Method for imaging residual oil in powder metallurgy component based on pulse thermal excitation

Abstract

The invention belongs to the technical field of nondestructive testing, and relates to a method for imaging residual oil in a powder metallurgy component based on pulse thermal excitation, which comprises the following steps: applying preheating disturbance pulse to a target area to generate a pre-excitation thermal field; applying a main detection pulse to the same target area at an interval of a preset time lag, and generating a transient composite hot spot; obtaining the whole process from formation to cooling dissipation of the transient composite hot spot, and generating a dynamic thermal field image sequence; images in a specific time window are extracted, and a key stage hot spot evolution image set is generated; the physical state of residual oil in the target area is judged, and a residual oil state label is generated; performing mapping association on the residual oil state label generated at each detection position and the two-dimensional coordinate of the residual oil state label to generate a residual oil state visual image; according to the invention, the problem of missing detection or misjudgment caused by difficulty in effectively amplifying weak thermal physical property difference between oil liquids in different states due to single excitation mode in the existing thermal imaging technology is solved.

Description

Technical Field

[0001] This invention belongs to the technical field of nondestructive testing and relates to a method for imaging residual oil inside powder metallurgy parts based on pulsed thermal excitation. Background Technology

[0002] Currently, in the quality control and service condition assessment of powder metallurgy parts, it is necessary to detect the state of residual oil in their internal pores as accurately as possible. For example, during service, residual oil relies on continuous seepage to achieve self-lubrication. However, residual oil is prone to aging and forming gel-like polymers under high temperature / oxidation environments, leading to pore blockage and loss of lubrication function. The state of these residual oils, whether they are in a fresh liquid state to maintain lubrication performance or have aged and deteriorated into a gel that affects performance, directly affects the lifespan and reliability of the parts.

[0003] Because it exists within the internal structure of parts, the methods commonly used in the industry include gravimetric analysis, chemical solvent extraction, or simple visual inspection. Among these, gravimetric analysis can only determine the total amount of residual oil and cannot distinguish its physical state; although chemical extraction can analyze the oil composition, the process is destructive and cannot be used for online or batch testing of finished products. There are also some single-pulse thermal imaging detection methods in the existing technology, but they are difficult to distinguish between different oil states due to weak signals and low signal-to-noise ratios.

[0004] Existing thermal imaging technologies, due to their single excitation method, are unable to effectively amplify the subtle differences in thermophysical properties between oils in different states, resulting in insufficient sensitivity and specificity of the detection signal. They are also susceptible to interference from the inhomogeneity of the material itself and environmental noise, leading to missed detections or misjudgments. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides a method for imaging residual oil inside powder metallurgy parts based on pulsed thermal excitation.

[0006] A method for imaging residual oil inside powder metallurgy parts based on pulsed thermal excitation includes the following steps:

[0007] S1. Apply a preheating disturbance pulse to the target area on the surface of the powder metallurgy part to be tested to generate a pre-excitation thermal field carrying the initial response information of the internal oil.

[0008] S2. After a preset time delay, apply the main detection pulse to the same target area and superimpose it with the pre-excitation thermal field to generate a transient composite hot spot.

[0009] S3. A high-frequency infrared camera is used to simultaneously acquire the entire process of transient composite hot spots from formation to cooling and dissipation, generating a dynamic thermal field image sequence;

[0010] S4. Extract images within a specific time window after the end of the main probe pulse from the dynamic thermal field image sequence to generate a hot spot evolution atlas for key stages;

[0011] S5. Based on the morphological evolution of hot spots in the hot spot evolution map of key stages, determine the physical state of residual oil in the target area and generate residual oil state labels.

[0012] S6. Obtain the two-dimensional coordinates of each detection location, repeat steps S1 to S5, and map and associate the residual oil status label generated at each detection location with its two-dimensional coordinates to generate a visual image of the residual oil status.

[0013] A further aspect of the present invention generates a pre-excited thermal field, comprising the following steps:

[0014] A preheating disturbance pulse is applied to the target area and acts on the fresh liquid oil molecules in the pores inside the target area;

[0015] During the preheating disturbance pulse, a pre-excitation thermal field is formed based on the surface temperature distribution of the target area and the response state of the internal oil.

[0016] A further aspect of the present invention generates transient composite hot spots, comprising the following steps:

[0017] After the preheating disturbance pulse ends, wait for a preset time delay of 10 to 30 milliseconds;

[0018] A main detection pulse is applied to the target area to generate a transient thermal response based on the residual oil state, thus creating a transient composite hot spot.

[0019] A further aspect of the present invention generates a dynamic thermal field image sequence, comprising the following steps:

[0020] Before applying the main detection pulse, the high-frequency infrared camera is activated to acquire images of the target area;

[0021] Starting from the injection of the main probe pulse, the two-dimensional spatial distribution of the surface temperature field of the target area and its changes over time are continuously recorded to form a dynamic thermal field image sequence that includes information on the evolution of hot spot brightness, morphology and size.

[0022] A further aspect of the present invention involves generating a hotspot evolution atlas for key stages, comprising the following steps:

[0023] Identify the termination time point corresponding to the end of the main probe pulse in the dynamic thermal field image sequence;

[0024] Starting from the termination time, select a time window with a duration of 5 to 50 milliseconds.

[0025] All infrared image frames captured within the time window were extracted and combined to form a hotspot evolution atlas for key stages.

[0026] A further aspect of the present invention involves determining the physical state of residual oil within a target area, including the following steps:

[0027] If the hot spot evolution atlas at the key stage shows that the difference in brightness decay rate between the hot spot center region and the edge region is less than the first preset threshold, then the residual oil status label will be determined as fresh liquid.

