Aerodynamic topological optimization method for airfoil design

Abstract

The invention discloses an aerodynamic topology optimization method for airfoil profile design. The method comprises the following steps: generating a plurality of normal airfoil profiles, namely initial solutions; generating an initial sub-domain around each initial solution according to a specified interval length, and performing Latin hypercube sampling in the initial sub-domain to obtain a plurality of samples; performing grid division on each sample obtained by sampling to obtain a grid sample, inputting a preset Mach number, preset resistance and preset lift force into the grid sample, and outputting a penalty objective function; and constructing a proxy model for the penalty target function, and performing sub-domain optimization on the proxy model to obtain an optimized penalty target function. The method does not depend on the initial shape any more, the design space is huge, the design freedom degree is extremely high, and an existing aerodynamic layout design normal form is broken through.

Description

Technical Field

[0001] The invention belongs to the technical field of aircraft design, and in particular relates to an aerodynamic topology optimization method for airfoil design. Background Art

[0002] In the development of aircraft, the improvement of aerodynamic performance requirements has continuously driven aircraft designers to make innovative designs based on design experience and understanding of flow physics. The review paper by Bravo Mosquera et al. summarizes various unconventional aerodynamic layout forms that have been proposed for civil aviation aircraft and proposes the trend of aerodynamic layout changes. So far, various researchers in academia and industry have proposed different unconventional aerodynamic layouts, each of which has its advantages and disadvantages. Bai Peng et al. systematically summarized the aerodynamic layout based on the principle of supersonic beneficial interference. The basic idea is to use the wave interference between aircraft components to obtain performance benefits such as increased lift or reduced drag. The development history of the concept of supersonic beneficial interference aerodynamic design is systematically sorted out, and typical configurations that apply the principle of supersonic beneficial interference are summarized, such as supersonic biplanes, flat-top configurations, ring wing and semi-ring wing configurations, parachute wing configurations, high-pressure capture wing configurations, etc., and the basic principles and aerodynamic characteristics of typical configurations are analyzed. The future of the concept of supersonic beneficial interference design is prospected, and related issues that need to be studied are outlined.

[0003] So far, the aerodynamic layout design of aircraft has always been a designer-driven task. However, new design methods, such as multidisciplinary design optimization (MDO), are methods for finding the best design based on first principles. There have been a lot of research on the application of MDO in the aerodynamic optimization design of aircraft; however, using this method, shape optimization can only be performed based on a given baseline shape. It is difficult to make a big breakthrough in aerodynamic performance improvement under this design framework and concept, and it is impossible to obtain a disruptive aerodynamic layout. For traditional aircraft, designers can achieve the established design goals based on existing design experience and combined with MDO. However, with the continuous development of aerospace, the aerodynamic performance requirements of aircraft are constantly increasing. The traditional design concept driven by design experience can no longer meet the growing design needs, such as wide-speed range aircraft, upper atmosphere aircraft, etc. Summary of the invention

[0004] The technical problem solved by the present invention is: to overcome the deficiencies of the prior art and provide an aerodynamic topology optimization method for airfoil design, which is no longer dependent on the initial shape, has a huge design space, and has a very high degree of design freedom, thus breaking through the existing aerodynamic layout design paradigm.

[0005] The object of the present invention is achieved through the following technical scheme: an aerodynamic topology optimization method for airfoil design, comprising: generating a plurality of normal wing airfoils, i.e., initial solutions; generating an initial subdomain with a specified interval length around each initial solution, performing Latin hypercube sampling in the initial subdomain to obtain a plurality of samples; gridding each sample obtained by sampling to obtain a grid sample, inputting a preset Mach number, a preset drag, and a preset lift into the grid sample, and outputting a penalized objective function; constructing a proxy model for the penalized objective function, performing subdomain optimization on the proxy model, and obtaining an optimized penalized objective function.

[0006] In the above-mentioned aerodynamic topology optimization method for airfoil design, generating multiple normal wing airfoils includes: generating a level set function according to preset design variables; truncating the level set function with a preset plane to obtain a truncated boundary, i.e., the wing airfoil; and removing the erroneous airfoil area in the wing airfoil to obtain a normal wing airfoil.

