Optimized Mask-RCNN model for particle chain segmentation based on improved online ferrograph sensor

Ferrograph-based wear debris analysis(WDA)provides significant information for wear fault analysis of mechanical equipment.After decades of offline application,this conventional technology is being driven by the online ferrograph sensor for real-time wear state monitoring.However,online ferrography has been greatly limited by the low imaging quality and segmentation accuracy of particle chains when analyzing degraded lubricant oils in practical applications.To address this issue,an integrated optimization method is developed that focuses on two aspects:the structural re-design of the online ferrograph sensor and the intelligent segmentation of particle chains.For enhancing the imaging quality of wear particles,the magnetic pole of the online ferrograph sensor is optimized to enable the imaging system directly observe wear particles without penetrating oils.Furthermore,a light source simulation model is established based on the light intensity distribution theory,and the LED installation parameters are determined for particle illumination uniformity in the online ferrograph sensor.On this basis,a Mask-RCNN-based segmentation model of particle chains is constructed by specifically establishing the region of interest(ROI)generation layer and the ROI align layer for the irregular particle morphology.With these measures,a new online ferrograph sensor is designed to enhance the image acquisition and information extraction of wear particles.For verification,the developed sensor is tested to collect particle images from different degraded oils,and the images are further handled with the Mask-RCNN-based model for particle feature extraction.Experimental results reveal that the optimized online ferrography can capture clear particle images even in highly-degraded lubricant oils,and the illumination uniformity reaches 90%in its imaging field.Most importantly,the statistical accuracy of wear particles has been improved from 67.2%to 94.1%.

Understanding the microscopic moisture migration in pore space using DEM simulation

The deformation of soil skeleton and migration of pore fluid are the major factors relevant to the triggeringof and damages by liquefaction. The influence of pore fluid migration during earthquake has beendemonstrated from recent model experiments and field case studies. Most of the current liquefactionassessment models are based on testing of isotropic liquefiable materials. However the recent NewZealand earthquake shows much severer damages than those predicted by existing models. A fundamentalcause has been contributed to the embedded layers of low permeability silts. The existence ofthese silt layers inhibits water migration under seismic loads, which accelerated liquefaction and causeda much larger settlement than that predicted by existing theories. This study intends to understand theprocess of moisture migration in the pore space of sand using discrete element method （DEM） simulation.Simulations were conducted on consolidated undrained triaxial testing of sand where a cylindersample of sand was built and subjected to a constant confining pressure and axial loading. The porositydistribution was monitored during the axial loading process. The spatial distribution of porosity changewas determined, which had a direct relationship with the distribution of excess pore water pressure. Thenon-uniform distribution of excess pore water pressure causes moisture migration. From this, themigration of pore water during the loading process can be estimated. The results of DEM simulationshow a few important observations： （1） External forces are mainly carried and transmitted by the particlechains of the soil sample; （2） Porosity distribution during loading is not uniform due to nonhomogeneoussoil fabric （i.e. the initial particle arrangement and existence of particle chains）; （3）Excess pore water pressure develops differently at different loading stages. At the early stage of loading,zones with a high initial porosity feature higher porosity changes under the influence of external loading,which leads to a larger pore pressure variation （increase or decrease） in such zones. As the axial strainincreases, particle rearrangement occurs and final porosity distribution has minor relationship with theinitial condition, and the pore pressure distribution becomes irregular. The differences in the porepressure development imply that water will migrate in the pore space in order to balance the pore waterpressure distribution. The results of this simulation offer an insight on the microscale water migration inthe soil pore space, which is important for holistic description of the triggering of soil liquefaction in lightof its microstructure. 2015 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Production and hosting byElsevier B.V. All rights reserved.

A numerical study of macro-mesoscopic mechanical properties of gangue backfill under biaxial compression

Based on the Particle Flow Code(PFC^(2D)) program,we set up gangue backfill models with different gangue contents and bond strength,and studied the stress-strain behaviours,the pattern of shear band and force chains,motion and fragmentation of particles under biaxial compression.The results show that when the bond strength or contents of gangue are high,the peak strength is high and the phenomena of post-peak softening and fluctuation are obvious.When gangue contents are low,the shape of the shear band is symmetrical and most strong force chains transfer in soil particles.With an increase in gangue content,the shape of the shear band becomes irregular and the majority of strong force chains turn to transfer in gangue particles gradually,most of which distribute along the axial direction.When the gangue content is higher than 50%,the interconnectivity of strong force chains decreases gradually:at the same time,the strong force chains become tilted and the stability of the system tends to decrease.With an increase in external loading,the coordination numbers of the system increase at first and then decrease and the main pattern of force chains changes into columnar from annular.However,after the forming of the advantageous shear band,the force chains external to the shear band maintain their columnar shape while the inner ones bend obviously.As a result,annular force chains form.

Accessing the syndemic of COVID-19 and malaria in tervention in Africa

Background:The pandemic of the coronavirus disease 2019(COVID-19)has caused substantial disruptions to health services in the low and middle-income countries with a high burden of other diseases,such as malaria in sub-Saharan Africa.The aim of this study is to assess the impact of COVID-19 pandemic on malaria transmission potential in malaria-endemic countries in Africa.Methods:We present a data-driven method to quantify the extent to which the COVID-19 pandemic,as well as various non-pharmaceutical interventions(NPIs),could lead to the change of malaria transmission potential in 2020.First,we adopt a particle Markov Chain Monte Carlo method to estimate epidemiological parameters in each country by fitting the time series of the cumulative number of reported COVID-19 cases.Then,we simulate the epidemic dynamics of COVID-19 under two groups of NPIs:(1)contact restriction and social distanci ng,and(2)early ide ratification and isolation of cases.Based on the simulated epidemic curves,we quantify the impact of COVID-19 epidemic and NPIs on the distribution of insecticide-treated nets(ITNs).Finally,by treating the total number of ITNs available in each country in 2020,we evaluate the negative effects of COVID-19 pandemic on malaria transmission potential based on the notion of vectorial capacity.Results:We con duct case studies in four malaria-endemic coun tries,Ethiopia,Nigeria,Tanza nia,and Zambia,in Africa.The epidemiological parameters(i.e.;the basic reproduction number R°and the duration of infection D1)of COVID-19 in each country are estimated as follows:Ethiopia(Rq=1.57,D1=5.32),Nigeria(Ro=2.18,D1=6.58),Tanzania(Ro=2.47,D1=6.01),and Zambia(R0=2.12,D1=6.96).Based on the estimated epidemiological parameters,the epidemic curves simulated under various NPIs indicated that the earlier the interventions are implemented,the better the epidemic is controlled.Moreover,the effect of combined NPIs is better than contact restriction and social di st a ncing only.By treating the total number of ITNs available in each country in 2020 as a baseline,our results show that even with stringent NPIs,malaria transmission potential will remain higher than expected in the second half of 2020.Conclusions:By quantifying the impact of various NPI response to the COVID-19 pandemic on malaria transmission potential,this study provides a way tojointly address the syndemic between COVID-19 and malaria in malariaendemic countries in Africa.The results suggest that the early intervention of COVID-19 can effectively reduce the scale of the epidemic and mitigate its impact on malaria transmission potential.