A rotating machinery friction fault detection method based on strain test technology

Abstract

The application discloses a rotating machinery friction fault detection method based on a strain test technology, relates to the technical field of friction fault detection, and solves the technical problem of low recognition accuracy of friction faults, and the technical solution is characterized in that the strain test technology is used to convert a bending strain signal into a voltage signal, and the interference of other factors in the vibration signal is avoided. The strain technology has high precision and can reflect extremely tiny thermal bending deformation, and can identify the friction in the early stage. The bending strain of the rotor in the running state is collected, and the change trend is observed, and when periodic fluctuation or divergence occurs, the occurrence of the friction fault can be judged. Meanwhile, the friction range can be roughly judged according to the strain values of different sections, and the diagnosis reference for the rotor friction fault treatment is provided.

Description

Technical Field

[0001] This application relates to the field of friction fault detection technology, and in particular to a method for detecting friction faults in rotating machinery based on strain testing technology. Background Technology

[0002] Rotating machinery, as a key piece of equipment in national infrastructure and industry, plays an increasingly important role in sectors such as power, machinery, and aviation. With the trend towards larger units and increasingly demanding operating parameters, the performance requirements for these units are also rising. To improve efficiency, the dynamic and static clearances of rotating machinery are often designed to be very small, thus increasing the possibility of dynamic and static friction. After friction occurs, the rotor is simultaneously affected by force and thermal shock effects. The frictional force causes the rotor to produce complex nonlinear motion patterns. Frictional thermal shock brings heat flow, creating a non-uniform temperature field at the friction cross-section, causing rotor thermal deformation, leading to periodic fluctuations or divergent vibrations, and affecting the safe operation of the unit.

[0003] In the existing technology, the detection techniques for friction faults include: (1) Spectrum analysis technology, which uses fast Fourier transform to perform spectrum analysis on vibration signals and identify dynamic and static friction characteristics. However, this technology is mainly suitable for analyzing stationary signals. For non-stationary signals, the important features of the signal may be averaged during the Fourier transform process, resulting in feature loss. (2) Wavelet analysis technology, which uses wavelet transform waveform decomposition to process the dynamic and static friction vibration signals of steam turbine generator sets to achieve friction diagnosis. However, the wavelet basis is not adaptable to the local signal, and it is difficult to select the optimal wavelet basis for a specific signal. It is also difficult to distinguish non-stationary signals. (3) Acoustic emission technology, which is often accompanied by strong acoustic emission signals when friction occurs. By identifying the signal and extracting features, the detection and diagnosis of friction faults can be achieved. However, acoustic emission technology is sensitive to material properties and is easily interfered with by various noises. At the same time, the attenuation and reflection of waves during propagation will also disturb the original signal emitted by the acoustic emission source, affecting the identification of the acoustic emission source.

[0004] Therefore, there is an urgent need for a new method for detecting friction faults in rotating machinery to overcome the problems existing in the current technology. Summary of the Invention

[0005] This application provides a method for detecting friction faults in rotating machinery based on strain testing technology. The technical purpose of this method is to avoid the feature identification of complex rotor vibration signals and realize the detection of friction faults.

[0006] The above-mentioned technical objective of this application is achieved through the following technical solution:

[0007] A method for detecting frictional faults in rotating machinery based on strain testing technology includes:

[0008] S1: Determine the measuring points on the rotating machinery, and symmetrically attach the strain gauge structure to the upper and lower sides of the measuring point section;

[0009] S2: Fix a wireless strain signal node near the measuring point section. The input end of the wireless strain signal node is connected to the lead wire of the strain gauge structure, and the output end is connected to the controller.

[0010] S3: When the rotating machinery is in the low-speed turning gear state before startup, the controller collects the signals transmitted by the wireless strain signal nodes to obtain the bending strain signal ε of each section in each rotation cycle. i , then ε i =ε0; where i represents the section of the i-th measuring point, and ε0 represents the initial value of the bending strain signal under the low-speed turning state before startup;

[0011] S4: When the rotating machinery is in a high-speed rotation state, the controller collects the signals transmitted by the wireless strain signal nodes to obtain the bending strain signal ε of each section in each rotation cycle. t and the bending moment value M of the section;

[0012] S5: Monitor the change in the effect of thermal effects on rotor bending during the transition of rotating machinery from low-speed turning gear to high-speed rotation, expressed as: Δε=ε t -ε0;

[0013] S6: When Δε shows periodic fluctuations or divergence, the rotating machinery has a friction fault. The bending moment value M of each section is obtained, and the location of friction is determined by the bending moment value M.

