Water jacket CHT automatic calibration method based on 3D printing and FBG

Abstract

This application relates to the field of engine testing and simulation technology, and discloses an automatic CHT calibration method for water jackets based on 3D printing and FBG. This method utilizes 3D printing to form conformal microchannels inside an engine prototype, embedding a fiber Bragg grating sensor with a polyimide coating in a free state. Continuous temperature data of the water jacket's inner wall is collected during bench testing, generating a temperature gradient curve for the test space. A conjugate heat transfer digital simulation model containing the corresponding microchannel structure is established, and simulation probe lines with equivalence in spatial coordinates are defined and the calculated temperature curve is extracted. The root mean square error between the experimental and simulation curves is calculated using an automated control script, and an adaptive search algorithm is used to correct the boiling model parameters in the simulation model based on the error. Iterative calculations are executed repeatedly until the error meets the convergence condition, and the corresponding parameters are output to complete the calibration. This invention achieves continuous spatial temperature acquisition without flow field disruption and automated closed-loop optimization of simulation parameters.

Description

Technical Field

[0001] This invention relates to the field of engine testing and simulation technology, specifically to an automatic calibration method for water jacket CHT based on 3D printing and FBG. Background Technology

[0002] In the internal combustion engine R&D process, conjugate heat transfer simulation technology is applied to the structural design and heat transfer performance verification of cooling water jackets. Accurate conjugate heat transfer simulation results depend on the precise setting of the internal boundary conditions and boiling model parameters of the simulation model. Existing simulation parameter calibration schemes are usually implemented during the engine bench test phase. The operation involves mechanically drilling holes in the metal shell of the cylinder head or cylinder block down to the water jacket wall and placing K-type thermocouples in the holes. The test system obtains single-point temperature data at coordinate locations through thermocouples, and then operators manually compare the results with the calculation results output by the fluid dynamics simulation software, and manually adjust the values ​​of the heat transfer coefficient or boiling model parameters to achieve error convergence.

[0003] The existing calibration methods based on thermocouple temperature measurement and manual parameter adjustment have limitations in measurement and process loop. Regarding the measurement structure, the outer diameter of traditional thermocouple probes is typically greater than 2 mm, making them difficult to place in the compact internal space of modern high-heat-load engines. Drilling through the water jacket wall would compromise the seal integrity of the casing, and probe penetration into the water jacket would alter the fluid boundary layer thickness at the bottom of the coolant and disturb the local flow field, causing the acquired temperature data to deviate from the actual heat transfer environment. Furthermore, due to probe size limitations, the temperature sensing element cannot penetrate narrow areas such as the exhaust valve nose area, where space is confined and localized boiling phase transitions are likely to occur.

[0004] Regarding data matching and calibration processes, single-point temperature measurement can only output discrete spatial coordinate temperature values, failing to reconstruct a continuous temperature gradient distribution in three-dimensional space. Because discrete data lacks continuous spatial variation trends, simulators cannot accurately capture the critical spatial points where the coolant undergoes phase transitions such as initial boiling and bubble shedding based on limited scattered data. Furthermore, the current parameter adjustment process relies on the operator's engineering experience. After modifying the boiling model parameters in the simulation software, the system needs to re-execute the fluid dynamics calculation and manually re-compare the temperature errors. This manual trial-and-error mechanism disrupts data interaction between the testing hardware and the simulation software, resulting in long calibration iteration cycles and hindering automated optimization and convergence of parameter variables. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an automatic calibration method for water jacket heat transfer temperature (CHT) based on 3D printing and FBG. This method solves the problems of existing single-point temperature probes penetrating the water jacket and disrupting the local flow field, failing to obtain continuous spatial temperature gradients, and relying on manual trial and error to adjust conjugate heat transfer simulation parameters based on discrete data, resulting in long calibration cycles and the inability to automate closed-loop iterative optimization.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] This invention provides an automatic calibration method for water jacket CHT based on 3D printing and FBG, comprising the following steps:

[0008] An engine prototype was formed using 3D printing technology, and a conformal microchannel was formed inside the metal body of the engine prototype. A fiber Bragg grating sensor with a polyimide coating was then inserted into the conformal microchannel.

[0009] The engine sample was installed on an engine bench for operation testing. The temperature distribution data along the inner wall of the water jacket was collected by the fiber Bragg grating sensor to generate a temperature gradient curve of the test space.

[0010] A conjugate heat transfer digital simulation model containing the conformal microchannel is established, a simulation probe line is defined in the conjugate heat transfer digital simulation model, and the temperature curve of the simulation probe line obtained by iterative calculation is extracted.

[0011] The root mean square error between the simulated probe line temperature curve and the experimental space temperature gradient curve is calculated using an automated control script. An adaptive search algorithm is then used to reverse-correct the boiling model parameters in the conjugate heat transfer digital simulation model based on the root mean square error.

