Search

CN-122021304-A - P91 pipeline aging prediction method

CN122021304ACN 122021304 ACN122021304 ACN 122021304ACN-122021304-A

Abstract

The invention relates to a P91 pipeline aging prediction method, and belongs to the field of P91 pipeline aging prediction. The method comprises the steps of obtaining micro magnetic signals (such as Barkhausen noise) of the P91 pipeline through a micro magnetic detector, extracting characteristic parameters related to the aging state, inputting the parameters into a pre-trained aging level prediction model, and directly outputting the aging level (1-5 levels) of the pipeline. The predictive model is built based on a machine learning algorithm (e.g., reliefF feature selection and BP neural network), and the training data is derived from accelerated aging specimens simulated by specific heat treatments, with well-defined microscopic organization (characterized by EBSD) and macroscopic mechanical property associations. The invention realizes the rapid, lossless and quantitative prediction of the aging state of the P91 pipeline, and provides a reliable basis for the safety evaluation and residual life prediction of the in-service pipeline.

Inventors

  • SUN YUNFEI
  • ZHANG YANFEI
  • XIE LIMING
  • TAN XIAOMENG
  • ZHANG TAO
  • LV LEI
  • YUN FENG

Assignees

  • 内蒙古电力(集团)有限责任公司内蒙古电力科学研究院分公司

Dates

Publication Date
20260512
Application Date
20260129

Claims (10)

  1. 1. A method for predicting aging of a P91 pipeline, comprising: Acquiring a micro-magnetic signal of a P91 pipeline, wherein the micro-magnetic signal comprises at least one of tangential magnetic field intensity, barkhausen noise, incremental magnetic permeability and multi-frequency eddy current signals; Extracting characteristic parameters from the micro-magnetic signals, wherein the characteristic parameters comprise at least one of Barkhausen noise maximum value BN_Max, barkhausen noise average value BN_mean, barkhausen noise remanence BN_Mr and Barkhausen noise peak width BN 75; the characteristic parameters are input into a pre-trained aging level prediction model, and the aging level of the P91 pipeline is output, wherein the aging level is divided into 1 level to 5 level, the 1 level represents unaged, and the 5 level represents completely aged.
  2. 2. The method for predicting aging of a P91 pipeline according to claim 1, wherein the micro magnetic signals are obtained under stress load, the stress load range covers an elastic stage and a plastic stage of the P91 pipeline, and a stepping tensile test method is adopted when the micro magnetic signals are obtained, and the micro magnetic signals are kept and measured under a preset stress level.
  3. 3. The P91 pipeline aging prediction method according to claim 1, wherein the aging level prediction model is obtained by training based on a machine learning algorithm, the machine learning algorithm comprises ReliefF feature selection algorithm and BP neural network model, the ReliefF algorithm is used for screening micro-magnetic feature parameters with high correlation with the aging level, and the BP neural network is used for constructing a nonlinear mapping relation between the micro-magnetic feature parameters and the aging level.
  4. 4. The method of claim 1, wherein the training data of the aging level prediction model is derived from an accelerated aging test that simulates different aging states of the P91 pipe by heat treatment, and the heat treatment parameters include: Stage 1 aging, namely, no heat treatment; 2-stage aging, namely carrying out furnace cooling after heat preservation for 180min at a high-temperature normalizing temperature of 810 ℃; 3-stage aging, namely, furnace cooling after heat preservation for 90min at the high temperature of normalizing 820 ℃; 4-stage aging, namely, carrying out furnace cooling after heat preservation for 160min at a high-temperature normalizing temperature of 820 ℃; 5-stage aging, namely normalizing 1055 ℃ to +/-15 ℃ and preserving heat for 130min, and then cooling to 740 ℃ and preserving heat for 480 min.
  5. 5. The method of claim 1, wherein the aging level is associated with a room temperature tensile strength threshold of the P91 pipe, wherein: the tensile strength corresponding to the 1-level aging is more than or equal to 650MPa; the tensile strength corresponding to the 2-grade aging is between 610MPa and 650 MPa; the 3-stage aging corresponds to the tensile strength of 585MPa to 610 MPa; The tensile strength corresponding to the 4-grade aging is between 550MPa and 585 MPa; The 5-grade aging corresponds to the tensile strength of less than or equal to 550MPa.
  6. 6. The method of claim 1, further comprising obtaining microstructure parameters of the P91 pipe by electron back scattering diffraction EBSD technique, wherein the microstructure parameters include at least one of average value, area weighted average value, maximum value, and average deviation of equivalent circle diameters of grains, and the method is used for assisting in verifying or calibrating the aging level prediction result.
  7. 7. The method of claim 6, wherein the detection parameters of the EBSD technique include 1021×765 grating dimensions, 0.25 μm step size, and 10.0 ° threshold angle, and the correspondence between the microstructure parameters and the aging level is: grade 1 aging, wherein the average diameter of the equivalent circle of the crystal grain is 2.4 mu m, the area weighted average value is 5.5 mu m, the maximum value is 16.4 mu m, and the average deviation is 1.8 mu m; 2-stage aging, namely, the average diameter of equivalent circles of crystal grains is 2.6 mu m, the area weighted average value is 6.1 mu m, the maximum value is 20.3 mu m, and the average deviation is 2.0 mu m; 3-stage aging, namely 2.8 μm of average diameter of equivalent circle of crystal grain, 10.3 μm of area weighted average value, 27.5 μm of maximum value and 2.9 μm of average deviation; 4-stage aging, namely, the average diameter of equivalent circles of crystal grains is 4.3 mu m, the average weighted average value of areas is 18.4 mu m, the maximum value is 48.1 mu m, and the average deviation is 5.1 mu m; 5-stage aging, average diameter of equivalent circle of crystal grain is 4.9 μm, average weighted area is 26.4 μm, maximum value is 71.1 μm, and average deviation is 6.5 μm.
  8. 8. The P91 pipeline aging prediction method according to claim 1, further comprising performing safety assessment on high temperature mechanical properties of the P91 pipeline based on a weibull reliability model, wherein a statistical distribution of high temperature tensile strength or yield strength is used for calculating failure probability, the failure probability is associated with an aging grade, the safety grade is from grade 1 to grade 5, grade 1 is safest, and grade 5 is safest.
  9. 9. The method for predicting aging of the P91 pipeline according to claim 1, wherein the micromagnetic signal acquisition uses a multifunctional micromagnetic nondestructive detector, the detection position is located in an axial area of the P91 pipeline, the influence of welding seams is avoided, and the surface of the pipeline is cleaned before detection to ensure the fitting degree of the probe.
  10. 10. A computer readable storage medium having stored thereon computer program instructions executable by a processor, which when executed by the processor are capable of carrying out the steps of the method according to any one of claims 1-9.

