Research On Measurement Of Solid Particles In Jet Fuel With Multi-angle Dynamic Scattering Method

Aviation kerosene would go through storage,transportation and other links from production to application with a long cycle,which should inevitably be contaminated by foreign impurity particles.Once too many solid particles are mixed in the aircraft engine,it would seriously affect the aircraft flight safety.According to the statistics,about 50%of engine failures are caused by poor fuel cleanliness^[1].The internal structure of the engine is complex with mostly precision components,the maintenance of which is difficult and expensive,therefore,the aviation kerosene input to the fuel tank should be ensured to be clean.Because of its wide applicability,wide measuring range and high measuring precision,light scattering method is gradually becoming an important method to measure particulates in gas or liquid.It is difficult to measure all the kerosene because of the problem of randomicity in sampling static detection,so the cleanliness of aviation kerosene in fuel tank can not be fully guaranteed.The existing dynamic measurement methods can measure only a small number of samples per unit time,and can not meet the requirements of a large number of aviation kerosene samples.Therefore,the measurement of impurity particles in aviation kerosene with large sample size has attracted many attentions.In this paper,the particle size and mass concentration of impurity particles in aviation kerosene have been measured by Mie scattering and extinction method to solve the problem of dynamic measurements.Here the theoretical research and experimental verification are carried out from the following five aspects:simulation model establishments,scattering characteristic simulations,optical-mechanical structure designs,experiment verification and simulation rules,and parameter inversions.The main contents of this paper are as follows:Firstly,in this paper,Mie scattering and extinction theory has been used to deduce the light flux calculation formula of multi-scattered particles with three angle directions in a large space range.Continuous-uniform-distribution random method has been used to simulate the position of each particle scattered in the light path,thus the complexity of the calculation has been simplified.The calculation model of particle group scattering light flux in aviation kerosene in a large space range has been constructed.It lays a foundation for measuring particle sizes and mass concentrations by multi-angle light scattering method.The simulation results show that the larger the particle size is,the more concentrated the scattered light intensity is in the small forward angle,and the weaker the backscattered light intensity is.In the case of the same particle size,the scattering law of the particle group in a large space is linear relationships between the concentration and the signal value received by the instrument,in which the scattered flux presents a linear positive correlation with the increase of the concentration,while the transmitted flux signal presents a linear negative correlation with the increase of the concentration,and the ratio between the luminous flux and the mass concentration presents an exponential relationship with the increase of the particle size.Then,according to the above theory,the optical design of multi-angle and large sampling dynamic scattering measurement device and the research of four-channel synchronous signal acquisition system have been carried out.A multi-angle large sample dynamic scattering measuring device for measuring solid particles in aviation kerosene has been developed.The rotational dynamic measurement method,instead of the cyclic dynamic measurement method,has been used to simulate the equivalent flow rate of 187L/min,250L/min and 310L/min.The scattering law in the aviation kerosene measured by the test is consistent with the simulation law.Finally,support vector machine regression method has been used to invert the mean particle size and mass concentration synchronously.First,a model has been established and the relationships between the four-channel data and the mean particle size and mass concentration have been estimated,with goodness-of-fit results of 99%for the four parameters obtained.Then,support vector machine regression method has been used to invert the mean particle size and mass concentration of Cu particles.By comparing the inversion errors of average particle size and mass concentration between two features and four features,it is found that the inversion accuracy is higher using four features.