[0012] The steps of extracting the temperature curve of the simulation probe line and correcting the boiling model parameters are repeated until the root mean square error meets the convergence condition, and the corresponding boiling model parameters are output to complete the calibration.

[0013] Preferably, inserting the fiber Bragg grating sensor with a polyimide coating into the conformal microchannel specifically includes:

[0014] The fiber Bragg grating sensor with polyimide coating is placed inside the space of the conformal microchannel;

[0015] By sealing both ends of the conformal microchannel, the fiber Bragg grating sensor with polyimide coating is left with a gap between the conformal microchannel and the metal inner wall and is in a non-fixed, free state.

[0016] Preferably, mounting the engine sample on an engine bench for operational testing specifically includes:

[0017] The test assembly, which includes the water jacket and the engine sample, is fixed to the engine test bench.

[0018] Coolant is introduced into the water jacket and the engine test bench is run, causing the coolant to undergo subcooled boiling in the water jacket inner wall region corresponding to the conformal microchannel, resulting in a step change in the wall heat flux density.

[0019] Preferably, the method of acquiring spatial temperature distribution data along the inner wall of the water jacket using the fiber Bragg grating sensor to generate a temperature gradient curve for the test space specifically includes:

[0020] Continuous high-frequency spatial temperature node data is acquired by the fiber Bragg grating sensor in a state where it is not in contact with the coolant.

[0021] The continuous high-frequency spatial temperature node data is converted into a continuous spatial temperature gradient curve, and the temperature characteristic inflection point corresponding to the sudden change in wall heat flux density is marked on the continuous spatial temperature gradient curve as the experimental space temperature gradient curve.

[0022] Preferably, establishing the conjugate heat transfer digital simulation model including the conformal microchannel specifically includes:

[0023] Establish a conjugate heat transfer digital simulation model that includes both fluid and solid domains;

[0024] Based on the actual three-dimensional coordinate dimensions of the conformal microchannel, a geometrically consistent model microchannel space is constructed within the solid domain of the conjugate heat transfer digital simulation model.

[0025] Preferably, in the conjugate heat transfer digital simulation model, a simulation probe line is defined, and the temperature curve of the simulation probe line obtained by iterative calculation is extracted, specifically including:

[0026] Based on the spatial coordinate correspondence, a simulation probe line with coincident coordinate positions is constructed in the conjugate heat transfer digital simulation model;

[0027] Import the data table of the temperature gradient curve of the test space into the fluid dynamics simulation software according to the data table format;

[0028] During the fluid dynamics iterative calculation process, the simulation calculation temperatures corresponding to multiple nodes on the simulation probe line are extracted to form the simulation probe line temperature curve.

[0029] Preferably, the step of calculating the root mean square error between the simulated probe line temperature curve and the experimental space temperature gradient curve using an automated control script specifically includes:

[0030] The automated control script, written in Python, is used in conjunction with a network-attached storage system to access the data table of the temperature gradient curve in the test space.

[0031] Extract the test temperature values ​​of the test nodes on the test space temperature gradient curve and the calculated temperature values ​​of the corresponding nodes on the simulated probe line temperature curve;

[0032] The root mean square error is obtained by calculating the sum of squares of the differences between the test temperature value and the calculated temperature value, taking the average value, and then performing a square root operation.

[0033] Preferably, the step of using an adaptive search algorithm to reverse-correct the boiling model parameters in the conjugate heat transfer digital simulation model based on the root mean square error specifically includes:

[0034] Set the vaporization core density multiplier and the bubble detachment diameter correction coefficient as parameter variables;

[0035] In the optimization manager of the fluid dynamics simulation software, the SHERPA hybrid adaptive search algorithm is invoked to dynamically adjust the distribution position of subsequent sampling points based on the root mean square error value output by the previous calculation.

[0036] The adjusted parameter variables are written into the conjugate heat transfer digital simulation model for the next fluid dynamics iteration calculation.

[0037] Preferably, the step of cyclically extracting the simulated probe line temperature curve and inversely correcting the boiling model parameters until the root mean square error meets the convergence condition specifically includes:

[0038] The reverse-corrected boiling model parameters are input into the conjugate heat transfer digital simulation model to perform fluid dynamics iterative calculations.

[0039] Extract the temperature curve of the simulated probe line output after the completion of the fluid dynamics iterative calculation;

[0040] Compare the numerical differences between the simulated probe line temperature curve and the experimental space temperature gradient curve in the segment containing the temperature characteristic inflection point;

[0041] When the root mean square error reaches a preset minimum value, and the simulated probe line temperature curve and the test space temperature gradient curve coincide in the segment containing the temperature characteristic inflection point, the convergence condition is determined to be met and the loop execution stops.

[0042] Preferably, the calibration of the boiling model parameters corresponding to the output specifically includes:

[0043] Extract the boiling model parameters from the conjugate heat transfer digital simulation model that satisfies the convergence condition;

[0044] The boiling model parameters that meet the convergence conditions are set as the standard simulation parameter set;

[0045] Store the standard simulation parameter set and set it as the input parameters for the conjugate heat transfer simulation of the thermal management of the corresponding engine model.