[0028] If the hot spot evolution atlas at the key stage shows that the brightness of the central region of the hot spot is higher than that of the edge region, and the difference in the brightness decay rate between the two is greater than the second preset threshold, then the residual oil status label will be determined as aged gum.

[0029] A further aspect of the present invention generates a visualization image of the residual oil state, comprising the following steps:

[0030] The control platform moves along a preset scanning path to cover the surface to be tested;

[0031] Repeat steps S1 to S5 for each point on the scanning path to obtain the two-dimensional coordinates of each point and its corresponding residual oil status label.

[0032] Based on the two-dimensional coordinates of each point, a preset color corresponding to the residual oil status label is used to mark it, generating a visual image of the residual oil status.

[0033] In a further aspect of the present invention, the pulse width of the preheating disturbance pulse is set based on a preset energy threshold that is sufficient to activate fresh liquid oil molecules but not enough to change the physical morphology of aged gum oil.

[0034] A further aspect of the present invention includes the following steps before generating the residual oil status label:

[0035] For each frame of the hotspot evolution atlas at the key stages, the average brightness of the hotspot center region and the average brightness of the edge region are calculated respectively.

[0036] Based on the calculated average brightness, generate brightness attenuation curves for the central region and edge region.

[0037] The difference in the decay rate between the generated brightness decay curve of the central region and the brightness decay curve of the edge region is used as the basis for judging the physical state of residual oil in the target region.

[0038] A further aspect of the present invention includes the following steps in determining the preset energy threshold:

[0039] Based on fresh liquid oil samples and aged gum oil samples, differential scanning calorimetry was used to test and obtain the heat flow-temperature curves of fresh liquid oil and aged gum oil.

[0040] Under the same detection environment, preheating perturbation pulses were applied to fresh liquid oil samples and aged gum oil samples to obtain pre-excited thermal field images.

[0041] By comparing and analyzing the pre-excitation thermal field response under different energy pulses, the preset energy threshold is obtained.

[0042] In summary, the present invention has the following beneficial technical effects:

[0043] 1. This invention improves the sensitivity and signal-to-noise ratio of detecting residual oil state inside powder metallurgy parts by using two pulse excitations and dynamic thermal field analysis. The preheating disturbance pulse first pre-processes the internal oil, selectively activating liquid oil molecules, which amplifies the difference in transient thermal response excited by the subsequent main detection pulse, thereby obtaining a clearer and more distinguishable feature signal on the thermal field image than that obtained by single pulse excitation.

[0044] 2. By analyzing the dynamic evolution pattern of hot spot morphology within a specific time window after the high-energy pulse ends, the thermophysical process is transformed into identifiable visual features. Based on the different manifestations of residual oil in different states during the initial stage of thermal diffusion, the interference caused by static factors such as uneven emissivity of the material surface is reduced, and the accuracy of the judgment results is improved.

[0045] 3. By scanning or array-based detection, the residual oil status labels obtained point by point are associated with their precise two-dimensional coordinates on the parts, and finally a visual image of the residual oil status is generated. This transforms the invisible internal defect information into an intuitive image, identifies the location, range and shape of the oil aging area, and provides a basis for further processing for subsequent maintenance, rework or process improvement. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. The drawings are used to provide a further understanding of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 A flowchart illustrating an embodiment of this application is disclosed.

[0048] Figure 2 A schematic diagram of module connections in an embodiment of this application is disclosed. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0050] The following is in conjunction with the appendix Figure 1 - Figure 2 A preferred description of the present invention is provided below.

[0051] See attached document Figure 1 This invention proposes a method for imaging residual oil inside powder metallurgy parts based on pulsed thermal excitation, comprising the following steps:

[0052] S1. Apply a preheating disturbance pulse to the target area on the surface of the powder metallurgy part to be tested to generate a pre-excitation thermal field carrying the initial response information of the internal oil.

[0053] S2. After a preset time delay, apply the main detection pulse to the same target area and superimpose it with the pre-excitation thermal field to generate a transient composite hot spot.

[0054] S3. A high-frequency infrared camera is used to simultaneously acquire the entire process of transient composite hot spots from formation to cooling and dissipation, generating a dynamic thermal field image sequence;

[0055] S4. Extract images within a specific time window after the end of the main probe pulse from the dynamic thermal field image sequence to generate a hot spot evolution atlas for key stages;

[0056] S5. Based on the morphological evolution of hot spots in the hot spot evolution map of key stages, determine the physical state of residual oil in the target area and generate residual oil state labels.

[0057] S6. Obtain the two-dimensional coordinates of each detection location, repeat steps S1 to S5, and map and associate the residual oil status label generated at each detection location with its two-dimensional coordinates to generate a visual image of the residual oil status.

[0058] In one embodiment of the present invention, step S1 includes the following steps: applying a preheating perturbation pulse to the target area and acting on the fresh liquid oil molecules in the pores inside the target area; during the application of the preheating perturbation pulse, forming a pre-excitation thermal field based on the surface temperature distribution of the target area and the response state of the internal oil; wherein the pulse width of the preheating perturbation pulse is set based on a preset energy threshold that can activate the fresh liquid oil molecules but is insufficient to change the physical morphology of the aged gum oil.

[0059] Specifically, for the powder metallurgy part to be tested, it is first securely mounted on the testing platform to prevent the part from shifting or vibrating during the subsequent application of thermal pulses and image acquisition. Then, the specific location on the part surface to be tested, i.e., the target area, is determined.

