A method for designing a double-wall load-bearing strut for a gas turbine

Abstract

The present application relates to the field of gas turbine design optimization, and particularly relates to a double-wall force support column design method for a gas turbine, comprising: obtaining complete constitutive curve data changing with temperature through high-temperature tensile experiment on all candidate materials; extracting principal stress direction vectors and forming a vector field; finally, after the skeleton model is solidified, assembling and performing Boolean operation with the heat shield entity and the fluid domains on the gas side and the cooling air side to obtain a complete double-wall initial calculation domain containing solid domains and fluid domains; establishing a three-dimensional coupled simulation model, and extracting key performance indicators through mesh division and multi-field solution on the simulation model. The present application forms a heat-force composite load path by combining heat flux and principal stress direction, and naturally forms a cooling air channel network in the low load area, thereby strengthening the heat dissipation effect through the internal cooling channel, realizing the collaborative optimization of mechanical bearing and thermal management, effectively improving the structural thermal stress distribution and improving the overall thermal stability.

Description

Technical Field

[0001] This invention relates to the field of gas turbine design optimization, and specifically to a design method for a double-walled load-bearing strut for gas turbines. Background Technology

[0002] As the power rating of gas turbines continues to increase, the initial gas temperature has entered a high-parameter era. The gas temperature in F-class gas turbines is around 1400℃, while the most advanced J-class gas turbines reach 1600℃. This significant increase in the initial gas temperature parameter presents severe challenges in gas turbine design, including material heat resistance, efficient cooling, and structural thermal stability. Among these challenges, the working environment of components such as load-bearing supports, which are in direct contact with high-temperature gas, is the most demanding.

[0003] As part of the gas turbine rotor support structure, the load-bearing strut is a crucial component connecting the "high-temperature gas passage" and the "low-temperature lubrication system." Its structure and function are highly complex and integrated: externally, it needs to connect to the casing that contains the high-temperature gas, and internally, it needs to precisely fit with the bearing housing that supports the rotor's rotation. This unique structural form dictates that the rotor support structure must "pass through" the high-temperature gas passage, directly bearing the continuous scouring of the high-temperature gas. Simultaneously, to ensure bearing lubrication and heat dissipation, the bearing cavity must maintain a low-temperature lubrication environment of 50℃-100℃. This "extremely large temperature difference between inside and outside" condition presents significant challenges to the design of the load-bearing strut. To overcome the design difficulties of load-bearing struts under high-temperature and large-temperature-difference conditions, this patent proposes a design method for a double-walled load-bearing strut for gas turbines. This method transforms the traditional single-structure load-bearing strut into a composite integrated structure with a heat insulation cover. The heat insulation cover separates the high-temperature gas from the load-bearing strut, preventing direct exposure and reducing the thermal load on the strut. This significantly improves the strength reserve and service life of the load-bearing strut. The design method for double-walled load-bearing columns proposed in this patent can guide the design work of subsequent new double-walled load-bearing columns, shorten the design cycle, and has important scientific significance and practical value. Summary of the Invention

[0004] This invention addresses the technical problems existing in the prior art by providing a design method for a double-walled load-bearing strut for gas turbines.

[0005] The technical solution of this invention to solve the above-mentioned technical problems is as follows: A design method for a double-walled load-bearing strut for a gas turbine, comprising the following steps: S1. After determining the global boundary conditions based on the rated operating conditions of the gas turbine, preliminary screening of candidate materials is carried out, and a set of candidate materials for load-bearing supports is obtained by constructing a three-dimensional interpolation surface of temperature-stress-life. Based on the temperature dependence curve of the thermal conductivity of the candidate materials, a set of candidate materials for heat insulation cover is formed, and complete constitutive curve data of temperature change are obtained for all candidate materials through high-temperature tensile tests. S2. Based on the overall structural layout of the gas turbine, read the absolute installation space coordinates of the load-bearing support and the geometry of the upper interface connected to the casing, as well as the geometry of the lower interface connected to the bearing housing. Construct a functional requirement skeleton model in the three-dimensional modeling environment, apply equivalent mechanical loads to the functional requirement skeleton model, extract the principal stress direction vectors, and form a vector field. S3. Based on the extreme values ​​of node coordinates in the circumferential and axial directions of the skeleton model, the maximum envelope size is determined and the envelope baseline is generated. On the windward side, the teardrop-shaped leading edge curve and teardrop-shaped cross-section curve are generated by the method of minimizing the pressure coefficient gradient integral. After generating the initial non-load-bearing teardrop-shaped heat shield inner surface model by surface lofting, the outer surface is generated by offsetting the unified initial thickness value along the normal direction at equal intervals based on the minimum thermal conductivity and minimum heat insulation gap value of the heat shield. Finally, the skeleton model is solidified and assembled with the heat shield solid and the fluid domains of the gas side and cooling air side by Boolean operation to obtain the complete double-wall initial calculation domain containing the solid domain and the fluid domain. S4. Based on the initial computational domain of the complete double-wall structure, perform thermal-mechanical synergistic variable density optimization of the load-bearing support. Design the connection form of the obtained solid model of the load-bearing support and the inner surface profile of the heat insulation cover, generate a color cloud map of the gap distribution, and perform Boolean fusion operation on the load-bearing support and the heat insulation cover based on the analysis of the color cloud map of the gap distribution to establish a three-dimensional coupled simulation model. Extract key performance indicators by meshing the simulation model and solving multiple fields.

[0006] In a preferred embodiment, S1 determines global boundary conditions based on the rated operating conditions of the gas turbine at the design point, including the total temperature, total pressure, and mass flow rate of the high-temperature gas, the characteristic curves of the gas viscosity, thermal conductivity, and specific heat capacity as a function of temperature obtained through gas component analysis, and the total temperature, total pressure, mass flow rate, and corresponding physical property parameter curves of the cooling air. Based on the parameters of global boundary conditions, preliminary screening of candidate materials is carried out. Specifically, based on the high-temperature material database and the functional requirements of the load-bearing column, the creep strength test data of high-temperature casting alloys mainly based on nickel and / or cobalt are retrieved. These data record the time until fracture of the material under different temperatures and stress levels. By constructing a three-dimensional interpolation surface of temperature-stress-life, isochrones corresponding to the design life time are intercepted on the three-dimensional interpolation surface. The material grade with the highest creep strength limit value in the working temperature range is found on the isochrones. At the same time, it is checked whether the oxidation weight gain rate of the material in the high-temperature oxygen environment is lower than the preset threshold. The material grade corresponding to the material grade is included in the candidate materials of the load-bearing column. Based on the functional requirements of the heat shield, the thermal conductivity temperature dependence curves and surface spectral emissivity data of the candidate materials are retrieved. The ratio of the average thermal conductivity to the average emissivity of each material within the operating temperature range is calculated. The deformable high-temperature alloys and / or ceramic matrix composites with the smallest ratio and good machinability are selected to form a set of candidate materials for the heat shield. Complete constitutive curves of elastic modulus, Poisson's ratio, coefficient of linear expansion, and yield strength as a function of temperature were obtained by fitting high-temperature tensile test data of all candidate materials.

