Study On The Influence Of Heating Conditions On The Multi-scale Structure Of Copper Wire And Image Recognition

With the increasing degree of electrification in the current society,a large number of electrical equipment appear in people ’s daily life,but also increase the fire hazard of modern buildings.In 2022,the National Fire and Rescue Team will receive and deal with alarms and fires issued by the Fire and Rescue Bureau of the Ministry of Emergency Management.Electricity is the primary cause of fires,accounting for 31.1%of the total number of fires.Larger fires are caused by one-third of the electrical causes,and the majority of electrical circuit failures,accounting for 80 % of the total number of electrical fires.The frequent occurrence of electrical fires puts forward new requirements for fire investigation.Copper wire is the most widely used wire type in electrical lines,and the metal can be completely retained after fire,so its fire trace plays a vital role in fire investigation.This thesis is based on the project of the Fire and Rescue Bureau of the Ministry of Emergency Management-the research and development of metal discoloration trace recognition and intelligent extraction equipment in narrow space.Through the actual fire situation and related experiments,the color performance and microstructure changes of copper wires under different fire and annealing conditions are summarized,and the corresponding recognition system is established to provide tools for fire investigation,update fire investigation mode,improve fire investigation efficiency,promote the development of fire accident cause identification,and provide scientific reference for fire evidence identification.The main research work and achievements are as follows :(1)The color change trend and macroscopic performance of copper at high temperature were obtained by theoretical analysis and experimental verification.Through the energy band theory and the principle of high temperature oxidation of substances,the oxidation products and coloration mechanism of copper at high temperature are analyzed and studied.The composition of oxidation products of copper at different temperatures is analyzed,and the causes of coloration of copper and its oxides are explained.The coloration of copper at different temperatures is obtained.The experimental platform of copper conductor fire heating is built.The macroscopic performance of copper conductor fire heating is summarized and analyzed from the surface color,oxide layer coverage and insulation layer damage.For the conductor without insulation layer,high temperature will cause different thickness and different composition oxide layers on the surface,thus showing different colors.For wires with insulating layer,high temperature will lead to different damage performance of PVC insulating layer,so that the copper core has different degrees of bare leakage and oxidation.The oxide layer thickness of copper wire without insulating layer increases first and then decreases with the increase of temperature,and reaches the maximum at500 °C.(2)Through theoretical analysis and experimental verification,the growth trend of copper wire grain size with the increase of fire temperature is obtained.The relationship between metal grain size and temperature is studied theoretically.It is found that it is related to nucleation rate and growth rate,and undercooling is an important factor of metal nucleation rate and grain growth rate.By combining the relationship between undercooling and metal heating temperature,the growth trend of metal grain size with the increase of fire temperature is obtained.The metallographic analysis experimental platform was built to analyze the microscopic performance of copper wire fire heating from metallographic structure,grain size,maximum grain related parameters and grain area distribution.The average grain diameter and maximum grain related parameters of copper wire gradually increase with the increase of temperature,and the overall trend is exponential,which matches the theoretical research results.Water cooling will affect the grain size of copper wires,resulting in smaller grains.Metallographic observation showed obvious damage,with multiple cracks of different sizes and widths.The grain area distribution changes with the increase of temperature,and the distribution area gradually moves backward.When the temperature is greater than 200 °C,the grain area distribution generally shows a normal distribution.(3)Based on the deep learning YOLOv5 algorithm,the image recognition function of copper wire heating conditions is realized.The image of copper wire fire condition collected by the experiment is screened and extracted,and the image is manually labeled.After data enhancement and other processing,the labeled samples are finally obtained and the data set of copper wire fire condition is made.The insulation layer and fire condition of copper wire are classified and processed by classification network.The effect of the classification model is evaluated by the test set,and the classification accuracy training curve of different rounds is drawn,which realizes the classification of the fire condition of the copper wire by deep learning.