[0060] The target area is selected and locked using a positioning system, such as a laser pointer linked to the heat source device or a selection tool on the software interface. After positioning, the non-contact heat source is activated, and a preheating disturbance pulse is emitted to the locked target area through the control system. The heat source can be a laser or a high-power flash lamp, characterized by its ability to heat the part without direct contact. This pulse is usually set to a low-energy, long-duration pulse, typically with a duration, i.e., a pulse width, controlled between 50 and 100 milliseconds. When the pulse width is less than 50 milliseconds, the instantaneous power required for effective activation will be too high, which may lead to surface overheating or interference with aged gum oil, resulting in the loss of selective activation. When the pulse width is greater than 100 milliseconds, heat conduction will make the thermal field too flat, which is not conducive to forming a pre-excitation thermal field with sufficient spatial resolution, and the entire detection cycle will be unnecessarily prolonged. The heating energy is controlled at a low level. The purpose is not to significantly raise the part's temperature, but rather to allow heat to penetrate below the surface, activating any fresh liquid oil that may exist in the internal pores. This means that slight heating makes the oil molecules more active in thermal motion, slightly reducing viscosity. For aged, deteriorated, gelatinous or solidified oil contaminants, the energy provided by the pulse is insufficient to change their physical form. Therefore, when the preheating disturbance pulse acts on the target area, heat is conducted across the part's surface, interacting with the thermal response behavior of the oil in different states inside. The fresh liquid oil inside is more easily thermally conductive or flows due to activation, while aged, gelatinous oil exhibits different thermal conductivity characteristics. This surface temperature distribution, determined by the differences in internal and external thermophysical properties, constitutes the pre-excitation thermal field carrying the initial response information of the internal oil, thus forming the basis for the subsequent high-energy detection pulse.

[0061] It should be noted that the powder metallurgy part to be tested refers to a metal part manufactured using powder metallurgy processes that requires internal residual oil condition detection. These parts typically contain micron-sized pores and may contain residual lubricating oil or rust-preventive oil. The target area is the selected region on the surface of the powder metallurgy part to be tested; its size and shape are set according to the required detection resolution, typically at the millimeter level or smaller. The preheating disturbance pulse is an energy pulse with specific parameters, designed to subtly alter the thermal motion state of the internal oil molecules without causing a drastic temperature rise. The pre-excitation thermal field is the instantaneous temperature distribution field formed on the surface of the target area under the action of the preheating disturbance pulse. It is a two-dimensional temperature data matrix, where the value of each pixel represents the temperature at that location, reflecting the initial information of the oil's response to heat within the pores of the part. The detection platform is a mechanical device used to fix the powder metallurgy part to be tested, providing a stable and vibration-free measurement environment. Non-contact heat sources are used to apply preheating perturbation pulses. These are devices capable of applying energy to the surface of an object over a distance, such as xenon flash lamps or semiconductor lasers. Low energy and long duration describe the energy level and duration of the preheating perturbation pulse. The pulse width is set to 50 to 100 milliseconds. The energy setting is based on experimental data of the thermophysical properties of fresh lubricating oil and aged gum oil in typical powder metallurgy parts. The critical energy point is typically sufficient to activate fresh oil but insufficient to activate aged oil. Generally, this critical point is obtained, but not limited to, by calibrating multiple groups of samples containing different oils and substrates using differential scanning calorimetry. This critical point is the preset energy threshold. An example of the steps is as follows:

[0062] At least two sets of representative powder metallurgy standard samples were prepared. The first set of samples was impregnated with fresh liquid lubricating oil (the normal oil to be tested) in its internal pores. The second set of samples was filled with aged gum oil generated through accelerated thermal oxidation. Differential scanning calorimetry (DSC) was used to test the oil samples extracted from the first and second sets of samples, obtaining heat flow-temperature curves for the fresh liquid oil and the aged gum oil. Analysis of the curves clarified the initiation temperature range for phase transitions or activation reactions in both oils. Typically, the activation initiation temperature of the fresh liquid oil is significantly lower than the physical morphological change temperature of the aged gum oil. An experimental platform consistent with the detection method of this invention was constructed. Under the same detection environment, a series of single preheating perturbation pulses of different energy levels were sequentially applied to the same positions on the surfaces of the first and second set of standard samples (i.e., without applying the main detection pulse), and the pre-excitation thermal field image formed after each pulse was recorded simultaneously. The energy level was controlled by adjusting the pulse width and power of the heat source. The pre-excitation thermal field responses of the two sets of samples under different energy pulses were compared and analyzed.

[0063] The preset energy threshold is determined as follows: at this energy level, when applied to the first group of samples, i.e., fresh oil samples, its pre-excitation thermal field image shows a recognizable change in thermal diffusion pattern due to oil activation, such as blurring of hot spot edges; while when applied to the second group of samples, i.e., aged gum oil samples, no obvious change in thermal response caused by changes in the physical state of the oil is observed in its pre-excitation thermal field image. This energy value is the critical energy point that can selectively activate fresh oil rather than aged oil, and is used as the preset energy threshold. The preset energy threshold obtained above is converted into control parameters of specific heat source devices, such as pulse width and voltage of a specific type of flash lamp or laser, and is solidified and applied to all subsequent detection processes.

[0064] An exemplary, feasible implementation example is as follows: A gear manufactured using an iron-based powder metallurgy process is taken as the powder metallurgy part to be tested, and a certain part is suspected of containing residual lubricating oil used during processing. The gear is placed on and clamped on a three-axis movable testing platform. A circular area with a diameter of 2 mm on the surface of the gear is selected as the target area using a camera coaxial with the heat source of the flash lamp. Subsequently, the control system drives the xenon flash lamp to emit a low-energy thermal pulse with a pulse width of 80 milliseconds towards this target area. According to prior calibration, this energy can cause the surface temperature of the target area to rise gently by about 5 to 10 degrees Celsius during the pulse. During this process, if there is fresh liquid oil in the pores below the target area, the fluidity of the oil will increase slightly due to the temperature rise, which will affect the way heat diffuses on the surface; if there is aged gum oil or no oil below, the heat diffusion mode will be different. The different temperature distribution patterns formed on the surface of the part in these three cases constitute the pre-excitation thermal field. For example, the pre-excitation thermal field generated in the fresh oil area is generally a circular temperature zone with a slightly lower center temperature and blurred edges, while the pre-excitation thermal field generated in the oil-free area is generally a temperature zone with a higher center temperature and clear edges.