[0007] In a preferred embodiment, step S2, based on the overall structural layout of the gas turbine, reads the absolute installation spatial coordinates of the load-bearing support, the geometry of the upper interface connected to the casing (including the spatial plane equation of the mating surface, the number, location, and diameter of bolt holes), and the geometry of the lower interface connected to the bearing housing from the overall design drawings. It then constructs a functional requirement skeleton model in a 3D modeling environment. Specifically, it establishes an upper interface reference point at the geometric center of the upper interface and creates its spatial coordinates in the 3D modeling environment. It also establishes a lower interface reference point at the geometric center of the lower interface, connects the upper and lower interface reference points based on a spatial curve (initially a straight line, serving as the core force transmission axis of the load-bearing support), and takes several equidistant cross-sectional positions from top to bottom based on the spatial curve. At each cross-sectional position, it calculates the required force according to the load-bearing support... The distance ratio between the upper and lower interfaces generates the cross-sectional profile. The cross-sectional profile is generated based on the principle that the cross-sectional shape is similar to the shape of the interface closer to the end. The cross-sectional area is determined according to the principle of equal strength design, that is, the cross-sectional area increases first and then decreases from the top to the middle and then to the end. The trend of change is assigned by the preliminary estimation of the axial force transmission path. All these cross-sectional profiles are connected by the lofting operation in 3D modeling to generate a hollow tubular geometry composed of facets. Based on the positions of the upper and lower interface reference points, the facet shape at the top of the hollow tubular geometry is adjusted to match the shape of the upper and lower interface joint surfaces. Nodes are created at each bolt hole position to obtain a functional requirement skeleton model composed of the upper and lower end faces and the connecting tubular curved surfaces. Its interior is hollow and its exterior shape is the preliminary mechanical transmission profile of the load-bearing column. Based on the functional requirement skeleton model, equivalent mechanical loads are applied to decompose the equivalent static forces and moments transmitted to the bearing housing via gas pressure and rotor vibration into nodal force vector groups acting on each bolt hole node at the bottom, based on the force translation theorem and the moment equivalence principle. After applying full-degree-of-freedom constraints to each bolt hole node at the top, linear static finite element method is used to solve the problem. The solution adopts shell element stiffness matrix assembly and elimination method based on thin shell theory. After the solution converges, the principal stress direction vectors are extracted one by one from the result file. The principal stress direction vectors are arranged according to their spatial positions to form a vector field. This vector field is the main path of mechanical transmission inside the load-bearing column. At the same time, the displacement amplitude response of the skeleton nodes is recorded, and the maximum value is taken as the initial reference of the overall stiffness.

[0008] In a preferred embodiment, S3 is based on the maximum and minimum absolute values ​​of all nodes of the functional requirement skeleton model in the circumferential coordinate of the gas turbine, and the maximum and minimum absolute values ​​in the axial coordinate, and adds a safety margin based on these extreme values. The safety margin is 15% of the difference between the extreme values, thereby obtaining the maximum envelope size of the load-bearing support in the circumferential and axial directions. Based on the upstream to downstream of the axial envelope of the load-bearing support in the direction perpendicular to the engine axis, several radial planes are taken at equal intervals to construct the initial non-load-bearing teardrop-shaped heat shield inner surface profile. After all the teardrop-shaped cross-sectional curves of the radial planes are generated, based on the teardrop-shaped cross-sectional curves as the cross-sectional profile and the engine axis as the lofting direction, a surface lofting operation is performed in the 3D modeling software to generate a continuous, smooth, and fully closed surface in 3D space. This is the initial non-load-bearing teardrop-shaped heat shield inner surface model.

[0009] In a preferred embodiment, step S3 calculates the equivalent thermal resistance based on the lowest thermal conductivity value and the minimum thermal insulation gap value among the candidate materials for the heat shield, and obtains a uniform initial thickness value based on the equivalent thermal resistance. Based on the initial non-load-bearing teardrop-shaped inner surface model of the heat shield, the uniform initial thickness value is offset outward along the normal direction of each surface using a surface normal equidistant algorithm to generate the outer surface model of the heat shield, forming a hollow sleeve structure of uniform thickness. The surface model of the skeleton of the functional requirement skeleton model is closed and solid voxels are generated, completing the solidification filling of the space occupied by the functional requirement skeleton model. The load-bearing support entity, the heat shield entity, the gas-side fluid domain formed by extending outward from the outer surface of the heat shield, and the cooling air-side fluid domain formed between the inner surface of the heat shield and the outer surface of the load-bearing support are assembled and Boolean operated in the three-dimensional modeling environment to obtain a complete double-wall initial calculation domain containing solid and fluid domains.

[0010] In a preferred embodiment, step S4 discretizes the solid space belonging to the load-bearing support in the initial computational domain of the complete double-walled structure into three-dimensional hexahedral finite elements. The total number of elements is determined by the computational resources and the expected resolution. Each three-dimensional hexahedral finite element is assigned a continuous design variable, the value of which is between a near-zero minimum value and a unit value representing a fully dense material. The minimum value is not zero, but is taken as one times ten to the power of negative cube of the unit value to avoid numerical singularities. When the design variable of any three-dimensional hexahedral finite element is less than the initial thickness value, the three-dimensional hexahedral finite element is marked as a cooling air channel. When the design variable of any three-dimensional hexahedral finite element is greater than or equal to the initial thickness value, the three-dimensional hexahedral finite element is marked as a load-bearing solid material. The heat flux vector and principal stress vector are extracted from the solid domain of the load-bearing column in three-dimensional hexahedral finite elements. After obtaining the heat flux vector and principal stress vector, the direction field of the thermo-mechanical composite load is generated by weighted fusion based on the energy ratio. The direction field of the thermo-mechanical composite load is then streamlined, and the set of all streamlines obtained is the path of the thermo-mechanical composite load. Finally, the direction field of the thermo-mechanical composite load is discretized by the isosurface extraction algorithm to obtain the pseudo-density field.

