A low cost mid-infrared sensor for on line contamination monitoring of lubricating oils in marine engines

Abstract

The chemical and physical condition of oils in marine engines must be monitored to ensure optimum performance of the engine and to avoid damage by degraded oil not adequately lubricating the engine. Routine monitoring requires expensive laboratory testing and highly skilled analysts. This work describes the adaptation and implementation of a mid infrared (MIR) sensor module for continued oil condition monitoring in two-stroke and four-stroke diesel engines. The developed sensor module will help to reduce costs in oil analysis by eliminating the need to collect and send samples to a laboratory for analysis. The online MIR-Sensor module measures the contamination of oil with water, soot, as well as the degradation indicated by the TBN (Total Base Number) value. For the analysis of water, TBN, and soot in marine engine oils, four spectral regions of interest have been identified. The optical absorption in these bands correlating with the contaminations is measured simultaneously by using a four-field thermopile detector, combined with appropriate bandpass filters. Recording of the MIR-absorption was performed in a transmission mode using a flow-through cell with appropriate path length. Since in this case no spectrometer is required, the sensor including the light source, the flowthrough-cell, and the detector can be realised at low cost and in a very compact manner. The optical configuration of the sensor with minimal component number and signal intensity optimisation at the four-field detector was implemented by using non-sequential ray tracing simulation. The used calibration model was robust enough to predict accurately the value for soot, water, and TBN concentration for two-stroke and four-stroke engine oils. The sensor device is designed for direct installation on the host engine or machine and, therefore, becoming an integral part of the lubrication system. It can also be used as a portable stand-alone system for machine fluid analysis in the field.

Non-Dispersive Infrared Sensor for Online Condition Monitoring of Gearbox Oil

“…Through the use of multivariate PLS regression for the sensor calibration, the information content of all infrared channels is used to determine oxidation, water content, and AN of the oil. The optical bandpass filters for NDIR condition monitoring of phosphate-ester hydraulic fluid [ 12 , 13 ], gas engine oil [ 14 ] and diesel engine oil [ 15 ] were selected on the basis of previous knowledge of the chemical reactions occurring during the oil aging process. We have shown that the bandpass filters can also be selected by means of a regression analysis, which means that no prior knowledge of the chemical aging behavior of the oil is necessary.…”

“…As infrared spectrometers are relatively expensive and are usually not robust enough to be operated in industrial environments, approaches that use non-dispersive infrared (NDIR) measurement setups have been made to monitor additive depletion [ 10 ] or oil oxidation [ 11 ]. NDIR setups for monitoring phosphate-ester hydraulic fluid [ 12 , 13 ], gas engine oil [ 14 ], and diesel engine oil [ 15 ] have been presented.…”

Asset Management

“…The Posseidon project (Progressive Oil Sensor System for Extended Identification ON-Line) addressed this by using sophisticated sensors to monitor the condition of lubricating oil. This provides not only a snapshot of the performance of the oil but also provides indications of other engine faults which result in contamination of the oil (Baglee and Knowles 2010a,b, Gorritaxetegi et al 2007, Mohammadi et al 2010. The sensors developed are capable of making measurements of oil properties including:…”

Study of on-line condition monitoring and fault feature extraction for marine diesel engines based on tribological information

“…The accelerated oxidation, fires and contamination of the lubrication oil could be detected by the OCM. Mohammadi et al 14 have designed a novel on-line mid-infrared sensor with low cost for oil monitoring of marine engines. Villar et al 15 designed a visible-near infrared (400–1100 nm) sensor to monitor the characteristics of in-use marine engine oil.…”