[0065] In one embodiment of the present invention, step S2 includes the following steps:

[0066] After the preheating disturbance pulse ends, wait for a preset time delay of 10 to 30 milliseconds; apply the main detection pulse to the target area to generate a transient thermal response based on the residual oil state, and generate a transient composite hot spot.

[0067] Specifically, after the preheating disturbance pulse ends and a pre-excitation thermal field carrying the initial response information of the internal oil is generated, a timing waiting phase begins. The duration of this waiting period is called the preset time delay, which is generally between 10 and 30 milliseconds. When the time delay is less than 10 milliseconds, the pre-excitation thermal field has not yet fully coupled with the internal oil, and the activation effect is not obvious, resulting in limited improvement in the signal-to-noise ratio of the subsequent main pulse excitation. When the time delay is greater than 30 milliseconds, the thermal field generated by the pre-excitation has spread to a large area, the surface temperature gradient decreases, and the superposition effect with the main detection pulse weakens, resulting in a significant decrease in the characteristics of the transient composite hot spot. The short pause is to allow the initial thermal effect caused by the preheating disturbance pulse to stabilize inside the part, so that the activation state of the oil molecules reaches an equilibrium point that can be used for subsequent detection, rather than being immediately interfered with by the next pulse. After the preset time delay ends, a second thermal energy injection is immediately applied to the same target area, i.e., a high-energy, short-duration main detection pulse is used. This pulse can be emitted by the same non-contact heat source, simply by adjusting its operating parameters to high-energy output mode, or another heat source specifically designed for high-energy output can be used, which can be selected as needed.

[0068] The pulse width of the main detection pulse is generally controlled between 1 and 5 milliseconds. When the pulse width is less than 1 millisecond, the instantaneous power requirement of the heat source equipment is high, and the resulting heat penetration depth may be insufficient. When the pulse width is greater than 5 milliseconds, the energy injection process is too long, and the superposition of the excited thermal response and the pre-excitation thermal field becomes complex, reducing the sharpness of transient features and hindering subsequent image analysis and state determination. The purpose is to inject a large amount of heat into the surface of the target area in a short time, thereby causing a rapid temperature jump. When this high-energy, short-duration pulse energy acts on the surface of the target area where a pre-excitation thermal field already exists, the two thermal effects are superimposed. The original pre-excitation thermal field represents the base temperature and the initial thermal state of the oil. Since fresh liquid oil and aged gum oil are in different thermophysical states after the pre-excitation stage, their responses to the second high-energy impact are also quite different. For example, activated fresh liquid oil may rapidly conduct heat to the interior through faster heat conduction or convection, while inert aged gum oil will accumulate more heat on the surface. The combined effect of this superposition and differential response generates a high signal-to-noise ratio instantaneous temperature peak distribution on the surface of the target area, which can reflect the differences in the internal oil. The resulting temperature field is the transient composite hot spot.

[0069] It should be noted that the preset time delay is a functional time parameter that defines the time interval between the end of the preheating perturbation pulse and the start of the main probe pulse. Its setting is based on an experimentally determined time window, such as 10 to 30 milliseconds. This time window maximizes the pre-excitation effect while avoiding excessive heat diffusion that could lead to information loss. The main probe pulse is a characteristic definition of the second thermal energy injection. Its high-energy attribute means that the energy injected per unit time is much greater than that of the preheating perturbation pulse, sufficient to cause a temperature rise on the surface. Its short-duration attribute refers to its pulse width, typically 1 to 5 milliseconds, to concentrate the heat effect within a short time. The transient composite hotspot is an instantaneous high-temperature region formed on the surface of the target area, generated by the combined effect of the pre-excitation thermal field and the main probe pulse. Its data structure is a two-dimensional temperature matrix, but its temperature values ​​are higher, its spatial gradient is steeper, and its morphology and brightness distribution strongly depend on the state of the residual oil below, thus exhibiting high recognizability or signal-to-noise ratio.

[0070] For example, continuing from the previous example, after a pre-excitation thermal field is generated on a target area with a diameter of 2 mm on the gear surface, the control system immediately starts timing, waiting for a preset time delay of 20 milliseconds. During this period, no operation is performed. Once 20 milliseconds have elapsed, the same xenon flash lamp used previously immediately flashes at a higher power, emitting a main detection pulse with a pulse width of 3 milliseconds towards the same target area, causing the surface temperature of that area to rise by 30 to 50 degrees Celsius in a short period of time. This rapid temperature rise, superimposed on the slight temperature rise previously caused by the pre-excitation thermal field, will cause the heat to diffuse rapidly to the surroundings and interior if the underlying material is fresh liquid oil, forming a transient composite hot spot with a high center brightness but relatively blurred edges; if the underlying material is aged gum oil, the heat will be highly concentrated on the surface, forming a transient composite hot spot with clear boundaries, high brightness, and uniformity.

[0071] In one embodiment of the present invention, step S3 includes the following steps:

[0072] Before applying the main detection pulse, a high-frequency infrared camera is activated to acquire images of the target area. Starting from the injection of the main detection pulse, the two-dimensional spatial distribution of the surface temperature field of the target area and its changes over time are continuously recorded to form a dynamic thermal field image sequence that includes information on the evolution of hot spot brightness, shape and size.