[0011] In a preferred embodiment, step S4 performs a fluid-thermal-solid three-field coupling solution based on the current pseudo-density field distribution, generating a continuous set of thermal-mechanical composite streamlines. Based on the physical field results, the temperature value at each finite element node within the solid domain of the load-bearing support is extracted, as is the equivalent stress value at each node. Based on the temperature and equivalent stress values, the heat flux vector of each three-dimensional hexahedral finite element is extracted. The candidate materials are guided to gather on the thermal-mechanical composite load path based on the thermal-mechanical composite streamlines, and the gathering amount is obtained according to the heat flux vector. After iterative optimization, the solid model of the load-bearing support is obtained.

[0012] In a preferred embodiment, step S4 designs the connection form based on the solid outline of the load-bearing pillar and the inner surface profile of the heat insulation cover. Based on several discrete local areas with negative gap values ​​in the gap distribution color cloud map, it performs overlapping, screw fastening, and through-hole creation matching. After completing the gap distribution color cloud map analysis, the solid model of the load-bearing pillar and the inner surface profile of the heat insulation cover are imported into the same part file. Based on the transition area between the solid model of the load-bearing pillar and the inner surface profile of the heat insulation cover, an intermediate equidistant surface is extracted, and a Gyroid-type curved surface is generated on the intermediate equidistant surface. The transition area is then merged with the solid model of the load-bearing pillar and the inner surface profile of the heat insulation cover using a Boolean fusion operation to form a part model composed of a single voxel.

[0013] In a preferred embodiment, S4 is based on a three-dimensional parametric modeling platform to assemble the solid model of the load-bearing pillar with the part model to obtain the gas-side fluid domain. Then, through Boolean operations, all solid domain entities are subtracted from the closed space between the inner surface of the heat insulation cover and the outer surface of the load-bearing pillar to establish a simulation model that includes the internal flow channel domain of the cooling air side and the gas-side fluid domain.

[0014] In a preferred embodiment, S4 uses a regional mesh to divide the outer surface regions of the heat shield and the load-bearing column based on the simulation model, generates a body-fitting anisotropic hexahedral structured mesh using the lay-up method, and performs fluid, heat transfer and mechanical solutions based on the anisotropic hexahedral structured mesh to obtain all physical quantities of the fluid domain, the complete three-dimensional temperature of the solid domain and the stress tensor of each mesh node as key performance indicators, and adjusts the teardrop-shaped leading edge curve, teardrop-shaped cross-sectional curve, inner surface profile of the heat shield and solid model of the load-bearing column based on the key performance indicators.

[0015] The beneficial effects of this invention are as follows: by adopting a variable density optimization method, heat flux and principal stress direction are integrated to form a thermal-mechanical composite load path, which guides the material to accumulate along the load transfer path. At the same time, a cooling air channel network is naturally formed in the low load area. Under the premise of ensuring the structural load-bearing capacity, the material utilization rate is improved, the structural weight is reduced, and the heat dissipation effect is enhanced through internal cooling channels. This achieves synergistic optimization of mechanical load-bearing and thermal management, effectively improves the distribution of structural thermal stress, and enhances the overall thermal stability. By accurately identifying assembly interference areas through gap distribution cloud maps, a reliable assembly of the heat insulation cover and load-bearing columns is achieved through overlapping and tight connections. At the same time, a smooth integration of the transition area is achieved through Gyroid-type curved surfaces, effectively alleviating the additional stress caused by temperature difference deformation incompatibility in the double-wall structure, improving the thermal fatigue resistance and structural integrity of the connection area, and constructing a functional requirement skeleton model based on actual installation interfaces and spatial constraints. The principal stress vector field is extracted through static solutions under equivalent loads, clarifying the main path of mechanical transmission within the load-bearing structure, and providing a clear force transmission basis for subsequent structural optimization. Attached Figure Description

[0016] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0017] The present invention will now be further described with reference to the accompanying drawings.

[0018] Please refer to the appendix. Figure 1 This embodiment provides a design method for a double-walled load-bearing strut for a gas turbine, comprising the following steps: S1. After determining the global boundary conditions based on the rated operating conditions of the gas turbine, preliminary screening of candidate materials is carried out, and a set of candidate materials for load-bearing supports is obtained by constructing a three-dimensional interpolation surface of temperature-stress-life. Based on the temperature dependence curve of the thermal conductivity of the candidate materials, a set of candidate materials for heat insulation cover is formed, and complete constitutive curve data of temperature change are obtained for all candidate materials through high-temperature tensile tests. S1 determines the global boundary conditions based on the rated operating conditions of the gas turbine at the design point, including the total temperature, total pressure, and mass flow rate of the high-temperature gas, the characteristic curves of the gas viscosity, thermal conductivity, and specific heat capacity as a function of temperature obtained through gas component analysis, and the total temperature, total pressure, and mass flow rate of the cooling air and their corresponding physical property curves. Based on the parameters of global boundary conditions, preliminary screening of candidate materials is carried out. Specifically, based on the high-temperature material database and the functional requirements of the load-bearing column, the creep strength test data of high-temperature casting alloys mainly based on nickel and / or cobalt are retrieved. These data record the time until fracture of the material under different temperatures and stress levels. By constructing a three-dimensional interpolation surface of temperature-stress-life, isochrones corresponding to the design life time are intercepted on the three-dimensional interpolation surface. The material grade with the highest creep strength limit value in the working temperature range is found on the isochrones. At the same time, it is checked whether the oxidation weight gain rate of the material in the high-temperature oxygen environment is lower than the preset threshold. The material grade corresponding to the material grade is included in the candidate materials of the load-bearing column. Based on the functional requirements of the heat shield, the thermal conductivity temperature dependence curves and surface spectral emissivity data of the candidate materials are retrieved. The ratio of the average thermal conductivity to the average emissivity of each material within the operating temperature range is calculated. The deformable high-temperature alloys and / or ceramic matrix composites with the smallest ratio and good machinability are selected to form a set of candidate materials for the heat shield. Complete constitutive curves of elastic modulus, Poisson's ratio, coefficient of linear expansion, and yield strength as a function of temperature were obtained by fitting high-temperature tensile test data of all candidate materials.