[0046] This invention provides an automatic calibration method for the water jacket CHT based on 3D printing and FBG. It has the following beneficial effects:

[0047] 1. This invention achieves isolation between the temperature sensing element and the coolant fluid by forming a conformal microchannel inside the metal body of the engine prototype and embedding a fiber Bragg grating sensor. This avoids the damage to the local flow field and temperature boundary layer of the coolant caused by the intrusion of traditional temperature probes into the water jacket, ensuring the consistency between the bench test environment and the flow field boundary conditions of the simulation model. Simultaneously, the sensor is sealed at both ends within the microchannel and remains in a free state, cutting off the path for the thermal expansion of the metal shell to transmit mechanical strain to the sensor, eliminating the error interference caused by mechanical stretching leading to grating wavelength drift.

[0048] 2. This invention utilizes a fiber Bragg grating sensor to collect temperature data along the inner wall of the water jacket, transforming discrete single-point temperature measurement into continuous spatial line field temperature measurement. The continuous spatial temperature gradient curve generated from the acquired high-frequency spatial temperature node data can completely record the temperature gradient variation along the line and directly capture the temperature characteristic inflection point corresponding to the step change in wall heat flux density caused by local subcooling boiling of the coolant. This provides fluid dynamics simulation software with continuous spatial comparison benchmark data containing critical phase transition characteristics.

[0049] 3. This invention utilizes automated control scripts to interface with fluid dynamics simulation software, constructing a closed-loop calibration workflow driven by an adaptive search algorithm. The root mean square error between the experimental space temperature gradient curve and the simulated probe line temperature curve is set as the objective function. Based on the error values ​​output by iterative calculations, the boiling model parameters, such as the vaporization core density multiplier, are automatically inferred and corrected. This method replaces manual comparison and trial-and-error parameter tuning with data-driven approaches, eliminating reliance on human experience and achieving automated optimization and convergence of model parameters, thus shortening the parameter calibration cycle for conjugate heat transfer simulations. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0051] Figure 2 This is a cross-sectional schematic diagram of the cylinder head of the present invention. Detailed Implementation

[0052] The technical solutions in 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, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] Example:

[0054] Reference Figures 1-2 This invention provides an automatic calibration method for the water jacket CHT based on 3D printing and FBG, including the following steps:

[0055] An engine prototype was formed using 3D printing technology, and a conformal microchannel was formed inside the metal body of the engine prototype. A fiber Bragg grating sensor with a polyimide coating was then inserted into the conformal microchannel.

[0056] The engine prototype was installed on an engine test bench for operation testing. The temperature distribution data along the inner wall of the water jacket was collected by a fiber Bragg grating sensor to generate a temperature gradient curve for the test space.

[0057] A conjugate heat transfer digital simulation model containing conformal microchannels was established. A simulation probe line was defined in the conjugate heat transfer digital simulation model, and the temperature curve of the simulation probe line obtained by iterative calculation was extracted.

[0058] The root mean square error between the simulated probe line temperature curve and the experimental space temperature gradient curve is calculated using an automated control script. An adaptive search algorithm is then used to reverse-correct the boiling model parameters in the conjugate heat transfer digital simulation model based on the root mean square error.

[0059] The steps of extracting the simulation probe line temperature curve and reversing the boiling model parameters are executed repeatedly until the root mean square error meets the convergence condition, and the corresponding boiling model parameters are output to complete the calibration.

[0060] The aforementioned automatic CHT calibration method for water jackets based on 3D printing and FBG is established in a hardware testing environment consisting of an engine test bench, an engine prototype, and a fiber Bragg grating sensor with a polyimide coating. A conformal microchannel is used to isolate the sensor from the coolant fluid inside the water jacket. The sensor output, containing the actual boundary phase transition characteristics, is used as the comparison input data for subsequent closed-loop calculations.

[0061] The aforementioned automatic CHT calibration method for water jackets based on 3D printing and FBG relies on fluid dynamics simulation software to construct a software simulation environment. The conjugate heat transfer digital simulation model maps the position of the conformal microchannel in three-dimensional spatial coordinates to generate the corresponding simulation probe line temperature curve as the calculation output data in the iterative process.

[0062] The automated control script establishes a closed-loop data flow relationship with central scheduling. Executed outside the fluid dynamics simulation software, the script extracts the experimental space temperature gradient curve output from the hardware testing environment and the simulation probe line temperature curve output from the software simulation environment, and performs root mean square error calculation.

[0063] The automated control script calls the adaptive search algorithm to change the values ​​of the boiling model parameters. The changed boiling model parameters are then input into the fluid dynamics simulation software to perform the next iteration calculation, thus constructing a continuous data interaction loop until the convergence condition is met and the iteration stops.

