures. While this methodology will get the job done, it may take longer and produce a less optimum product than one created through a DOE methodology. This study illustrates the benefits of a well-laid-out DOE plan to design a new PAO-PTFE formulation for the lubrication of automotive steering shafts and other components. Data from the DOEs were analyzed by RA to construct empirical models. Then, some models were optimized using an EA. The final results were predictions of the optimal formulation and manufacturing procedures for this grease. These predictions were compared with data for validation. Results from this study support the application of DOE, RA, and EA methodologies to develop formulations and manufacturing procedures.
Introduction
There are many goals to be considered when designing a new product. During the development of any material or product, careful attention must be paid to raw materials,
manufacturing process, and validation testing. However, it is not always feasible to perform extensive experimental designs including all the validation testing for every possible variable due to time and cost constraints. Many times in industrial laboratories, future generations of products are developed by modifying existing ones, which inherit their advantages (and sometimes their problems). This creates a situation where a new formulation is not optimized and scaling up is difficult because it is not optimum for the new formulation. A low yield in scale-up or production, cost variances due to increased amounts of raw materials or time, and inconsistent product performance data can occur when a formulation or production process is not robust A robust product or process design is insensitive to changes caused by normal variability.
Background
This study applies a design of experiments (DOE) methodology to develop and optimize a new lubricating grease formulated using a fluoropolymer thickener (polytetrafluoroethylene, PTFE) to bind a polyalphaolefin (PAO) oil. This grease is intended for automotive steering applications. The manufacturing process for these materials can range from very simple mixing to a much more complicated process that involves specialized mixers, homogenizing equipment, and milling processes. A DOE approach was used to develop this product with more control and less variability than typical trial-and-error approaches. Design of Experiments Methodology - Basics DOE is a set of tools and guidelines for running experiments in a controlled fashion. DOEs can provide insight into the factors that affect the outcome of a process or product. In this study, DOE is used to develop the formulation and processing conditions for a PAO-PTFE grease that meets targets for anti-wear and rheology performance.
- 22 VOLUME 81, NUMBER 6