[0019] S2. Based on the overall structural layout of the gas turbine, read the absolute installation space coordinates of the load-bearing support and the geometry of the upper interface connected to the casing, as well as the geometry of the lower interface connected to the bearing housing. Construct a functional requirement skeleton model in the three-dimensional modeling environment, apply equivalent mechanical loads to the functional requirement skeleton model, extract the principal stress direction vectors, and form a vector field. Based on the overall structural layout of the gas turbine, S2 retrieves the absolute installation spatial coordinates of the load-bearing struts and the geometry of the upper interface connecting to the casing from the overall design drawings, including the spatial plane equation of the mating surface, the number, location, and diameter of bolt holes, and the geometry of the lower interface connecting to the bearing housing. A functional requirement skeleton model is then constructed in the 3D modeling environment. Specifically, an upper interface reference point is established at the geometric center of the upper interface, and its spatial coordinates are created based on the 3D modeling environment. A lower interface reference point is also established at the geometric center of the lower interface. A spatial curve connects the upper and lower interface reference points; the initial shape of this curve is a straight line, serving as the core force transmission axis of the load-bearing strut. Several equidistant cross-sectional positions are taken from top to bottom based on the spatial curve, and at each cross-sectional position, the geometry is determined according to the connections to the upper and lower interfaces. The distance ratio generates the cross-sectional profile. The cross-sectional profile is generated based on the principle that the cross-sectional shape is similar to the shape of the interface at the closer end. The cross-sectional area is determined according to the principle of equal strength design, that is, from the top to the middle and then to the end, the cross-sectional area first increases and then decreases. The trend of change is assigned by the preliminary estimation of the axial force transmission path. All these cross-sectional profiles are connected by the lofting operation in 3D modeling to generate a hollow tubular geometry composed of facets. Based on the positions of the upper and lower interface reference points, the facet shape at the top of the hollow tubular geometry is adjusted to match the shape of the upper and lower interface joint surfaces. Nodes are created at each bolt hole position to obtain a functional requirement skeleton model composed of the upper and lower end faces and the connecting tubular curved surfaces. Its interior is hollow and its exterior shape is the preliminary mechanical transmission profile of the load-bearing column. Based on the functional requirement skeleton model, equivalent mechanical loads are applied to decompose the equivalent static forces and moments transmitted to the bearing housing via gas pressure and rotor vibration into nodal force vector groups acting on each bolt hole node at the bottom, based on the force translation theorem and the moment equivalence principle. After applying full-degree-of-freedom constraints to each bolt hole node at the top, linear static finite element method is used to solve the problem. The solution adopts shell element stiffness matrix assembly and elimination method based on thin shell theory. After the solution converges, the principal stress direction vectors are extracted one by one from the result file. The principal stress direction vectors are arranged according to their spatial positions to form a vector field. This vector field is the main path of mechanical transmission inside the load-bearing column. At the same time, the displacement amplitude response of the skeleton nodes is recorded, and the maximum value is taken as the initial reference of the overall stiffness.

[0020] S3. Based on the extreme values ​​of node coordinates in the circumferential and axial directions of the skeleton model, the maximum envelope size is determined and the envelope baseline is generated. On the windward side, the teardrop-shaped leading edge curve and teardrop-shaped cross-section curve are generated by the method of minimizing the pressure coefficient gradient integral. After generating the initial non-load-bearing teardrop-shaped heat shield inner surface model by surface lofting, the outer surface is generated by offsetting the unified initial thickness value along the normal direction at equal intervals based on the minimum thermal conductivity and minimum heat insulation gap value of the heat shield. Finally, the skeleton model is solidified and assembled with the heat shield solid and the fluid domains of the gas side and cooling air side by Boolean operation to obtain the complete double-wall initial calculation domain containing the solid domain and the fluid domain. S3 is based on the maximum and minimum absolute values ​​of all nodes in the functional requirement skeleton model of the gas turbine in the circumferential coordinate and the axial coordinate. Based on these extreme values, a safety margin is added, which is 15% of the difference between the extreme values, so as to obtain the maximum envelope size of the load-bearing support in the circumferential and axial directions. Based on the uppermost to the lowermost end of the axial envelope range of the load-bearing support in the direction perpendicular to the engine axis, take several radial planes at equal intervals to construct the initial non-load-bearing teardrop-shaped heat shield inner surface profile. After all the teardrop-shaped cross-sectional curves of the radial planes are generated, based on the teardrop-shaped cross-sectional curves as the cross-sectional profile and the engine axis as the lofting direction, perform a surface lofting operation in the 3D modeling software to generate a continuous, smooth, and fully closed surface in 3D space. This is the initial non-load-bearing teardrop-shaped heat shield inner surface model. It should be noted that the initial non-load-bearing teardrop-shaped heat insulation cover inner surface profile includes: Step 1: Based on each radial plane, establish a local two-dimensional coordinate system on the plane, with the origin located at the intersection of the engine axis and the plane. The horizontal axis represents the circumferential arc length, and the vertical axis represents the radial height. Project the intersection line of the skeleton model from Step 2 and the plane onto this local coordinate system to obtain a set of closed or open contour points, which serves as the area occupied by the load-bearing support on the plane. Step 2: Generate the envelope baseline based on the occupied area. This is achieved using a two-dimensional rolling sphere algorithm: A sphere is set with a radius equal to a preset minimum thermal insulation gap value. This minimum thermal insulation gap value is determined as follows: using the thermal resistance requirement of the candidate thermal insulation material obtained in Step 1 at the lowest thermal conductivity, combined with steady-state one-dimensional heat conduction estimation, the required insulation distance from the gas temperature to the target temperature on the inner surface of the thermal insulation cover is taken as the larger of this distance and the structural assembly tolerance. The virtual rolling sphere is then rolled along the outer boundary of the occupied area contour. The trajectory swept by the center of the rolling sphere is an equidistant line at a distance from the occupied area contour that is exactly one minimum thermal insulation gap. This equidistant line is the envelope baseline. Step 3: Generate a teardrop-shaped leading edge curve on the windward side of the envelope baseline. First, calculate the outward normal direction at the leading edge point of the envelope baseline. Starting from this leading edge point, extend forward along the outward normal direction by a leading edge extension amount, which is half of the maximum circumferential width of the envelope baseline. Using this extension point as the new leading edge vertex and the leading edge point of the envelope baseline as the tangent point, generate a second-order Bézier curve using the Bézier curve construction method. The location of its control point is determined by solving a minimization problem: using the coordinates of the control point as variables and minimizing the integral of the pressure coefficient gradient of the curve in the direction of the gas flow as the objective, the control point location is searched using the coordinate rotation method, thus generating a convex curve from the leading edge vertex to the leading edge point of the envelope baseline with the pressure distribution as gentle as possible as the teardrop-shaped leading edge. Step 4: Generate a teardrop-shaped trailing edge curve on the leeward side of the envelope baseline. Starting from the maximum width point of the envelope baseline, extend it backward along the flow direction to a predetermined trailing edge endpoint. The flow direction distance from the trailing edge endpoint to the trailing edge point of the envelope baseline is twice the maximum circumferential width of the envelope baseline. From the maximum width point to the trailing edge endpoint, generate a cubic spline curve. The starting tangent of this spline curve is continuous with the tangent of the envelope baseline at the maximum width point, and the ending tangent points in the opposite direction of the incoming flow. The radius of curvature at the endpoint is set to one-fifth of the maximum width of the envelope baseline to ensure that the trailing edge angle is within an acute angle range that can suppress flow separation. Determine all control points of the spline curve by solving the cubic spline interpolation equation, thereby generating the leeward side profile.