[0073] Specifically, before the main detection pulse is launched, a high-frequency infrared camera is pre-activated, and its lens is pre-adjusted to cover the target area and its surrounding area to capture the heat diffusion process. The image acquisition rate, or frame rate, is set to no less than 200 frames per second (200 Hz) to ensure sufficient temporal resolution to capture the millisecond-level transient thermal phenomena caused by the main detection pulse. The moment the main detection pulse is injected into the target area, a synchronization signal triggers the high-frequency infrared camera to begin recording data. From that moment, the camera continuously captures and saves a series of images of the surface temperature field of the target area. Each frame is a two-dimensional snapshot of the temperature distribution, fully recording the brightness, shape, and size of the transient composite hotspot at that precise point in time. This continuous recording process continues until the energy of the transient composite hotspot is largely dissipated, and the surface temperature distribution gradually returns to near its initial state before the application of the thermal pulse. After acquisition, all captured infrared image frames are processed, and a timestamp accurate to milliseconds or higher is added to each frame. Finally, based on these timestamps, all image frames are strictly arranged in chronological order from morning to night, forming a coherent image set. This final, chronologically arranged image set demonstrates the entire dynamic process of transient composite hotspots from their birth and development to their eventual cooling and disappearance, which is the dynamic thermal field image sequence.

[0074] It should be noted that a high-frequency infrared camera is a device capable of sensing infrared radiation from an object's surface and converting it into a visualized temperature image. The high-frequency attribute specifically refers to its image acquisition rate, i.e., a frame rate of no less than 200 Hz, enabling it to capture rapidly changing temperature processes. The dynamic thermal field image sequence is the final output of this step. Its data structure is an ordered image set, where each element is an infrared image frame, each associated with a unique timestamp. The entire sequence completely records the evolution information of transient composite hotspots in both time and space, including changes in brightness, contour morphology, and size expansion and contraction.

[0075] For example, before the control system emits a 3-millisecond main detection pulse to a target area with a diameter of 2 mm on the gear surface, a high-frequency infrared camera with a frame rate set to 250 Hz has already been activated and aimed at the area. The recording function of the camera is triggered synchronously the instant the main detection pulse begins to be injected. For instance, in the next approximately 100 milliseconds, the camera continuously captures 25 infrared images. These images clearly demonstrate how, if the underlying material is fresh liquid oil, a transient composite hot spot with a high center brightness but blurred edges rapidly forms, and then how its center brightness and overall size rapidly decay; and how, if the underlying material is aged gum oil, a transient composite hot spot with clear boundaries and high brightness forms, maintaining its high-brightness core area relatively stable while the surrounding areas cool down first. These 25 infrared images, precisely marked with their capture time and arranged sequentially, together constitute a dynamic thermal field image sequence.

[0076] In one embodiment of the present invention, step S4 includes the following steps:

[0077] Identify the termination time point corresponding to the end of the main detection pulse in the dynamic thermal field image sequence; select a time window with a duration of 5 to 50 milliseconds from the termination time point; extract and combine all infrared image frames captured within the time window to form a hot spot evolution atlas for key stages.

[0078] Specifically, based on the aforementioned dynamic thermal field image sequence, precise temporal localization is performed on the image sequence encompassing the entire process from hot spot formation to dissipation. By analyzing the brightness changes in each image frame of the sequence, the time point at which the main detection pulse ends is identified. This time point typically corresponds to the moment when the hot spot brightness reaches its peak or the peak plateau just ends in the entire sequence, serving as the zero-time reference for subsequent analysis. After determining this zero-time reference point, the next step is to define a time window of a specific length starting from this point on the time axis. Generally, the duration of this time window is set between 5 and 50 milliseconds. This duration is based on experimental observations of the initial patterns of heat diffusion from the surface to the interior and surrounding areas, ensuring that the selected time period most effectively reflects the differences in thermal diffusion behavior caused by different internal oil states. A time window that is too short may not capture sufficient change information, while a time window that is too long will result in blurred features due to excessive heat diffusion.

[0079] After setting the time window, all image frames whose timestamps fall within this specific time window are extracted from the dynamic thermal field image sequence. These extracted image frames constitute a new and smaller subset of images. This subset no longer includes the intense process of hot spot formation, nor the calm phase after the hot spot dissipates. Instead, it focuses on the hot spot morphology in the initial stage after the main detection pulse ends, when the enormous energy carried by the hot spot begins to diffuse outward, especially the relative brightness change between the center and the edge. This is also key information for determining the internal oil state.

[0080] Finally, the subset of images extracted and used to describe the morphological characteristics of the initial stage of heat diffusion is integrated to form an atlas for subsequent analysis, which is the hot spot evolution atlas of the key stage.

[0081] It should be understood that, since the hot spot evolution atlas of the key stages is a subset extracted from the dynamic thermal field image sequence, its data structure is still a collection of image frames ordered by time. Its purpose is to characterize the dynamic behavior of the hot spot in the early stage of cooling diffusion, especially the morphological and brightness evolution characteristics of the central and edge regions of the hot spot.

[0082] For example, taking the dynamic thermal field image sequence obtained in the above example, for a dynamic thermal field image sequence consisting of 25 infrared images, assuming the pulse lasts for 3 milliseconds, the frame rate is 250 Hz (approximately one frame), and the hot spot brightness reaches its peak in the third image frame, the time of the third image frame is defined as time zero. Then, starting from this time zero, a time window of 40 milliseconds is selected, corresponding to 10 frames, meaning all image frames from the third to the thirteenth image need to be extracted. These 11 consecutive image frames are extracted. The resulting subset of 11 images no longer contains the process of the hot spot appearing from nothing, but rather records how the center brightness and edge contour of the hot spot decay and diffuse within the first 40 milliseconds after the brightness reaches its peak. This atlas containing 11 images is the key stage hot spot evolution atlas used for subsequent analysis.