[0014] Furthermore, the method also includes:

[0015] S7: Upon observing a friction fault, the rotating machinery stops high-speed rotation and slows down to a low-speed turning gear state. The controller collects signals transmitted from the wireless strain signal nodes to obtain the bending strain signal ε′ of each section in each rotation cycle. i , will ε′ i The bending strain signal ε under high-speed rotation t A comparison was made to eliminate the influence of self-weight and imbalance.

[0016] Furthermore, the strain gauge structure is a full-bridge circuit, which is a four-arm bridge circuit, where R1, R2, R3, and R4 are all resistance strain gauges. The output voltage is then expressed as:

[0017]

[0018] Among them, K s U represents the sensitivity coefficient;out Indicates the output voltage; U in Indicates input voltage; ε e ε represents the strain of a component caused by changes in ambient temperature; F The values ​​represent the strain produced when a component is subjected to force; ε1, ε2, ε3, and ε4 represent the strain on the surface of the component felt by R1, R2, R3, and R4, respectively.

[0019] Furthermore, the bending moment value M of the section is expressed as:

[0020]

[0021] Where D represents the outer diameter of the measuring point section; d represents the inner diameter of the measuring point section; E represents the elastic modulus of the material; S 180° S represents the strain signal value when the strain gauge structure is rotated to the downward position. 0° This indicates the strain signal value when the strain gauge structure is rotated to the top.

[0022] Furthermore, in step S6, when the shaft of the rotating machinery bulges downwards, the bending moment value M of the cross section is positive; when the shaft of the rotating machinery bulges upwards, the bending moment value M of the cross section is negative.

[0023] Furthermore, the bending strain signal ε of each section during each rotation cycle is expressed as:

[0024]

[0025] Among them, S 180° S represents the strain signal value when the strain gauge structure is rotated to the downward position. 0° This indicates the strain signal value when the strain gauge structure is rotated to the top.

[0026] The beneficial effects of this application are as follows: The rotating machinery friction fault detection method based on strain testing technology described in this application differs from traditional processing methods that extract features from vibration signals. It uses strain testing technology to convert bending strain signals into voltage signals, avoiding interference from other factors in the vibration signal. Strain testing technology offers high precision, capable of detecting extremely minute thermal bending deformations, allowing for identification at the early stages of friction. By collecting bending strain data during rotor operation and observing its changing trends, the occurrence of friction faults can be determined when periodic fluctuations or divergent phenomena appear. Furthermore, the range of friction locations can be roughly determined based on the strain values ​​at different cross-sections, providing a diagnostic reference for rotor friction fault management. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of a full-bridge circuit;

[0028] Figure 2 A schematic diagram of a specific implementation scheme for wireless strain testing technology;

[0029] Figure 3 A schematic diagram for collecting bending strain signals at the measuring point cross section;

[0030] Figure 4 A schematic diagram showing the strain symbols at different rotational positions. Detailed Implementation

[0031] The technical solution of this application will be described in detail below with reference to the accompanying drawings.

[0032] The rotating machinery friction fault detection method based on strain testing technology described in this application includes:

[0033] S1: Determine the measuring point on the rotating machinery, and symmetrically attach the strain gauge structure to the upper and lower sides of the measuring point section.

[0034] Specifically, when determining the measuring points, select the exposed parts of the shaft and select as many measuring point sections as possible within the limits of available conditions.

[0035] The strain gauge structure is a full-bridge circuit, such as Figure 1 As shown, this full-bridge circuit is a four-arm bridge, with R1, R2, R3, and R4 all being resistance strain gauges. While eliminating errors caused by changes in ambient temperature, it also possesses high sensitivity. Therefore, the output voltage is expressed as:

[0036]

[0037] Among them, K s U represents the sensitivity coefficient; out Indicates the output voltage; U in Indicates input voltage; ε e ε represents the strain of a component caused by changes in ambient temperature; F The values ​​represent the strain produced when a component is subjected to force; ε1, ε2, ε3, and ε4 represent the strain on the surface of the component felt by R1, R2, R3, and R4, respectively.