[0021] Step 5: Combine the leading edge curve on the windward side, the middle section of the envelope baseline, and the trailing edge curve on the leeward side into a closed teardrop-shaped cross-sectional curve.

[0022] S3 calculates the equivalent thermal resistance based on the lowest thermal conductivity value and the minimum thermal insulation gap value among the candidate materials for the heat shield. Based on the equivalent thermal resistance, a unified initial thickness value is obtained. Based on the initial non-load-bearing teardrop-shaped inner surface model of the heat shield, the unified initial thickness value is offset outward along the normal direction of each surface using the curved surface normal equidistant algorithm to generate the outer surface model of the heat shield, forming a hollow sleeve structure of equal thickness. The surface model of the skeleton of the functional requirement skeleton model is closed and solid voxels are generated to complete the solidification filling of the space occupied by the functional requirement skeleton model. The load-bearing support solid, the heat shield solid, the gas side fluid domain formed by the outer surface of the heat shield extending outward, and the cooling air side fluid domain formed between the inner surface of the heat shield and the outer surface of the load-bearing support are assembled and Boolean operations are performed in the 3D modeling environment to obtain the complete double-wall initial calculation domain containing solid domain and fluid domain. The process for determining the initial thickness value is as follows: using the thermal conductivity value of the candidate heat shield material with the lowest thermal conductivity, combined with the determined minimum heat insulation gap value as the characteristic length, the ratio of this characteristic length to the thermal conductivity is calculated as the equivalent thermal resistance; assuming that the outer surface of the heat shield is subjected to the full temperature of the combustion gas and the inner surface is subjected to half the temperature of the cooling air, under the assumption of steady-state one-dimensional heat conduction, the required heat shield thickness is estimated by dividing the equivalent thermal resistance by the heat flux density, where the heat flux density is obtained by dividing the difference between the combustion gas temperature and the cooling air temperature by the total thermal resistance. This estimated value is then rounded up to the nearest integer multiple of 0.1 millimeters to obtain a uniform initial thickness value.

[0023] S4. Based on the initial computational domain of the complete double-wall structure, perform thermal-mechanical synergistic variable density optimization of the load-bearing support. Design the connection form of the obtained solid model of the load-bearing support and the inner surface profile of the heat insulation cover, generate a color cloud map of the gap distribution, and perform Boolean fusion operation on the load-bearing support and the heat insulation cover based on the analysis of the color cloud map of the gap distribution to establish a three-dimensional coupled simulation model. Extract key performance indicators by meshing the simulation model and solving multiple fields.

[0024] S4 discretizes the solid space belonging to the load-bearing support in the initial computational domain of the complete double-walled structure into three-dimensional hexahedral finite elements. The total number of elements is determined by the computational resources and the expected resolution. Each three-dimensional hexahedral finite element is assigned a continuous design variable, the value of which is between a near-zero minimum value and a unit value representing fully dense material. The minimum value is not zero, but is taken as one times ten to the power of negative cube of the unit value to avoid numerical singularities. When the design variable of any three-dimensional hexahedral finite element is less than the initial thickness value, the three-dimensional hexahedral finite element is marked as a cooling air channel. When the design variable of any three-dimensional hexahedral finite element is greater than or equal to the initial thickness value, the three-dimensional hexahedral finite element is marked as a load-bearing solid material. The heat flux vector and principal stress vector are extracted from the solid domain of the load-bearing column in three-dimensional hexahedral finite elements. After obtaining the heat flux vector and principal stress vector, the direction field of the thermo-mechanical composite load is generated by weighted fusion based on the energy ratio. The direction field of the thermo-mechanical composite load is then streamlined, and the set of all streamlines obtained is the path of the thermo-mechanical composite load. Finally, the direction field of the thermo-mechanical composite load is discretized by the isosurface extraction algorithm to obtain the pseudo-density field.