[0083] In one embodiment of the present invention, step S5 includes the following steps:

[0084] For each frame of the hot spot evolution atlas at the key stage, the average brightness of the hot spot center region and the average brightness of the edge region are calculated respectively. Based on the calculated average brightness, the brightness decay curve of the center region and the brightness decay curve of the edge region are generated. The difference in decay rate between the generated brightness decay curve of the center region and the brightness decay curve of the edge region is compared, and this difference is used as the basis for judging the physical state of residual oil inside the target region.

[0085] If the hot spot evolution atlas at the key stage shows that the difference in brightness decay rate between the central and edge regions of the hot spot is less than a first preset threshold, the residual oil status label is determined to be fresh liquid. If the hot spot evolution atlas at the key stage shows that the brightness of the central region of the hot spot is higher than that of the edge region, and the difference in brightness decay rate between the two is greater than a second preset threshold, the residual oil status label is determined to be aged gum. Judging from the difference in decay rate: if the decay rate of the brightness decay curve of the central region is basically the same as that of the brightness decay curve of the edge region, the residual oil status label is determined to be fresh liquid. If the decay rate of the brightness decay curve of the central region is lower than that of the brightness decay curve of the edge region, the residual oil status label is determined to be aged gum.

[0086] Among them, the first and second preset thresholds used to determine the state of residual oil are quantitative criteria determined by statistical analysis after conducting a large number of experimental tests on standard samples with known states.

[0087] The specific calibration method is as follows:

[0088] First, multiple sets of powder metallurgy standard samples impregnated with fresh liquid oil and standard samples filled with aged gum oil that has undergone accelerated aging treatment were prepared. All standard samples were tested using the method described above to obtain a key stage hotspot evolution atlas for each set of samples, and the brightness attenuation curves of the central and edge regions were calculated. For each attenuation curve, its attenuation rate was obtained through linear or exponential fitting. Then, the attenuation rate of the central region was calculated. With the decay rate of the edge region The relative difference value D is calculated using the following formula: The relative difference value D of all fresh liquid oil samples was statistically analyzed, and it was found that its distribution was concentrated in a small positive range or close to zero. The upper limit of this distribution range was taken as a first preset threshold, for example, plus three standard deviations. For example, the first preset threshold can be set to 15%. That is, when D < 15%, it is judged as fresh liquid. The relative difference value D of all aged gum oil samples was statistically analyzed, and it was found that its distribution was concentrated in a large positive range. The lower limit of this distribution range was taken as a second preset threshold. For example, the second preset threshold can be set to 30%. That is, when D > 30%, it is judged as aged gum. Finally, in actual testing, if the calculated relative difference value D is less than the first preset threshold, such as 15%, it is judged as fresh liquid; if D is greater than the second preset threshold, such as 30%, it is judged as aged gum. The region between the two can be marked as an uncertain or transitional state for further analysis.

[0089] It should be understood that the specific values ​​of the first and second preset thresholds can be fine-tuned according to the specific material, porosity, and oil type of the part to be tested, but the physical meaning and judgment logic of uniform diffusion and central aggregation they represent remain unchanged.

[0090] Specifically, each frame of the hotspot evolution atlas at the key stages is processed to quantify the brightness and contour features of the central and edge regions of the hotspot. For each frame, firstly, the centroid of the hotspot is located as the center point, and small shaped regions, such as circular regions, are defined as the central region; concentric rings surrounding the central region are defined as the edge regions. Then, the average brightness values ​​of the central region and the edge regions are calculated and recorded. Since the hotspot evolution atlas at the key stages is sorted chronologically, connecting the calculated center and edge brightness values ​​of each frame in chronological order yields two curves describing the decay of center and edge brightness over time. Next, a judgment is made based on the dynamic comparison of these two brightness decay curves and the observation of changes in the hotspot contour.

[0091] Judgment Method: If, in the initial frames of the image set, the hot spot does not form a particularly bright, concentrated core, but instead presents a soft halo with a relatively large area and uniform brightness distribution from the beginning, and the brightness decay curves of the center and the edge almost overlap, both showing a rapid and synchronous decreasing trend, and the calculated relative difference in decay rate is less than a first preset threshold, such as 15%, then the target area is determined to be a fresh liquid oil zone. If, in the initial images of the image set, the hot spot center quickly forms a clearly defined boundary with higher brightness than the surrounding structure, and the decrease rate of the center brightness decay curve is slower than that of the edge brightness decay curve, exhibiting a pattern where the brightness of the central area remains high for a long time, while the brightness of the edge area decreases rapidly first, and the calculated relative difference in decay rate is greater than a second preset threshold, such as 30%, then the target area is determined to be an aged gum oil zone. Finally, a text string is generated as the judgment result, which is the residual oil status label of the current target area.

[0092] The residual oil status label is a text-based data label, with the data value indicating whether it is in fresh liquid or aged gum. This label provides a clear classification of the internal physical state of residual oil at a single detection point. The morphological evolution pattern refers to the specific pattern of change in the visual characteristics of a hot spot, such as brightness, size, and outline, over time within the time window covered by the hot spot evolution atlas at key stages. Different patterns correspond to different internal thermophysical processes.

