Quantitative Detection Of Fatigue Cracks And Stress Research Of Steel Wire Rope Based On The Parameters Of The Magnetic Properties

Wire rope as a important bearing tool widely used in many fields of nationalproduct.Wire rope safety condition for people’s life and property safety is the vital. Theferromagnetic material wire rope under a variety of uncertainty in the process of work loadand bad work environment easily lead to cross section of wear, corrosion, fatigue andoverload damage. In the event of rupture it caused immeasurable loss to people’s life andproperty security. For these injuries, artificial detection has a large labor intensity andblindness, damage degree is influenced by subjective factors of faults, easy to cause wrongtargets. At present, although there are a few departments use wire rope testing instrument, butthey are widespread, such as low detecting precision and intelligence degree of difference,detection signal of a single, quantitative of damage is not accurate, and can not be real-timeonline monitoring of the stress. Therefore research of intelligent and efficient wire ropedamage and stress the quantitative technique is necessary.In this paper, based on the magnetic properties of ferromagnetic materials, therelationship between the parameters and the qualitative change of fatigue crack is analyzedand the influence of tensile stress on the material magnetic. Fatigue and stress testingsoftware/hardware system is designed, including according to the experiment request thepreparation of test specimens; For commonly used excitation detection sensor simulationoptimization; Characteristics, signal acquisition and process design, and discussescharacteristic extraction method; Choose the reasonable excitation device, test structures,signal conditioning circuit, the signal preprocessing. In the end of this paper, we analyzed thecharacteristic data filtering, and choose quantitative identification method based on neuralnetwork, fatigue quantitative detection system is designed, the multivariate linear regressionanalysis between stress and magnetic parameters, stress and magnetic parameters of themultiple regression model is set up, and get the fitting curve.First of all, according to the principle of magnetism, from the micro magnetism ofatoms, theoretically expounds the origin of the magnetic material. To study the influences of different shape materials for magnetic field back. Analyzed the fatigue crack on internal localmagnetic field, the influence of ferromagnetic materials and crack leads to the formation of amagnetic field annealing process. At the same time, the internal stress of ferromagneticmaterials is deduced and the relation between permeability and coercive force. Determine thefatigue crack and stress is the important cause leading to the magnetic parameters. For theparameters of the magnetic properties of the wire rope damage and stress changes theorybasis.Second, design the magnetic parameters of ferromagnetic material fatigue damage andstress signal acquisition system. The commonly used form of excitation are simulated,determine the optimal structural form. Using LabVIEW program signal acquisition andfeature extraction. Characteristics of signals are extracted, and according to the characteristicsof the degree of differentiation, and screening reflect the characteristics of damagedconditions. Determine the coercive force, remanence, saturated magnetic induction intensityfor quantitative recognition characteristics.Finally, research the method of neural network theory and network algorithm. Analyzethe characteristics of wire rope and filter, keep on crack damage recognition features highamount. In the correct classification rate, the average absolute error, as the maindiscrimination on the basis of network training time, determine the number of hidden layernodes of neural network, the test set and the training set, the output layer node number.Established between stress and magnetic parameters of multiple regression model. Accordingto the parameters of magnetic signal acquisition, feature extraction, quantitative identificationof neural networks in sequentiality and inheritance, development procession of wire ropefatigue crack quantitative detection of virtual instrument and stress.