[0025] The specific operation is as follows: In the post-processing stage, the heat flux vector is extracted from the solid domain of the load-bearing support element by element. This heat flux vector is obtained by multiplying the temperature gradient at the center point of the element by the equivalent thermal conductivity of the element point by point. Its direction points to the direction of the fastest temperature drop, and its modulus represents the density of heat transfer at that point. A weighted fusion algorithm based on energy ratio is adopted, specifically as follows: For each element, its thermal strain energy density and mechanical strain energy density are calculated. The thermal strain energy density is obtained by integrating the element's temperature, coefficient of thermal expansion, and thermal stress, while the mechanical strain energy density is obtained by integrating the element's mechanical stress and mechanical strain. The ratio of the thermal strain energy density to the total strain energy density is used as the thermal weight, and the ratio of the mechanical strain energy density to the total strain energy density is used as the force weight. Then, the heat flux vector and principal stress vector are weighted and summed: the heat flux vector is multiplied by the force weight, and the principal stress vector is multiplied by the thermal weight. The sum of the two vectors yields a new vector that is neither purely heat flux nor purely stress direction. This process is repeated for all elements, and the resulting vector field is the direction field of the thermo-mechanical composite load. S4 performs a fluid-thermal-solid three-field coupled solution based on the current pseudo-density field distribution, generating a continuous set of thermo-mechanical composite streamlines. The solution process is as follows: a steady-state solution is performed on the fluid region in the entire computational domain, solving for the Reynolds-averaged mass conservation equation, momentum conservation equation, and energy conservation equation. A shear stress transport model is used for turbulent closure. The solver discretizes the governing equations using the finite volume method, employs a coupled implicit scheme for time advancement, and utilizes algebraic multigrid technology to accelerate convergence. After the solution is completed, the velocity vector, static pressure, total pressure, and static temperature values ​​at each volume center of the fluid domain are output. In the solid domain of the load-bearing column, the steady-state heat conduction equation is solved. At the fluid-solid coupling interface, the solid side receives the convective heat transfer coefficient and near-wall fluid temperature from the fluid side, and the convective heat flux density is calculated. Simultaneously, on the solid wall of the gas combustion engine, radiative heat transfer is solved. Based on the wall emissivity and angular coefficient matrix, the ray tracing method is used to calculate the radiative heat transfer between the walls and between the wall and the gas combustion engine. The non-uniform temperature values ​​of all nodes in the solid domain and the pressure values ​​of the fluid domain acting on all fluid-solid interfaces are used as load inputs. Based on the thermoelastic small deformation assumption, the displacement finite element method is used to assemble the stiffness matrix and load vector including thermal strain contributions. The displacement solutions of all nodes are obtained through a sparse matrix direct solver. Then, the strain tensor and stress tensor at all element integration points are calculated based on the displacement solutions. At this point, the three-field coupling solution is completed. Based on the physical field results, the temperature value at each finite element node within the solid domain of the load-bearing column is extracted, as is the equivalent stress value at each node. Based on the temperature and equivalent stress values, the heat flux vector for each three-dimensional hexahedral finite element is extracted. The calculation process for the heat flux vector involves calculating the average temperature gradient at the eight nodes of the element, which is then used as the temperature gradient vector for that element. The magnitude of this temperature gradient vector is then multiplied by the equivalent thermal conductivity of the element, which is determined by the pseudo-density value and the thermal conductivity of the dense material according to a penalty relationship. The direction of the heat flux vector is taken as the opposite direction of the temperature gradient vector. Finally, the principal stress vector for each element is extracted, taking the direction corresponding to the principal stress with the largest absolute value in the element's stress tensor. Based on the thermo-mechanical composite streamline guiding the candidate material to gather on the thermo-mechanical composite load path, and the gathering amount is obtained according to the heat flux vector, after iterative optimization, the solid model of the load-bearing column is obtained, which contains a non-uniform cooling channel network that conforms to the thermo-mechanical composite path.

[0026] S4 designs the connection form based on the solid outline of the load-bearing column and the inner surface profile of the heat insulation cover. First, it calculates the spatial distance field between the outer surface of the load-bearing column and the inner surface of the heat insulation cover. A large number of dense points are sampled on the outer surface of the load-bearing column. For each sampling point, a ray is emitted along its outer normal. The intersection of the ray and the inner surface of the heat insulation cover is calculated. The Euclidean distance between the sampling point and the intersection point is the gap value at that point. A color cloud map of the gap distribution on the entire outer surface is generated. Based on the gap distribution color cloud map, several discrete local areas with negative gap values ​​are identified and matched by overlapping, screw fastening, and through-hole creation. Specifically, in most areas of the global region, the gap is positive, and there are several discrete local areas with negative gap values, representing geometric interference. These local areas are identified and numbered. For each such area, a cylindrical or square overlapping boss is generated on the side of the load-bearing column. The height of the boss extends to be flush with the inner surface of the heat insulation cover, and a bolt through-hole is created in the center of the boss. Matching through-holes are created at the corresponding positions of the heat insulation cover. The overlapping-screw fastening method is adopted, and high-temperature alloy bolts are used for connection. A corrugated thin metal gasket is designed between the overlapping boss and the inner surface of the heat insulation cover. After completing the color cloud map analysis of the gap distribution, the solid model of the load-bearing column and the inner surface profile of the heat insulation cover are imported into the same part file. Based on the transition area between the solid model of the load-bearing column and the inner surface profile of the heat insulation cover, the intermediate equidistant surface is extracted, and a Gyroid-type surface is generated on the intermediate equidistant surface. The transition area is then merged with the solid model of the load-bearing column and the inner surface profile of the heat insulation cover using a Boolean fusion operation to form a part model composed of a single voxel.

[0027] Based on a 3D parametric modeling platform, S4 assembles the solid model of the load-bearing support with the part models. In the assembly environment, an envelope of the gas meridional flow channel is created. The radial range of the envelope extends from the hub to the casing, and the axial range covers the area from three times the chord length upstream to five times the chord length downstream of the load-bearing support. Using Boolean subtraction, the solid models of the load-bearing support and the heat shield are subtracted from the envelope to obtain the gas-side fluid domain. Then, through Boolean operations, all solid domain entities are subtracted from the closed space between the inner surface of the heat shield and the outer surface of the load-bearing support to establish a simulation model that includes the internal flow channel domain of the cooling air side and the gas-side fluid domain.

[0028] S4 uses a regional meshing method to divide the outer surface regions of the heat shield and load-bearing columns based on the simulation model. It generates a body-fitting anisotropic hexahedral structured mesh using the lay-up method. Based on the anisotropic hexahedral structured mesh, it performs fluid, heat transfer and mechanical solutions to obtain all physical quantities of the fluid domain, the complete three-dimensional temperature of the solid domain and the stress tensor of each mesh node, which are key performance indicators. Based on the key performance indicators, it adjusts the teardrop-shaped leading edge curve, teardrop-shaped cross-sectional curve, inner surface profile of the heat shield and solid model of the load-bearing column.

[0029] First, this application initiates a fluid dynamics solution, simultaneously performing steady-state solutions for the fluid domains on both the gas and cooling air sides. A second-order upwind scheme is used for discretization, and a coupled algorithm is employed for pressure-velocity coupling. The convergence criterion is a decrease of five orders of magnitude in the root mean square residuals of all governing equations. Simultaneously, the average heat transfer coefficient of the load-bearing support wall is monitored; convergence is determined when its relative change is less than 0.05% in the last 500 iterations. All physical quantities of the fluid domain are output.