Description

P91 pipeline aging prediction method Technical Field The invention belongs to the field of P91 pipeline aging prediction, and particularly relates to a P91 pipeline aging prediction method. Background P91 steel (10 Cr9Mo1 VNbN) is used as a martensitic heat-resistant steel, and is widely applied to high-temperature and high-pressure key components such as main steam pipelines, headers and the like of supercritical and ultra-supercritical thermal power units due to excellent high-temperature strength, creep resistance and oxidation resistance. The parts operate in high temperature (generally 535-538 ℃) and high pressure environment for a long time, the materials inevitably undergo aging and evolution of microstructures, so that the mechanical properties gradually decline, and the safe operation life of the unit is further threatened. At present, for the aging state evaluation of a P91 steel pipeline, the industry is greatly dependent on methods specified by standards such as DL/T884-2019 thermal power plant metallographic examination and evaluation technical guidelines, DL/T2219-2021 10Cr9Mo1VNbN steel microstructure aging rating for thermal power plants and the like. The method is mainly based on metallographic microscope observation, and the aging degree is classified into different grades from grade 1 (unaged) to grade 5 (severely aged) by qualitatively analyzing the characteristics of the shape integrity of martensite laths, the dispersion degree of lath positions, the size, distribution and spheroidization/chaining states of carbides (such as M 23C6 and MX phases) and the like. However, there are several significant limitations to the conventional metallographic evaluation methods described above: the subjectivity is strong, the judgment of the aging grade is highly dependent on the experience and subjective judgment of the inspector, and the interpretation of the same microstructure among different operators can be different, so that the consistency and the repeatability of the evaluation result are difficult to ensure. The quantitative method is insufficient, the evaluation standard is mainly descriptive language, the accurate quantitative index support is lacking, and the aging degree in the critical state or the fine change is difficult to accurately distinguish and predict the trend. The mechanism association is weak, although the standard describes the change of macroscopic tissue morphology, the lack of direct and quantitative association with the micro-dynamics mechanism (such as carbide dissolution, aggregation, coarsening and grain boundary evolution caused by oswald ripening) behind material aging limits the deep understanding of the cause of performance degradation from the mechanism level. The traditional method belongs to lossy detection, pipeline samples need to be intercepted, and the requirements of online nondestructive detection and life prediction of in-service pipelines cannot be met. In recent years, some non-destructive testing techniques, such as micromagnetic testing techniques, have been attempted for material property assessment due to their sensitivity to material microstructure states and stress changes. The micromagnetic signals (such as Barkhausen noise, tangential magnetic field strength harmonic waves and the like) contain abundant material microstructure information. Meanwhile, advanced microscopic characterization techniques such as Electron Back Scattering Diffraction (EBSD) can provide accurate quantitative data such as grain orientation, grain size (e.g., average of equivalent circle diameters, area weighted average, etc.), grain boundary type, distribution, etc. Although the EBSD technology is applied in the material science research, how to establish a reliable and systematic corresponding relation between the obtained quantized microscopic parameters and the macroscopic aging level of the P91 steel, and further correlate the quantized microscopic parameters with a micromagnetic detection signal which can be applied on site, and construct a complete aging state evaluation system from microscopic quantitative characterization to macroscopic lossless prediction is still a key problem to be solved in the technical field. Therefore, the method for quickly, nondestructively and accurately predicting the aging state of the P91 pipeline can be developed, overcomes the defects of strong subjectivity and quantification deficiency in the prior art, combines a microscopic mechanism and macroscopic performance, and has important engineering practical significance for guaranteeing safe and economic operation of the thermal power generating unit. Disclosure of Invention The invention aims to provide a P91 pipeline aging prediction method, which realizes the rapid, lossless and quantitative prediction of the P91 pipeline aging state and provides a reliable basis for the safety evaluation and residual life prediction of in-service pipelines. In order to achieve the purpose, the t