[0093] For example, continuing from the previous example, after receiving the key stage hotspot evolution atlas consisting of 11 images, these 11 images are processed sequentially. If the processing result shows that from the 1st to the 11th image, the hotspot never exhibits any sharp bright spots, and the brightness decay rate of its center and edges is basically the same, such as decaying by more than 80% within approximately 60 milliseconds, then a residual oil status label indicating fresh liquid is generated for this detection point. Conversely, if the processing result shows that a relatively bright center appears in the 1st image, and the brightness of this center decays slowly in the subsequent 10 images, for example, its brightness still remains above 70% of the peak value after 40 milliseconds, while its surrounding halo has already become dim, then a residual oil status label indicating aged colloid is generated for this detection point.

[0094] In one embodiment of the present invention, step S6 includes the following steps:

[0095] The control detection platform moves along a preset scanning path to cover the surface to be tested; steps S1 to S5 are repeated for each point on the scanning path to obtain the two-dimensional coordinates of each point and its corresponding residual oil status label; based on the two-dimensional coordinates of each point, a preset color corresponding to the residual oil status label is used to mark it, generating a visual image of the residual oil status.

[0096] Specifically, by moving the powder metallurgy part to be tested or moving the detection head, a scanning path needs to be pre-set before the detection begins. This path planning determines the distribution of detection points on the part surface. It can be point-by-point scanning, that is, moving and detecting one point at a time according to a preset grid; or line-by-line scanning, moving continuously along a straight line and triggering detection. The goal of path planning is to completely cover all surface areas that need to be detected. After the path planning is completed, the scanning process officially begins. In this process, whether it is point-by-point or line-by-line scanning, whenever the detection head moves to a new position, the entire process from step S1 to step S5 will be completely repeated. That is, for each new detection point, a preheating disturbance pulse will be applied sequentially to generate a pre-excitation thermal field, then a main detection pulse will be applied to generate a transient composite hot spot, followed by the acquisition of a dynamic thermal field image sequence by a high-frequency infrared camera, then the hot spot evolution atlas of key stages will be extracted, and finally, the residual oil status label of fresh liquid or aged gum will be generated by analyzing the atlas. During this process, the two-dimensional coordinate position of each detection point on the part surface is recorded simultaneously. Once the entire preset scan path has been executed, a complete dataset is obtained. This dataset contains the two-dimensional coordinates of all detection points and a residual oil status label corresponding to each coordinate. Finally, the data is visualized and integrated.

[0097] These data are processed, and the residual oil status of each test point is visually marked on the pre-acquired contour map or 3D model of the powder metallurgy part to be tested. For example, green is used to mark all points determined to be in a fresh liquid state, while red is used to mark all points determined to be aged colloids. After all points are marked, an intuitive, color map is formed, which clearly shows the physical state and spatial distribution of the residual oil beneath the entire surface to be tested, i.e., a residual oil status visualization image.

[0098] The residual oil visualization image is a two-dimensional or three-dimensional image. Its data structure consists of a set of marker points bound to the geometry of the part and encoded with color or grayscale. It visually displays the distribution of residual oil in different areas beneath the part's surface, allowing workers to identify areas with aging and failed oil contamination. It should be understood that the two-dimensional coordinates are a pair of values ​​used to precisely describe the position of each detection point on the surface of the powder metallurgy part being tested, typically corresponding to the X and Y axes of the inspection platform's movement.

[0099] For example, based on the above example, a comprehensive inspection of a specific tooth surface of the entire gear is performed. The control system drives the inspection platform, moving the inspection head point by point on the tooth surface according to a grid path with a spacing of 1 mm. Assuming the tooth surface size is 10 mm by 10 mm, a total of 100 points need to be inspected. At the first coordinate point (0,0), S1 to S5 are executed completely, assuming the final residual oil status label is fresh liquid. The coordinate (0,0) corresponding to fresh liquid is recorded. Then the inspection head moves to the second coordinate point (0,1), and S1 to S5 are executed completely again, assuming the residual oil status label generated this time is aged gum. The coordinate (0,1) corresponding to aged gum is recorded. This process is repeated 100 times until all grid points have been inspected. Finally, on the gray outline map of the gear tooth surface, a green dot is marked at the (0,0) position, a red dot is marked at the (0,1) position, and so on, marking all 100 inspection results on the outline map with dots of different colors. The final result is an intuitive visualization of the residual oil state, indicating the areas on the tooth surface where the oil has aged and the areas that still retain fresh liquid.

[0100] See appendix Figure 2 The present invention also proposes a residual oil detection system for powder metallurgy parts based on pulsed thermal excitation, comprising the following modules:

[0101] The pre-excitation thermal field generation module is used to apply a preheating disturbance pulse to the target area on the surface of the powder metallurgy part to be tested, and generate a pre-excitation thermal field carrying the initial response information of the internal oil.

[0102] The transient composite hot spot excitation module obtains a pre-excited thermal field, then applies a main detection pulse to the same target area after a preset time delay, and superimposes it with the pre-excited thermal field to generate a transient composite hot spot.

[0103] The thermal field image sequence acquisition module uses a high-frequency infrared camera to synchronously acquire the entire process of transient composite hot spots from formation to cooling and dissipation, generating a dynamic thermal field image sequence;

[0104] The hot spot evolution atlas extraction module is used to extract images within a specific time window after the end of the main probe pulse from the dynamic thermal field image sequence, and generate hot spot evolution atlases for key stages.

[0105] The residual oil status label generation module determines the physical state of residual oil within the target area based on the morphological evolution of hot spots in the hot spot evolution map set at key stages, and generates residual oil status labels.

[0106] The visualization image generation module is used to obtain the two-dimensional coordinates of each detection location, generate corresponding residual oil status labels one by one, and map and associate the residual oil status labels generated at each detection location with their two-dimensional coordinates to generate a visualization image of the residual oil status.