[0030] The heat transfer solution is initiated, receiving the near-wall temperature and convective heat transfer coefficient from the fluid domain, and calculating convective heat transfer. The radiation module is then invoked, and based on the banded gray body model, the infrared radiation band is discretized into four spectral bands. Within each band, the radiation transfer equation is solved to calculate the radiative heat transfer between the walls and the radiative absorption and emission of the participating media (water vapor, carbon dioxide) in the combustion gas. The interface temperature and heat flux between the fluid and solid are iterated repeatedly until the maximum temperature difference between the interface between two adjacent coupled iteration steps is less than 0.1 Kelvin. At this point, the heat transfer field converges, and the complete three-dimensional temperature distribution of the solid domain is output.

[0031] The mechanical solution is initiated, reading in the temperature fields of all nodes in the solid domain as thermal loads and the static pressure field output by the fluid solver at the fluid-solid interface as pressure loads. The solver assembles the global stiffness matrix based on isotropic linear elastic-thermoelastic constitutive relations. Constraints are applied to all six degrees of freedom at the top interface of the load-bearing column. A parallel direct sparse solver is used to solve the displacement equations. After obtaining the displacement solutions, the displacement vector of each node and the six-component stress tensor at each element integration point are calculated and output.

Claims

1. A design method for a double-walled load-bearing strut for a gas turbine, characterized in that, Includes the following steps: S1. After determining the global boundary conditions based on the rated operating conditions of the gas turbine, preliminary screening of candidate materials is carried out, and a set of candidate materials for load-bearing supports is obtained by constructing a three-dimensional interpolation surface of temperature-stress-life. Based on the temperature dependence curve of the thermal conductivity of the candidate materials, a set of candidate materials for heat insulation cover is formed, and complete constitutive curve data of temperature change are obtained for all candidate materials through high-temperature tensile tests. S2. Based on the overall structural layout of the gas turbine, read the absolute installation space coordinates of the load-bearing support and the geometry of the upper interface connected to the casing, as well as the geometry of the lower interface connected to the bearing housing. Construct a functional requirement skeleton model in the three-dimensional modeling environment, apply equivalent mechanical loads to the functional requirement skeleton model, extract the principal stress direction vectors, and form a vector field. S3. Based on the extreme values ​​of node coordinates in the circumferential and axial directions of the skeleton model, the maximum envelope size is determined and the envelope baseline is generated. On the windward side, the teardrop-shaped leading edge curve and teardrop-shaped cross-section curve are generated by the method of minimizing the pressure coefficient gradient integral. After generating the initial non-load-bearing teardrop-shaped heat shield inner surface model by surface lofting, the outer surface is generated by offsetting the unified initial thickness value along the normal direction at equal intervals based on the minimum thermal conductivity and minimum heat insulation gap value of the heat shield. Finally, the skeleton model is solidified and assembled with the heat shield solid and the fluid domains of the gas side and cooling air side by Boolean operation to obtain the complete double-wall initial calculation domain containing the solid domain and the fluid domain. S4. Based on the initial computational domain of the complete double-wall structure, perform thermal-mechanical synergistic variable density optimization of the load-bearing support. Design the connection form of the obtained solid model of the load-bearing support and the inner surface profile of the heat insulation cover, generate a color cloud map of the gap distribution, and perform Boolean fusion operation on the load-bearing support and the heat insulation cover based on the analysis of the color cloud map of the gap distribution to establish a three-dimensional coupled simulation model. Extract key performance indicators by meshing the simulation model and solving multiple fields.

2. The design method for a double-walled load-bearing strut for a gas turbine according to claim 1, characterized in that, S1 determines the global boundary conditions, including the high-temperature gas characteristic curve and the cooling air physical property parameter curve, based on the rated operating conditions of the gas turbine at the design point. Based on the parameters of global boundary conditions, preliminary screening of candidate materials is carried out. Specifically, based on the high-temperature material database and the functional requirements of the load-bearing column, the creep strength test data of high-temperature casting alloys mainly based on nickel and / or cobalt are retrieved. By constructing a three-dimensional interpolation surface of temperature-stress-life, isochrones corresponding to the design life time are intercepted on the three-dimensional interpolation surface. The material grade with the highest creep strength limit value in the working temperature range is found on the isochrone, and the material grade corresponding to the material grade is included in the candidate materials of the load-bearing column. Based on the functional requirements of the heat shield, the thermal conductivity temperature dependence curves and surface spectral emissivity data of the candidate materials are retrieved. The ratio of the average thermal conductivity to the average emissivity of each material within the operating temperature range is calculated. The deformable high-temperature alloys and / or ceramic matrix composites with the smallest ratio and good machinability are selected to form a set of candidate materials for the heat shield. Complete constitutive curve data were obtained by fitting high-temperature tensile test data of all candidate materials.

3. The design method for a double-walled load-bearing strut for a gas turbine according to claim 1, characterized in that, S2, based on the overall structural layout of the gas turbine, reads the absolute installation spatial coordinates of the load-bearing support, the geometry of the upper interface connected to the casing, and the geometry of the lower interface connected to the bearing housing from the overall design drawings, and constructs a functional requirement skeleton model in the 3D modeling environment. Specifically, it establishes an upper interface reference point at the geometric center of the upper interface and creates the spatial coordinates of the upper interface reference point based on the 3D modeling environment. It also establishes a lower interface reference point at the geometric center of the lower interface, connects the upper and lower interface reference points based on a spatial curve, and takes several cross-sectional positions at equal intervals from top to bottom based on the spatial curve. At each cross-sectional position, it generates a cross-sectional profile according to the distance ratio to the upper and lower interfaces. It connects all these cross-sectional profiles through a lofting operation in 3D modeling to generate a hollow tubular geometry composed of facets. Based on the positions of the upper and lower interface reference points, it adjusts the shape of the facets at the top of the hollow tubular geometry to match the shapes of the upper and lower interface mating surfaces. It creates nodes at each bolt hole position to obtain a functional requirement skeleton model composed of upper and lower end faces and connecting tubular curved surfaces. Based on the functional requirement skeleton model, equivalent mechanical loads are applied to decompose the equivalent static forces and moments transmitted to the bearing housing via gas pressure and rotor vibration into nodal force vector groups acting on each bolt hole node at the bottom, based on the translation theorem of forces and the principle of moment equivalence. After applying full degree of freedom constraints to each bolt hole node at the top, the solution is obtained by linear static finite element method. After the solution converges, the principal stress direction vectors are extracted one by one from the result file, and the principal stress direction vectors are arranged according to their spatial positions to form a vector field.

