Lubricating oil is used to separate two surfaces in relative motion. To adequately reduce wear and maintain optimum efficiency, it is vital to know the viscosity of the lubricating oil. Most lubricating oils are non-Newtonian, meaning the viscosity is dependent on the shear rate. This thesis develops a tool to carry out viscosity measurements of lubricating oils at varying high shear rates in-situ.
To fully understand how the lubricant will react in-situ it is necessary to subject the oil to the harsh conditions found in an engine. However, this is very challenging to replicate. If viscosity measurements are completed in-situ, it negates the need to replicate the harsh conditions. Due to the many advantages of ultrasound, including low cost, ability to complete measurements in harsh conditions and continuous measurements, the ultrasound method is the most promising technique to complete high shear
in-situ measurements. The shear reflection method obtains the viscosity of the lubricating oil by measuring the returned amplitude of an emitted shear ultrasound wave. When the emitted wave reaches a boundary, a proportion of the wave will be reflected, and a proportion will be transmitted. The
proportion reflected is dependent upon the acoustic impedance of the media on either side of the boundary. For lubricating oil, its acoustic impedance is related to its viscosity.
The ultrasonic shear reflection viscometer had to be adapted to measure the viscosity of lubricating oils because the majority of the signal was reflected as the difference in the acoustic impedance of the piezoelectric crystal and oil was too great. This made it impossible to determine the viscosity of lubricating oils. By installing an intermediary layer, known as the matching layer, between the piezoelectric crystal and the lubricating oil, a greater proportion of the signal was transmitted, and the reflection coefficient could be used to determine the viscosity of lubricating oils.
This thesis focuses on understanding the shear rate exerted on the test oil by the ultrasonic shear reflection viscometer, as without this information the viscometer is unable to be used to measure the viscosity of non-Newtonian fluids, such as engine oils. In this study, a mathematical model was developed to determine the viscosity of Newtonian and non-Newtonian fluids from the pressure reflection coefficient for a three-layered system. The model illustrated the advantages of the three-layered system, and a sensitivity analysis was completed to understand the relationship between frequency and reflection coefficient.
This model was validated by measuring the viscosity of Newtonian fluids with a viscosity range of 16 − 710cP, using an ultrasound viscometer over the frequencies 1 − 10MHz and comparing with viscosity results from a low shear Couette viscometer. This multiple-frequency matching layer viscometer was instrumented with different frequency transducers and different thickness matching layers to obtain maximum sensitivity. Using the multiple frequency matching layer viscometer different input signal settings were altered to find the optimal settings.
A further model was then developed from Stokes’ second problem, which is used to describe the flow from an oscillating surface, to understand the shear rate exerted on the lubricating oil from the ultrasound matching layer viscometer. Previous researchers have successfully used the Cox-Merz rule
for simple fluids at low shear rates which states that the shear rate for a steady and oscillatory shear can be related by, η( ˙ γ) = η(ω), where η is the dynamic viscosity, γ˙ is the shear rate, and ω is the angular velocity. This thesis argues that for high shear rate measurements, the shear rate from the ultrasonic shear reflection viscometer is dependent on the viscosity and velocity of the fluid and the frequency of the shear wave.
To determine the shear rate exerted on the lubricating oil using the derivation from Stokes’ second problem, the velocity of the lubricating oil was required. Due to the non-reflective nature of lubricating oil, it was indirectly measured using a laser vibrometer and was found to be dependant on the excitation voltage.
Viscosity measurements of non-Newtonian lubricants were completed on the multiple frequency matching layer viscometer and compared with results obtained on an ultra-shear Couette viscometer. The results were more similar using Stokes’ second problem than Cox-Merz, which validated the shear rate model. Finally using this model, a shear rate viscosity plot was created for Newtonian and non-Newtonian oils.