[0007] In the above-mentioned aerodynamic topology optimization method for airfoil design, removing erroneous airfoil areas in the wing airfoil includes: detecting two self-intersecting airfoil areas and two self-contained airfoil areas in the wing airfoil; for multiple intersection areas in the two self-intersecting airfoil areas, retaining the largest intersection area and removing the remaining intersection areas; for the two self-contained airfoil areas, removing the contained airfoil area.

[0008] In the above-mentioned aerodynamic topology optimization method for airfoil design, a graphics detection algorithm is used to detect two self-intersecting airfoil regions and two self-contained airfoil regions in the wing airfoil.

[0009] In the above-mentioned aerodynamic topology optimization method for airfoil design, subdomain optimization of the proxy model includes: after the subdomain converges, the subdomain is moved and reduced, and after the maximum number of iterations or overall convergence is reached, the optimized penalized objective function is obtained.

[0010] An aerodynamic topology optimization system for airfoil design comprises: a first module for generating a plurality of normal wing airfoils, i.e., initial solutions; a second module for generating an initial subdomain with a specified interval length around each initial solution, and performing Latin hypercube sampling in the initial subdomain to obtain a plurality of samples; a third module for performing grid division on each sample obtained by sampling to obtain a grid sample, inputting a preset Mach number, a preset drag, and a preset lift into the grid sample, and outputting a penalized objective function; and a fourth module for constructing a proxy model for the penalized objective function, performing subdomain optimization on the proxy model, and obtaining an optimized penalized objective function.

[0011] In the above-mentioned aerodynamic topology optimization system for airfoil design, generating multiple normal wing airfoils includes: generating a level set function according to preset design variables; truncating the level set function with a preset plane to obtain a truncated boundary, i.e., the wing airfoil; and removing an erroneous airfoil area in the wing airfoil to obtain a normal wing airfoil.

[0012] In the above-mentioned aerodynamic topology optimization system for airfoil design, removing erroneous airfoil areas in the wing airfoil includes: detecting two self-intersecting airfoil areas and two self-contained airfoil areas in the wing airfoil; for multiple intersection areas in the two self-intersecting airfoil areas, retaining the largest intersection area and removing the remaining intersection areas; for the two self-contained airfoil areas, removing the contained airfoil area.

[0013] In the aerodynamic topology optimization system for airfoil design, a graphics detection algorithm is used to detect two self-intersecting airfoil regions and two self-contained airfoil regions in the wing airfoil.

[0014] In the aerodynamic topology optimization system for airfoil design, subdomain optimization of the proxy model includes: after the subdomain converges, the subdomain is moved and reduced, and after the maximum number of iterations or overall convergence is reached, the optimized penalized objective function is obtained.

[0015] Compared with the prior art, the present invention has the following beneficial effects:

[0016] (1) The present invention utilizes the high degree of freedom of topological design, basically frees the artificial constraints of the design space, and no longer relies on the traditional experience of designers, so that a series of innovative aerodynamic shapes with better aerodynamic performance can be found in a larger design domain, breaking through the existing aerodynamic layout design paradigm;

[0017] (2) The present invention uses the intersection line of a fixed plane and a level set function to characterize the aerodynamic shape, realizes the dynamic evolution of the topological boundary of the aerodynamic shape, and obtains a topological parameterized modeling method that can characterize different topological structures. By introducing an adaptive large neighborhood search algorithm, the problem of huge sample size caused by topological deformation is solved, and the application of aerodynamic topology optimization design on different aerodynamic shapes is realized. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0019] Figure 1 is a flow chart of an aerodynamic topology optimization method for airfoil design provided by an embodiment of the present invention;

[0020] Figure 2 is a schematic diagram of a parameterized level set method provided by an embodiment of the present invention;

[0021] Figure 3 is a topology boundary evolution diagram of the aerodynamic topology optimization process provided by an embodiment of the present invention;

[0022] Figure 4 It is an initial sample diagram of the pneumatic topology design process provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0023] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present disclosure and to be able to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that, in the absence of conflict, the embodiments of the present invention and the features described in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0024] Aerodynamic topology optimization is a new design paradigm. Different from traditional size optimization and shape optimization, aerodynamic topology optimization does not rely on the initial shape. As long as the optimization goals and constraints are given, it is possible to design an aircraft layout that meets the performance requirements. Aerodynamic topology optimization can create topological components that the initial shape does not have. For example, from the fuselage of a single connected domain, it is possible to optimize the control surface, vertical tail and even the engine nacelle. Therefore, the design space of aerodynamic topology optimization is huge, and the design freedom is extremely high. It is expected to break through the existing aerodynamic layout design paradigm, discover new layout forms and even create new design ideas.