4. The design method for a double-walled load-bearing strut for a gas turbine according to claim 1, characterized in that, The S3 is based on the maximum and minimum absolute values ​​of all nodes in the functional requirement skeleton model in the circumferential coordinate of the gas turbine, as well as the maximum and minimum absolute values ​​in the axial coordinate, and obtains the maximum envelope size of the load-bearing support in the circumferential and axial directions according to the safety margin. Based on the uppermost to the lowermost end of the axial envelope range of the load-bearing support in the direction perpendicular to the engine axis, take several radial planes at equal intervals to construct the initial non-load-bearing teardrop-shaped heat shield inner surface profile. After all the teardrop-shaped cross-sectional curves of the radial planes are generated, based on the teardrop-shaped cross-sectional curves as the cross-sectional profile and the engine axis as the lofting direction, perform surface lofting operation in the 3D modeling software to generate the non-load-bearing teardrop-shaped heat shield inner surface model.

5. The design method for a double-walled load-bearing strut for a gas turbine according to claim 4, characterized in that, S3 calculates the equivalent thermal resistance based on the thermal conductivity, minimum thermal conductivity, and minimum thermal gap values ​​of the candidate materials for the heat shield. Based on the equivalent thermal resistance, a uniform initial thickness value is obtained. Based on the initial non-load-bearing teardrop-shaped inner surface model of the heat shield, the uniform initial thickness value is offset outward along the normal direction of each surface using a surface normal equidistant algorithm to generate the outer surface model of the heat shield, forming a hollow sleeve structure of uniform thickness. The surface model of the skeleton of the functional requirement skeleton model is closed and solid voxels are generated, completing the solidification filling of the space occupied by the functional requirement skeleton model. The load-bearing support entity, the heat shield entity, the gas-side fluid domain formed by extending outward from the outer surface of the heat shield, and the cooling air-side fluid domain formed between the inner surface of the heat shield and the outer surface of the load-bearing support are assembled and Boolean operations are performed in the three-dimensional modeling environment to obtain a complete double-wall initial calculation domain containing solid and fluid domains.

6. The design method for a double-walled load-bearing strut for a gas turbine according to claim 5, characterized in that, S4 discretizes the solid space belonging to the load-bearing column in the initial calculation domain of the complete double-walled structure into three-dimensional hexahedral finite elements, and assigns continuous design variables to each three-dimensional hexahedral finite element. When the design variable of any three-dimensional hexahedral finite element is less than the initial thickness value, the three-dimensional hexahedral finite element is marked as a cooling air channel. When the design variable of any three-dimensional hexahedral finite element is greater than or equal to the initial thickness value, the three-dimensional hexahedral finite element is marked as a load-bearing solid material. The heat flux vector and principal stress vector are extracted from the solid domain of the load-bearing column in three-dimensional hexahedral finite elements. After obtaining the heat flux vector and principal stress vector, the direction field of the thermo-mechanical composite load is generated by weighted fusion based on the energy ratio. The direction field of the thermo-mechanical composite load is then streamlined, and the set of all streamlines obtained is the path of the thermo-mechanical composite load. Finally, the direction field of the thermo-mechanical composite load is discretized by the isosurface extraction algorithm to obtain the pseudo-density field.

7. The design method for a double-walled load-bearing strut for a gas turbine according to claim 1, characterized in that, S4 performs a fluid-thermal-solid three-field coupling solution based on the current pseudo-density field distribution, generating a continuous set of thermal-mechanical composite streamlines; Based on the physical field results, the temperature value at each finite element node in the solid domain of the load-bearing column is extracted, the equivalent stress value at each node is extracted, and the heat flux vector of each three-dimensional hexahedral finite element is extracted based on the temperature value and the equivalent stress value. Candidate materials are guided to aggregate on the thermal-mechanical composite load path based on the thermal-mechanical composite streamline, and the aggregation amount is obtained according to the heat flux vector. After iterative optimization, the solid model of the load-bearing column is obtained.

8. The design method for a double-walled load-bearing strut for a gas turbine according to claim 1, characterized in that, S4 is designed based on the solid outline of the load-bearing column and the inner surface profile of the heat insulation cover to generate a color cloud map of the gap distribution. Based on several discrete local areas with negative gap values ​​existing in the gap distribution color cloud map, overlap-screw fastening and through-hole matching are performed; After completing the color cloud map analysis of the gap distribution, the solid model of the load-bearing column and the inner surface profile of the heat insulation cover are imported into the same part file. Based on the transition area between the solid model of the load-bearing column and the inner surface profile of the heat insulation cover, the intermediate equidistant surface is extracted, and a Gyroid-type surface is generated on the intermediate equidistant surface. The transition area is then merged with the solid model of the load-bearing column and the inner surface profile of the heat insulation cover using a Boolean fusion operation to form a part model composed of a single voxel.

9. The design method for a double-walled load-bearing strut for a gas turbine according to claim 1, characterized in that, The S4 model is based on a three-dimensional parametric modeling platform. It assembles the solid model of the load-bearing support column with the part model to obtain the gas-side fluid domain. Then, through Boolean operations, it subtracts all solid domain entities from the closed space between the inner surface of the heat shield and the outer surface of the load-bearing support column to establish a simulation model that includes the internal flow channel domain of the cooling air side and the gas-side fluid domain.

10. A design method for a double-walled load-bearing strut for a gas turbine according to claim 9, characterized in that, The S4 model uses a regional mesh to divide the outer surface regions of the heat shield and the load-bearing column based on the simulation model. It generates a body-fitting anisotropic hexahedral structured mesh using the lay-up method. Based on the anisotropic hexahedral structured mesh, it performs fluid, heat transfer, and mechanical solutions to obtain all physical quantities of the fluid domain, the complete three-dimensional temperature of the solid domain, and the stress tensor of each mesh node, which are key performance indicators. Based on the key performance indicators, it adjusts the teardrop-shaped leading edge curve, the teardrop-shaped cross-sectional curve, the inner surface profile of the heat shield, and the solid model of the load-bearing column.

Images