[0038] Before attaching the strain gauges, check that the resistance values ​​are correct and that there are no air bubbles, rust spots, or other defects. Before attaching, accurately position the cross-section and mark the angles. Grind away any dirt and paint from the shaft surface and select a smooth area for attachment. After attachment, ensure proper alignment and a tight fit. Once everything is confirmed to be correct, secure the strain gauges and leads and protect them from moisture.

[0039] S2: Fix a wireless strain signal node near the measuring point section. The input end of the wireless strain signal node is connected to the lead wire of the strain gauge structure, and the output end is connected to the controller.

[0040] Specifically, the wireless strain testing scheme is as follows: Figure 2As shown, strain gauges are symmetrically arranged on the upper and lower sides of the measuring point section of the rotating shaft. Wireless strain signal nodes are fixed near the measuring point section and connected to the strain gauge leads. The wireless strain signal nodes transmit signals to a remote computer (i.e., the controller), thereby realizing the measurement of bending strain during shaft operation.

[0041] S3: When the rotating machinery is in the low-speed turning gear state before startup, the controller collects the signals transmitted by the wireless strain signal nodes to obtain the bending strain signal ε of each section in each rotation cycle. i , then ε i =ε0; where i represents the measuring point section and ε0 represents the initial value of the bending strain signal under low-speed turning state before startup.

[0042] like Figure 3 As shown, strain signals were collected for each section. Since the strain signal has an initial value ε0 under the low-speed turning state before startup, the existence of this initial value ε0 is due to the initial load of the shaft's own weight.

[0043] S4: When the rotating machinery is in a high-speed rotation state, the controller collects the signals transmitted by the wireless strain signal nodes to obtain the bending strain signal ε of each section in each rotation cycle. t And the bending moment value M of the section.

[0044] Specifically, according to the definition of the sign of the bending moment of a cross section in mechanics of materials: when the shaft is convex downwards, the bending moment of the cross section is positive; when the shaft is convex upwards, the bending moment of the cross section is negative. Simultaneously, when the shaft is convex downwards, the upper surface of the shaft experiences compressive stress and negative strain, while the lower surface experiences tensile stress and positive strain.

[0045] like Figure 4 As shown, the strain gauge structure is attached to point A. When point A is directly above, the rotation axis is at 0°, and the strain output is negative. When point A is directly below, the rotation axis is at 180°, and the strain output is positive.

[0046] Therefore, assuming the measured strain signal values ​​at 0° and 180° are S 0° and S 180° To align with the sign of the bending moment value at the cross section, the strain value (i.e., the bending strain signal) of that cross section is defined as:

[0047]

[0048] According to Hooke's Law, the formulas for calculating the bending stress and bending moment of a cross-section can be obtained as follows:

[0049] σ = E·ε;

[0050] M = W·σ;

[0051] Where E represents the material's elastic modulus and W represents the section bending coefficient.

[0052] The shaft system of a steam turbine generator set is usually a circular cross-section, therefore the bending moment value M can be derived as follows:

[0053]

[0054] Where D represents the outer diameter of the measuring point section; d represents the inner diameter of the measuring point section.

[0055] S5: Monitor the change in the effect of thermal effects on rotor bending during the transition of rotating machinery from low-speed turning gear to high-speed rotation, expressed as: Δε=ε t -ε0.

[0056] The strain signal at this moment is subtracted from the initial average strain ε0 under low-speed turning conditions to eliminate the influence of factors such as gravity on subsequent measurements and calculations.

[0057] S6: When Δε shows periodic fluctuations or divergence, the rotating machinery has a friction fault. The bending moment value M of each section is obtained, and the location of friction is determined by the bending moment value M.