[0064] Reference Figure 2 ,Should Figure 2 A schematic diagram of a partial cross-sectional structure of the interface between the fluid cavity and the solid metal entity inside the engine prototype is shown. The cavity structure shown in the figure includes a water jacket, an intake duct, and an exhaust duct.

[0065] Microchannels (i.e., conformal microchannels in the aforementioned embodiment) are formed within the metal entity between the intake duct and the water jacket, and within the metal entity near the periphery of the exhaust duct (corresponding to the area around the exhaust valve seat). These conformal microchannels are arranged in three-dimensional space near the boundary of the water jacket and maintain a parallel correspondence with the inner wall shape of the bottom of the water jacket. The wall thickness of the metal entity between the outer wall of the conformal microchannel and the inner wall of the water jacket is between 1.5 mm and 2.0 mm, thereby establishing a temperature measurement channel that does not disrupt the fluid boundary layer of the coolant inside the water jacket.

[0066] Reference Figures 1-2 The method of using 3D printing technology to form engine prototypes and then forming conformal microchannels inside the metal entity of the engine prototypes specifically involves using laser melting 3D printing equipment to form engine prototypes, setting aluminum alloy powder or cast iron metal powder as the forming raw material, and extracting unformed powder from the unmelted areas inside the metal entity during the manufacturing process of melting and forming the metal entity of the engine prototype layer by layer, directly forming conformal microchannels in a hollow state.

[0067] The conformal microchannels are arranged around the exhaust valve seat of the engine prototype. The three-dimensional spatial orientation of the conformal microchannels is parallel to the shape of the inner wall of the water jacket of the engine prototype. The thickness of the solid metal wall between the outer wall of the conformal microchannel and the inner wall of the water jacket is set to be between 1.5 mm and 2.0 mm. The inner diameter of the conformal microchannel is limited to less than 1 mm. Under the premise of maintaining the original fluid boundary layer thickness of the coolant and preventing the temperature sensing element from intruding into the water jacket, the above dimensional conditions and spatial parallel position establish temperature sensing channels that are equidistantly distributed from the heat exchange interface.

[0068] For the placement of the temperature sensing element, a fiber Bragg grating sensor with a polyimide coating on its outer surface was selected. The polyimide coating has the characteristic of withstanding temperatures above 300°C, which meets the testing environment conditions of the high-temperature area of ​​the engine cylinder head. The fiber Bragg grating sensor with the polyimide coating was placed into the space inside the conformal microchannel from the external opening end of the conformal microchannel.

[0069] After insertion, a high-temperature resistant sealing material is used to seal the openings at both ends of the conformal microchannel. Because the outer diameter of the fiber Bragg grating sensor with polyimide coating is smaller than the inner diameter of the conformal microchannel, an air gap remains between the polyimide-coated fiber Bragg grating sensor body and the surrounding metal inner wall after sealing. This air gap ensures that the polyimide-coated fiber Bragg grating sensor is not structurally fixed to the metal inner wall, keeping the sensor in a free, non-fixed state.

[0070] The free state structurally blocks the transmission path of axial tensile stress from the mechanical deformation caused by thermal expansion of the metal entity to the fiber Bragg grating sensor. The center wavelength drift of the fiber Bragg grating sensor is affected by both temperature change and axial strain change, and the corresponding relationship for the grating center wavelength drift is:

[0071] ;

[0072] In the formula, This indicates the center wavelength shift of the fiber Bragg grating sensor; This represents the initial center wavelength of the fiber Bragg grating when it is not subject to external interference. This represents the coefficient of thermal expansion of the optical fiber material; This represents the thermo-optic coefficient of the optical fiber material; This indicates the change in external temperature. Indicates the effective elastic-optical coefficient of the optical fiber material; This represents the change in mechanical strain acting on the optical fiber axis.

[0073] Under the free-state conditions provided within the conformal microchannel, the metal entity of the engine prototype did not exert a mechanical tensile force on the fiber Bragg grating sensor with a polyimide coating, resulting in a change in mechanical strain. The value is zero. This includes the change in mechanical strain. The product term is zeroed out in the above equation. The center wavelength shift of the fiber Bragg grating sensor output with a polyimide coating is only related to the temperature change. A linear correspondence is formed to eliminate the superimposed interference caused by the mechanical strain of the engine sample due to thermal expansion during bench operation on the grating wavelength drift value.

[0074] Reference Figures 1-2 The engine prototype was mounted on an engine bench for operational testing. Specifically, the test assembly, including the water jacket and engine prototype, was bolted to the engine bench. A dynamometer was connected, and the engine bench was run with target speed and load values ​​set. An external coolant circulation system was activated, and coolant at a set pressure and initial temperature was introduced into the water jacket.

