Research On Plastic Damage Assessment Technology Of X80 Pipeline Steel Based On Acoustic Emission

Oil and gas pipelines inevitably pass through polar regions,oceans and other geologically unstable regions,so their working environment as well as stress modes are complex and changeable.As one of the main failure forms of pipeline steel,plastic damage threatens the safe operation of oil and gas pipelines seriously.Therefore,it is urgent and necessary to monitor the healthy service state of pipeline steel effectively and evaluate its remaining life accurately.Acoustic emission technology,as the most effective means of dynamic monitoring,can realize material damage assessment by analyzing the damage information,which carried by the stress acoustic wave released when the internal structure of the material changes.In this paper,the widely used bainite + polygonal ferrite(B+PF)dual-phase structure X80 pipeline steel is taken as the research object.The tensile,shear and tension-shear mixed damage specimen taken as the test sample are designed,and the material’s damage factors are defined,different damage states have different damage factors.Acoustic emission on-line monitoring method was used during the test.characteristic parameter and waveform analysis was used to analyze the acoustic emission signals generated in the test process.The failure mode and damage evolution of X80 steel were studied.The research conclusions are as follows:1.The tensile and shear damage process of X80 steel was analyzed by the acoustic emission characteristic parameter analysis method,and it was found that the time domain changes of characteristic parameters such as acoustic emission amplitude,accumulated count and accumulated energy were similar to the deformation damage stage of X80 steel under single damage mode.Correspondingly,the research results provide a basis for determining the service load state of metal materials.2.The acoustic emission waveform signal collected during the test was analysed and the signal was denoised using the wavelet threshold denoising method.The denoised signal retained most of the detailed components of the original signal while removing clutter in a large area,and the signal-to-noise ratio and root mean square error were within the optimal range.The wavelet packet decomposition is applied to extract the energy values of different frequency bands as the eigenvalues of the signal.3.Combined with the energy dissipation model,the damage factor of X80 steel in a single damage mode was defined according to the damage evolution curve of X80 steel in tensile and shear states,and the damage degree of the material was quantified.The quantitative relationship between the acoustic emission signal waveform characteristics and the damage factor was established using a multi-layer perceptron network,and the monitoring and evaluation of the plastic damage process of X80 steel was achieved within an allowable error of 5%.4.The machine learning method of random forest classification was used to complete the classification of acoustic emission sources in the mixed damage mode of X80 steel pulling and shearing with an accuracy of 96.02%,and multi-layer perceptron network model was used to achieve the assessment of the degree of plastic damage of the classified signals with an overall error within 8.5%.The results provide a means of monitoring the extent of material damage in real time,identifying the sources of material damage,quantitatively assessing the service condition of pipelines and predicting their remaining life.