[0058] Specifically, when the unit is running smoothly, the bending strain of each cross-section remains constant. When a friction fault occurs, the rotor undergoes bending deformation due to thermal shock, causing changes in the bending strain of each cross-section, which then exhibits periodic fluctuations or divergence over time. Based on this characteristic, friction faults can be detected by observing the trend of bending strain changes over time. Furthermore, the range of friction locations can be determined based on the strain values ​​of different cross-sections, providing a basis for troubleshooting and diagnostic reference for rotor friction faults.

[0059] In a specific embodiment, the method further includes:

[0060] S7: Upon observing a friction fault, the rotating machinery stops high-speed rotation and slows down to a low-speed turning gear state. The controller collects signals transmitted from the wireless strain signal nodes to obtain the bending strain signal ε′ of each section in each rotation cycle. i , will ε′ i The bending strain signal ε under high-speed rotation t A comparison was made to eliminate the influence of self-weight and imbalance.

[0061] By collecting strain signals during low-speed turning before and after rotor operation, interference from factors such as self-weight and imbalance is eliminated.

[0062] The above are exemplary embodiments of this application, and the scope of protection of this application is defined by the claims and their equivalents.

Claims

1. A method for detecting frictional fault of a rotating machine based on strain test technique, characterized by, The method comprises the following steps: S1: determining a measuring point on the rotating machine, and symmetrically sticking strain gauge structures on upper and lower sides of the measuring point section; S2: fixing a wireless strain signal node near the measuring point section, an input end of the wireless strain signal node being connected with a lead wire of the strain gauge structure, and an output end of the wireless strain signal node being connected with a controller; S3: When the rotating machinery is in the low-speed turning state before starting, the controller collects the signals transmitted by the wireless strain signal nodes to obtain the bending strain signals ε of each section in each rotation period i Then ε i = ε0; wherein i represents the i-th measuring point section, and ε0represents the initial value of the bending strain signal in the low-speed turning state before starting. S4: When the rotating machine is in a high-speed rotating state, the signals transmitted by the wireless strain signal nodes are collected by the controller to obtain the bending strain signals ε of each section in each rotating period t and the section bending moment values M; S5: Monitoring the change of the influence of thermal effects on rotor bending during the process of rotating machinery from low-speed cranking state to high-speed rotating state, expressed as: Δε = ε - ε0 t - ε0; S6: when periodic fluctuation or divergence of Δε occurs, it is determined that a friction fault occurs in the rotating machine, and a section bending moment value M of each section is obtained to determine a friction occurrence position according to the section bending moment value M; The strain gauge structure is a full-bridge circuit, the full-bridge circuit is a four-bridge-arm electric bridge, R1, R2, R3 and R4 are all resistance strain gauges, and the output voltage is represented as: wherein K s represents the sensitivity coefficient; U out represents the output voltage; U in represents the input voltage; ε e represents the strain of the component due to the change in ambient temperature; ε F represents the strain of the component after being subjected to force; ε1, ε2, ε3, ε4 respectively represent the strain of the surface of the component sensed by R1, R2, R3, R4. The section bending moment value M is represented as: wherein D represents the outer diameter of the cross section of the measurement point; d represents the inner diameter of the cross section of the measurement point; E represents the elastic modulus of the material; S 180° represents the strain signal value when the strain gauge structure is turned to the positive upside, S 0° represents the strain signal value when the strain gauge structure is turned to the positive upside.

2. The method of claim 1, wherein, The method further comprises the following steps: S7: After observing the rubbing fault, the rotating machine stops high-speed rotation and reduces to low-speed rotation, the controller collects the signals transmitted by the wireless strain signal nodes, and obtains the bending strain signals ε' of each cross section in each rotation period i , and compares the bending strain signals ε' i and the bending strain signals ε t in the high-speed rotation state to exclude the effects of self-weight and unbalance factors.

3. The method of claim 1, wherein, In step S6, when the rotating shaft of the rotating machine is concave, the section bending moment value M is positive; and when the rotating shaft of the rotating machine is convex, the section bending moment value M is negative.

4. The method of claim 1, wherein, The bending strain signal ε of each section in each rotating cycle is represented as: where S 180° represents a strain signal value when the strain gauge structure turns to the positive upside, S 0° represents a strain signal value when the strain gauge structure turns to the positive upside.

Images