[0075] During the operation test, the combustion chamber top and exhaust valve seat area of ​​the engine sample conduct heat outward. When the local metal wall temperature of the water jacket inner wall area corresponding to the conformal microchannel exceeds the saturation temperature of the coolant under the corresponding pressure, and the temperature of the mainstream area of ​​the coolant inside the water jacket is still lower than the saturation temperature, the coolant in the water jacket inner wall area undergoes subcooling boiling.

[0076] During the sub-cold boiling stage, vaporization nuclei are generated on the inner wall surface of the water jacket. The movement of bubbles generated and detached at the solid-liquid interface disturbs the fluid boundary layer at the bottom of the coolant. Based on the laws of heat conduction, the single-phase forced convection heat transfer state transforms into a nucleation boiling heat transfer state, resulting in a step increase in the convective heat transfer coefficient in this local region. This step increase in the convective heat transfer coefficient directly leads to a step change in the wall heat flux density.

[0077] The acquisition of spatial temperature distribution data along the inner wall of the water jacket using a fiber Bragg grating sensor specifically involves connecting the pigtail of a fiber Bragg grating sensor with a polyimide coating to a grating demodulator. The demodulator acquires the continuous high-frequency center wavelength reflection signals from multiple grating nodes distributed along the fiber axis on the polyimide-coated fiber Bragg grating sensor. Since the sensor structure is not in contact with the coolant and mechanical strain interference is eliminated, the acquired wavelength signals are converted into spatial temperature node data using the following linear conversion formula:

[0078] ;

[0079] In the formula, Indicates the first Spatial coordinates of each grating node within the conformal microchannel; Spatial position coordinates are The real-time temperature values ​​obtained by measuring the grating nodes; This represents the initial reference temperature value determined under environmental calibration conditions. Spatial position coordinates are The real-time center wavelength value reflected by the grating node; This indicates the initial center wavelength value corresponding to the grating node; This indicates the temperature sensitivity coefficient of the fiber Bragg grating sensor, expressed in picometers per degree Celsius.

[0080] The specific steps for generating the experimental space temperature gradient curve include: using multiple discrete spatial temperature node data points obtained through calculation as discrete sample points, and using a spline interpolation algorithm to fit the discrete data into a continuous spatial temperature gradient curve, so that the curve has continuous spatial position coordinate variable characteristics.

[0081] In solid-state heat transfer mechanisms, the heat flux density step corresponding to subcooling boiling alters the heat loss boundary conditions for heat transfer from the inner wall of the water jacket to the coolant, causing a sudden change in the slope of the temperature gradient along the temperature measurement channel path within the metal entity. Spatial first derivative calculations are performed to analyze the slope changes on the continuous spatial temperature gradient curve, extracting the coordinate points where the first derivative values ​​undergo discontinuous jumps. These coordinate points are marked on the curve as the temperature characteristic inflection points corresponding to the abrupt change in heat flux density at the wall surface. The curve containing the coordinate information of these temperature characteristic inflection points is output as the experimental space temperature gradient curve.

[0082] Reference Figures 1-2 The establishment of a conjugate heat transfer digital simulation model encompassing both fluid and solid domains specifically involves importing the three-dimensional geometric digital model file of the engine prototype into the fluid dynamics simulation software. Within the fluid dynamics simulation software, a fluid domain mesh is created for the fluid space inside the engine's water jacket, and a solid domain mesh is created for the engine's metallic solid space.

[0083] Conjugate heat transfer energy calculation boundary conditions are set at the interface between the fluid domain mesh and the solid domain mesh. Within the conjugate heat transfer calculation framework, the interface between the solid and fluid domains follows energy conservation and temperature continuity properties, and the corresponding governing equations are as follows:

[0084] ;

[0085] ;

[0086] In the formula, This represents the calculated temperature value at the grid node side of the solid domain at the interface; This represents the calculated temperature value at the mesh node side of the fluid domain at the interface; This represents the specified thermal conductivity value of a metallic solid material; This indicates the set thermal conductivity value of the coolant fluid; This represents the normal direction vector perpendicular to the fluid-solid interface; This represents the calculated temperature gradient of the solid domain at the interface along the normal direction; This represents the calculated temperature gradient of the fluid domain at the interface along the normal direction. The governing equations described above mathematically ensure that the amount of heat transferred from the fluid to the metal wall via convection is equivalent to the amount of heat conducted from the wall to the interior of the metal.

[0087] Based on the actual three-dimensional coordinates of the conformal microchannel, a geometrically consistent model microchannel space is constructed within the solid domain of the conjugate heat transfer digital simulation model. This specifically involves performing a Boolean subtraction operation on the three-dimensional geometric model. Within the established three-dimensional mesh space of the solid domain, a cylindrical volume space that completely corresponds to the inner diameter and three-dimensional orientation of the conformal microchannel in step S1 is removed, ensuring exclusive mapping consistency at the entity attribute level.