[0025] Figure 1 1 is a flow chart of an aerodynamic topology optimization method for airfoil design provided by an embodiment of the present invention. Figure 1 As shown, the aerodynamic topology optimization method for airfoil design includes: generating multiple normal wing airfoils, i.e., initial solutions; generating an initial subdomain with a specified interval length around each initial solution, performing Latin hypercube sampling in the initial subdomain to obtain multiple samples; meshing each sample obtained by sampling to obtain a grid sample, inputting a preset Mach number, a preset drag, and a preset lift into the grid sample, and outputting a penalized objective function; constructing a proxy model for the penalized objective function, performing subdomain optimization on the proxy model, and obtaining an optimized penalized objective function.

[0026] Generating a plurality of normal wing airfoils comprises: generating a level set function according to preset design variables (i.e., initial wing airfoil variables); truncating the level set function with a preset plane to obtain a truncated boundary, i.e., the wing airfoil; and removing an erroneous airfoil region in the wing airfoil to obtain a normal wing airfoil.

[0027] Removing erroneous airfoil areas in a wing airfoil includes: detecting two self-intersecting airfoil areas and two self-contained airfoil areas in the wing airfoil; for multiple intersection areas in the two self-intersecting airfoil areas, retaining the largest intersection area and removing the remaining intersection areas; for the two self-contained airfoil areas, removing the contained airfoil areas. Specifically, a graphics detection algorithm is used to detect the self-intersection and self-containment phenomena of the airfoil areas in the wing airfoil. First, both the self-intersection and self-containment phenomena only need to judge the intersection area between multiple wing areas. When the intersection area is not 0, the airfoil area with the largest area is saved, and the remaining airfoil areas with smaller areas are automatically deleted.

[0028] The two self-intersecting airfoil regions and the two self-contained airfoil regions in the wing airfoil are detected using a graphics detection algorithm.

[0029] Subdomain optimization of the proxy model includes: after the subdomain converges, moving and reducing the subdomain, reaching the maximum number of iterations or overall convergence, and obtaining an optimized penalized objective function.

[0030] According to the traditional point density description method, the design variables are related to the resolution of the finite element, which can easily lead to the dimensional disaster phenomenon and is difficult to solve using non-gradient methods. The level set function can overcome this defect well and extract the airfoil boundary from the level set function determined by the design variables, so that more complex structures can be represented with fewer design variables.

[0031] The present embodiment adopts design variables to generate a level set function; truncates the generated level set function to obtain a topological structure; detects the self-intersection and self-containment of the generated wing airfoil, and removes erroneous airfoil forms; selects a part of the entire design domain, and constructs a proxy model in the subdomain, and then moves the subdomain after optimizing using the proxy model. In the optimization process, a non-gradient method is used to first "coarse search" and then "fine search", and the size of the search subdomain is gradually reduced in the optimization process to reduce the adaptive neighborhood; after the subdomain is specified, Latin hypercube sampling is used in the subdomain to calculate the response and objective function value of the sample; the grid is automatically divided and the boundary conditions are set; finite element calculation is performed and the calculation results are exported; the information value in the dat file is read and converted into an unconstrained objective function form; after calculating the objective function values ​​and constraint values ​​of all sample points in the subdomain, a penalized objective function form is constructed, and a low-cost and high-fidelity model of the original objective function is constructed using the proxy model; the parameterized level set method is combined with the optimization algorithm to realize the design of a new concept airfoil.

[0032] Figure 2 is a schematic diagram of a parameterized level set method provided by an embodiment of the present invention; Figure 3 is a topology boundary evolution diagram of the aerodynamic topology optimization process provided by an embodiment of the present invention; Figure 4 It is an initial sample diagram of the pneumatic topology design process provided by an embodiment of the present invention.