[0107] Each of the modules can be implemented in whole or in part through software, hardware, or a combination thereof. It supports hardware embedded in or independent of the processor in the computer device, and also supports software stored in the memory of the computer device, so that the processor can call and execute the operations corresponding to each of the above modules.

[0108] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for imaging residual oil inside a powder metallurgy part based on pulsed thermal excitation, characterized in that, The method comprises the following steps: S1, applying a pre-heating disturbance pulse to a target region on the surface of a powder metallurgy part to be tested to generate a pre-energized thermal field carrying initial response information of internal oil; S2, after a preset time lag, applying a main detection pulse to the same target region and superimposing it on the pre-energized thermal field to generate a transient composite thermal spot; S3, synchronously acquiring the whole process of the transient composite thermal spot from formation to cooling dissipation by using a high-frequency infrared camera device to generate a dynamic thermal field image sequence; S4, extracting images in a specific time window after the end of the main detection pulse from the dynamic thermal field image sequence to generate a key stage thermal spot evolution atlas; S5, judging the physical state of residual oil in the target region based on the morphological evolution of the thermal spot in the key stage thermal spot evolution atlas to generate a residual oil state label; S6, acquiring the two-dimensional coordinates of each detection position, repeating steps S1 to S5, and mapping and correlating the residual oil state label generated at each detection position with its two-dimensional coordinates to generate a residual oil state visualization image.

2. A method for imaging residual oil inside a powder metallurgy part based on pulsed thermal excitation according to claim 1, characterized in that, The pre-energized thermal field is generated by the following steps: applying a pre-heating disturbance pulse to the target region and acting on fresh liquid oil molecules in the pores in the target region; during the action of the pre-heating disturbance pulse, based on the surface temperature distribution of the target region and the response state of the internal oil, a pre-energized thermal field is formed.

3. A method for imaging residual oil inside a powder metallurgy part based on pulsed thermal excitation according to claim 1, characterized in that, The transient composite thermal spot is generated by the following steps: after the end of the pre-heating disturbance pulse, waiting for a preset time lag of 10 to 30 milliseconds; applying a main detection pulse to the target region to generate a transient thermal response based on the residual oil state to generate a transient composite thermal spot.

4. The method of imaging residual oil inside a powder metallurgy part based on pulsed thermal excitation according to claim 1, characterized in that, The dynamic thermal field image sequence is generated by the following steps: before applying the main detection pulse, starting the high-frequency infrared camera device to capture images of the target region; starting from the injection of the main detection pulse, continuously recording the two-dimensional spatial distribution of the surface temperature field of the target region and its data changing with time to form a dynamic thermal field image sequence including thermal spot brightness, morphology and size evolution information.

5. A method for imaging residual oil inside a powder metallurgy part based on pulsed thermal excitation according to claim 1, characterized in that, The key stage thermal spot evolution atlas is generated by the following steps: identifying the termination time point corresponding to the end of the main detection pulse in the dynamic thermal field image sequence; selecting a time window with a length of 5 to 50 milliseconds backward from the termination time point; extracting and combining all infrared image frames captured in the time window to form a key stage thermal spot evolution atlas.

6. A method of imaging residual oil inside a powder metallurgy part based on pulsed thermal excitation according to claim 1, characterized in that, The physical state of residual oil in the target region is judged by the following steps: if the key stage thermal spot evolution atlas shows that the brightness decay rate difference between the center region and the edge region of the thermal spot is less than a first preset threshold, the residual oil state label is determined to be fresh liquid; if the key stage thermal spot evolution atlas shows that the brightness of the center region of the thermal spot is higher than that of the edge region, and the brightness decay rate difference between the two is greater than a second preset threshold, the residual oil state label is determined to be aged gum.

7. A method for imaging residual oil inside a powder metallurgy part based on pulsed thermal excitation according to claim 1, characterized in that, The residual oil state visualization image is generated by the following steps: controlling the detection platform to move according to a preset scanning path to cover the surface to be tested; repeating steps S1 to S5 for each point on the scanning path to obtain the two-dimensional coordinates of each point and its corresponding residual oil state label; according to the two-dimensional coordinates of each point, using a preset color corresponding to the residual oil state label for marking to generate a residual oil state visualization image.

8. A method of imaging residual oil inside a powder metallurgy part based on pulsed thermal excitation according to claim 2, characterized in that, The pulse width of the preheating disturbance pulse is set according to a preset energy threshold value that can activate fresh liquid oil molecules but is insufficient to change the physical form of aged gelled oil.

9. A method of imaging residual oil inside a powder metallurgy part based on pulsed thermal excitation according to claim 1, characterized in that, Before generating the residual oil state label, the following steps are further included: For each frame image in the key stage thermal spot evolution atlas, the average brightness of the center region and the average brightness of the edge region are calculated respectively; According to the calculated average brightness, a center region brightness decay curve and an edge region brightness decay curve are generated; The decay rate difference between the generated center region brightness decay curve and the edge region brightness decay curve is compared, and the difference is used as a judgment basis for judging the physical state of the residual oil in the target region.

10. A method of imaging residual oil inside a powder metallurgy part based on pulsed thermal excitation according to claim 8, characterized in that, The determination of the preset energy threshold value includes the following steps: Based on fresh liquid oil samples and aged gelled oil samples, differential scanning calorimetry is used for testing to obtain the heat flow-temperature curve of fresh liquid oil and aged gelled oil; Under the same detection environment, preheating disturbance pulses are applied to the fresh liquid oil samples and the aged gelled oil samples to obtain pre-excitation thermal field images; By comparing and analyzing the pre-excitation thermal field responses under different energy pulses, the preset energy threshold value is obtained.

Images