[0088] In the conjugate heat transfer digital simulation model, defining the simulation probe line specifically involves extracting a set of spatial coordinate point arrays along the geometric central axis of the microchannel space in the model, according to the correspondence in a three-dimensional Cartesian coordinate system. The three-dimensional coordinate values ​​contained in this array are equal to the three-dimensional spatial coordinate values ​​of the location of the fiber Bragg grating sensor with a polyimide coating. A curve connection operation is performed to connect the spatial coordinate point arrays into a continuous three-dimensional spatial structure curve, which is then defined as the simulation probe line. This step maps the geometric coordinate attributes of the sensor placement location within the three-dimensional space of the digital simulation model, establishing a coordinate-equivalent data extraction path.

[0089] Importing the experimental space temperature gradient curve data table into the fluid mechanics simulation software according to the data table format specifically includes: storing the mapping table containing discrete spatial position coordinates and corresponding experimental temperature values ​​on the experimental space temperature gradient curve as a comma-separated text file; calling the data table input interface function module in the fluid mechanics simulation software to read the data from the text file in memory, and storing the array of experimental temperature values ​​contained therein in the memory variable space of the fluid mechanics simulation software.

[0090] In the iterative fluid dynamics calculation process, the simulated temperatures corresponding to multiple nodes on the simulated probe line are extracted to form the simulated probe line temperature curve. Specifically, this involves: at each steady-state calculation step of the fluid dynamics conservation control equation matrix solution algorithm, upon convergence of a single solution, calculating and extracting the discrete temperature numerical variables corresponding to the spatial coordinate point array on the simulated probe line using a grid interpolation algorithm. The extracted discrete temperature numerical variables are then combined according to the order of arrangement along the three-dimensional spatial coordinates to generate a simulated probe line temperature curve continuously arranged according to the one-dimensional spatial length distribution law. The simulated probe line temperature curve provides a benchmark sample of the calculation results output by the software simulation end in the subsequent adaptive calibration process.

[0091] Reference Figures 1-2The root mean square error (RMSE) between the simulated probe line temperature curve and the experimental space temperature gradient curve is calculated using an automated control script. Specifically, this involves running an automated control script written in Python in the background of the computing workstation's operating system. The script establishes a communication connection with a network attached storage system (NAS) by invoking the network file system protocol. It then reads a data table containing the experimental space temperature gradient curve from the NAS and loads the spatial three-dimensional coordinate data of the test nodes and their corresponding experimental temperature values ​​into the running memory.

[0092] When the fluid dynamics simulation software completes a single steady-state calculation iteration or reaches the set save step size, the automated control script sends data extraction commands to the fluid dynamics simulation software through the application programming interface. The automated control script extracts the calculated temperature values ​​of the corresponding coincident coordinate nodes on the simulation probe line temperature curve based on the spatial three-dimensional coordinate data loaded into memory.

[0093] After obtaining the experimental and calculated temperature values, the automated control script executes a mathematical operation to calculate the sum of squares of the differences between the two values ​​and uses the root mean square error (RMSE) as the objective function. The corresponding formula for calculating the RMS error is:

[0094] ;

[0095] In the formula, This represents the calculated root mean square error value. This indicates the total number of nodes extracted based on the matched coordinates; Indicates the first Spatial coordinates of each grating node within the conformal microchannel; This indicates the test temperature value corresponding to the extracted node coordinates; This represents the calculated temperature value corresponding to the same node coordinates. The above root mean square error values ​​objectively reflect the degree of global error in the temperature distribution of the simulated flow field output in the current iteration step, which deviates from the test results.

[0096] The boiling model parameters in the conjugate heat transfer digital simulation model are corrected in reverse based on the root mean square error using an adaptive search algorithm. Specifically, this includes extracting the vaporization nucleus density multiplier and the bubble escape diameter correction coefficient from the boiling model control panel of the fluid dynamics simulation software and setting them as independent variables. The vaporization nucleus density multiplier determines the base number of vaporization nuclei generated per unit area at the solid-liquid interface; the bubble escape diameter correction coefficient determines the calculated volume equivalent diameter of the bubble when it escapes the heated wall and enters the mainstream fluid region. These two parameters serve as the underlying calculation input conditions, determining the latent heat flux density distribution during the phase change process.

[0097] The built-in optimization manager module of the fluid dynamics simulation software is launched, and the SHERPA hybrid adaptive search algorithm is invoked. The vaporization core density multiplier and the bubble escape diameter correction coefficient, set as independent variables, are input into the optimization manager module as the design variable space of the adaptive search algorithm, and the root mean square error value of each calculation output is used as the response target of the adaptive search algorithm.

[0098] The SHERPA hybrid adaptive search algorithm receives the input root mean square error value and performs response surface fitting on the recorded historical error matrix to determine the gradient direction of the objective function in the variable space. Based on the gradient direction analyzed, the adaptive search algorithm adjusts the distribution of subsequent sampling points in the variable space, outputting a new set of back-corrected vaporization core density multipliers and bubble escape diameter correction coefficients. An automated control script extracts the back-corrected numerical variables, overwrites them in the solver control file of the conjugate heat transfer digital simulation model, and triggers the fluid dynamics simulation software to perform the next fluid dynamics iteration calculation operation according to the updated parameters.

[0099] Reference Figures 1-2 The steps of cyclically extracting the simulation probe line temperature curve and the reverse correction boiling model parameters specifically include writing the reverse correction boiling model parameters output by the SHERPA hybrid adaptive search algorithm into the control file of the conjugate heat transfer digital simulation model, and triggering the fluid dynamics simulation software to start the calculation of the fluid dynamics control equations for the current iteration step.

[0100] Once the fluid dynamics solution process reaches steady-state convergence, the fluid dynamics simulation software outputs the three-dimensional spatial calculated temperature field for the current iteration batch. An automated control script extracts nodal temperature data along a predefined spatial coordinate sequence of the simulation probe line, generating an updated simulation probe line temperature curve.

[0101] The comparison of the numerical difference between the simulated probe line temperature curve and the experimental space temperature gradient curve within the segment containing the temperature feature inflection point specifically involves executing local feature overlap comparison logic. The automated control script extracts the spatial coordinates of the temperature feature inflection point corresponding to the marked abrupt change in wall heat flux density. The spatial coordinates of the temperature feature inflection point are set as the center point, and a preset coordinate distance is extended along both ends of the simulated probe line to form a local comparison segment.

[0102] The automated control script calculates the absolute value of the temperature difference between the simulated probe line temperature curve and the experimental space temperature gradient curve at corresponding coordinate nodes within a local comparison section. The condition for determining local overlap is that the absolute temperature difference between all nodes within the local comparison section is less than a preset maximum allowable deviation threshold. This local feature comparison step is used to verify that the location of the fluid phase transition critical point within the simulation model coincides with the location of the sub-cold boiling initiation obtained from the test.

[0103] The determination of whether the root mean square error (RMSE) meets the convergence condition adopts a dual criterion combining global error and local characteristics. The first criterion is that the aforementioned local coincidence condition is met within the local comparison segment. The second criterion is that the global RMSE value monitored by the automated control script is less than the preset lower error threshold, or the absolute value of the change in the RMSE value within a consecutive preset number of iteration batches is lower than the smaller convergence value.

[0104] When both the first and second criteria are met simultaneously, the automated control script determines that the convergence condition has been met, issues a stop command to the optimization manager module, and terminates the iterative process of the fluid dynamics simulation software. The automated control script extracts the vaporization core density multiplier value and bubble escape diameter correction coefficient value corresponding to the final iteration batch that meets the convergence condition from the memory variables of the fluid dynamics simulation software.

[0105] The extracted vaporization core density multiplier values ​​and bubble escape diameter correction coefficient values ​​are combined and packaged into a standard simulation parameter set. This standard simulation parameter set is stored in a network-attached storage system using an Extensible Markup Language (EXPLAIN) file format. When performing subsequent thermal management conjugate heat transfer simulation tasks for the same engine model, the values ​​from the standard simulation parameter set are read according to the calling instructions, overriding the software's default boiling model parameters, thus completing the closed-loop solidification process of parameter calibration.

Claims

1. An automatic calibration method for water jacket CHT based on 3D printing and FBG, characterized in that, Includes the following steps: An engine prototype was formed using 3D printing technology, and a conformal microchannel was formed inside the metal body of the engine prototype. A fiber Bragg grating sensor with a polyimide coating was then inserted into the conformal microchannel. The engine sample was installed on an engine bench for operation testing. The temperature distribution data along the inner wall of the water jacket was collected by the fiber Bragg grating sensor to generate a temperature gradient curve of the test space. A conjugate heat transfer digital simulation model containing the conformal microchannel is established. A simulation probe line is defined in the conjugate heat transfer digital simulation model, and the temperature curve of the simulation probe line obtained by iterative calculation is extracted. The root mean square error between the simulated probe line temperature curve and the experimental space temperature gradient curve is calculated using an automated control script. An adaptive search algorithm is then used to reverse-correct the boiling model parameters in the conjugate heat transfer digital simulation model based on the root mean square error. The steps of extracting the temperature curve of the simulation probe line and correcting the boiling model parameters are repeated until the root mean square error meets the convergence condition, and the corresponding boiling model parameters are output to complete the calibration.

2. The automatic calibration method for water jacket CHT based on 3D printing and FBG according to claim 1, characterized in that, The specific steps of inserting the fiber Bragg grating sensor with a polyimide coating into the conformal microchannel include: The fiber Bragg grating sensor with polyimide coating is placed inside the space of the conformal microchannel; By sealing both ends of the conformal microchannel, the fiber Bragg grating sensor with polyimide coating is left with a gap between the conformal microchannel and the metal inner wall and is in a non-fixed, free state.

3. The automatic calibration method for water jacket CHT based on 3D printing and FBG according to claim 1, characterized in that, The specific steps of mounting the engine prototype on an engine bench for operational testing include: The test assembly, which includes the water jacket and the engine sample, is fixed to the engine test bench. Coolant is introduced into the water jacket and the engine test bench is run, causing the coolant to undergo subcooled boiling in the water jacket inner wall region corresponding to the conformal microchannel, resulting in a step change in the wall heat flux density.

4. The automatic calibration method for water jacket CHT based on 3D printing and FBG according to claim 1, characterized in that, The acquisition of spatial temperature distribution data along the inner wall of the water jacket using the fiber Bragg grating sensor, and the generation of a temperature gradient curve for the experimental space, specifically includes: Continuous high-frequency spatial temperature node data is acquired by the fiber Bragg grating sensor in a state where it is not in contact with the coolant. The continuous high-frequency spatial temperature node data is converted into a continuous spatial temperature gradient curve, and the temperature characteristic inflection point corresponding to the sudden change in wall heat flux density is marked on the continuous spatial temperature gradient curve as the experimental space temperature gradient curve.

5. The automatic calibration method for water jacket CHT based on 3D printing and FBG according to claim 1, characterized in that, The establishment of the conjugate heat transfer digital simulation model including the conformal microchannel specifically includes: Establish a conjugate heat transfer digital simulation model that includes both fluid and solid domains; Based on the actual three-dimensional coordinate dimensions of the conformal microchannel, a geometrically consistent model microchannel space is constructed within the solid domain of the conjugate heat transfer digital simulation model.

6. The automatic calibration method for water jacket CHT based on 3D printing and FBG according to claim 1, characterized in that, In the conjugate heat transfer digital simulation model, a simulation probe line is defined, and the temperature curve of the simulation probe line obtained from iterative calculation is extracted, specifically including: Based on the spatial coordinate correspondence, a simulation probe line with coincident coordinate positions is constructed in the conjugate heat transfer digital simulation model; Import the data table of the temperature gradient curve of the test space into the fluid dynamics simulation software according to the data table format; During the fluid dynamics iterative calculation process, the simulation calculation temperatures corresponding to multiple nodes on the simulation probe line are extracted to form the simulation probe line temperature curve.

7. The automatic calibration method for water jacket CHT based on 3D printing and FBG according to claim 1, characterized in that, The calculation of the root mean square error between the simulated probe line temperature curve and the experimental space temperature gradient curve using an automated control script specifically includes: The automated control script, written in Python, is combined with a network-attached storage system to call the data table of the temperature gradient curve in the test space; Extract the test temperature values ​​of the test nodes on the test space temperature gradient curve and the calculated temperature values ​​of the corresponding nodes on the simulated probe line temperature curve; The root mean square error is obtained by calculating the sum of squares of the differences between the test temperature value and the calculated temperature value, taking the average value, and then performing a square root operation.

8. The automatic calibration method for water jacket CHT based on 3D printing and FBG according to claim 1, characterized in that, The step of using an adaptive search algorithm to reverse-correct the boiling model parameters in the conjugate heat transfer digital simulation model based on the root mean square error specifically includes: Set the vaporization core density multiplier and the bubble detachment diameter correction coefficient as parameter variables; In the optimization manager of the fluid dynamics simulation software, the SHERPA hybrid adaptive search algorithm is invoked to dynamically adjust the distribution position of subsequent sampling points based on the root mean square error value output by the previous calculation. The adjusted parameter variables are written into the conjugate heat transfer digital simulation model for the next fluid dynamics iteration calculation.

9. The automatic calibration method for water jacket CHT based on 3D printing and FBG according to claim 1, characterized in that, The step of cyclically extracting the simulated probe line temperature curve and inversely correcting the boiling model parameters until the root mean square error meets the convergence condition specifically includes: The reverse-corrected boiling model parameters are input into the conjugate heat transfer digital simulation model to perform fluid dynamics iterative calculations. Extract the temperature curve of the simulated probe line output after the completion of the fluid dynamics iterative calculation; Compare the numerical differences between the simulated probe line temperature curve and the experimental space temperature gradient curve in the segment containing the temperature characteristic inflection point; When the root mean square error reaches a preset minimum value, and the simulated probe line temperature curve and the test space temperature gradient curve coincide in the segment containing the temperature characteristic inflection point, the convergence condition is determined to be met and the loop execution stops.

10. The automatic calibration method for water jacket CHT based on 3D printing and FBG according to claim 1, characterized in that, The calibration of the boiling model parameters corresponding to the output specifically includes: Extract the boiling model parameters from the conjugate heat transfer digital simulation model that satisfies the convergence condition; The boiling model parameters that meet the convergence conditions are set as the standard simulation parameter set; Store the standard simulation parameter set and set it as the input parameters for the conjugate heat transfer simulation of the thermal management of the corresponding engine model.

Images