[0033] The present embodiment also provides an aerodynamic topology optimization system for airfoil design, the system comprising: a first module, for generating a plurality of normal wing airfoils, i.e., initial solutions; a second module, for generating an initial subdomain with a specified interval length around each initial solution, and performing Latin hypercube sampling in the initial subdomain to obtain a plurality of samples; a third module, for gridding each sample obtained by sampling to obtain a grid sample, inputting a preset Mach number, a preset drag, and a preset lift into the grid sample, and outputting a penalized objective function; a fourth module, for constructing a proxy model for the penalized objective function, performing subdomain optimization on the proxy model, and obtaining an optimized penalized objective function.

[0034] This embodiment takes advantage of the high degree of freedom of topological design, basically frees the artificial constraints of the design space, and no longer relies on the traditional experience of designers, so that a series of innovative aerodynamic shapes with better aerodynamic performance can be found in a larger design domain, breaking through the existing aerodynamic layout design paradigm; this embodiment uses NURBS surfaces to express level set functions, and uses the intersection line of the 0 level set plane and the level set function to characterize the aerodynamic shape, so as to realize the dynamic evolution of the topological boundary of the aerodynamic shape, and obtain a topological parameterized modeling method that can characterize different topological structures. By introducing an adaptive large neighborhood search algorithm, the problem of huge sample size caused by topological deformation is solved, and the application of aerodynamic topology optimization design to different aerodynamic shapes is realized.

[0035] Although the present invention has been disclosed as above in the form of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art may make possible changes and modifications to the technical solution of the present invention by using the methods and technical contents disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall fall within the protection scope of the technical solution of the present invention.

Claims

1. An aerodynamic topology optimization method for airfoil design, characterized in that include: Generate multiple normal wing airfoils, i.e., initial solutions; Generate an initial subdomain with a specified interval length around each initial solution, and perform Latin hypercube sampling in the initial subdomain to obtain multiple samples; Gridding each sample obtained by sampling to obtain a grid sample, inputting a preset Mach number, a preset drag and a preset lift into the grid sample, and outputting a penalized objective function; A proxy model is constructed for the penalized objective function, and subdomain optimization is performed on the proxy model to obtain an optimized penalized objective function.

2. The aerodynamic topology optimization method for airfoil design according to claim 1, characterized in that: Generate multiple normal wing airfoils including: Generate a level set function based on the preset design variables; The level set function is truncated with a preset plane to obtain the truncated boundary, i.e., the wing airfoil; Remove the wrong airfoil area in the wing airfoil to obtain a normal wing airfoil.

3. The aerodynamic topology optimization method for airfoil design according to claim 2, characterized in that: The areas of the wing airfoil that need to be removed include: Two self-intersecting airfoil regions and two self-contained airfoil regions in the wing airfoil are detected; For multiple intersection areas in two self-intersecting airfoil regions, the largest intersection area is retained and the remaining intersection areas are removed; For two self-contained airfoil regions, the contained airfoil region is removed.

4. The aerodynamic topology optimization method for airfoil design according to claim 3, characterized in that: The two self-intersecting airfoil regions and the two self-contained airfoil regions in the wing airfoil are detected using a graphics detection algorithm.

5. The aerodynamic topology optimization method for airfoil design according to claim 1, characterized in that: Subdomain optimization of the proxy model includes: after the subdomain converges, moving and reducing the subdomain, reaching the maximum number of iterations or overall convergence, and obtaining an optimized penalized objective function.

6. An aerodynamic topology optimization system for airfoil design, characterized in that include: The first module is used to generate a plurality of normal wing airfoils, i.e., initial solutions; The second module is used to generate an initial subdomain with a specified interval length around each initial solution, and perform Latin hypercube sampling in the initial subdomain to obtain multiple samples; The third module is used to grid each sample obtained by sampling to obtain a grid sample, input a preset Mach number, a preset drag and a preset lift into the grid sample, and output a penalized objective function; The fourth module is used to construct a proxy model for the penalized objective function, perform subdomain optimization on the proxy model, and obtain an optimized penalized objective function.

7. The aerodynamic topology optimization system for airfoil design according to claim 6, characterized in that: Generate multiple normal wing airfoils including: Generate a level set function based on the preset design variables; The level set function is truncated with a preset plane to obtain the truncated boundary, i.e., the wing airfoil; Remove the wrong airfoil area in the wing airfoil to obtain a normal wing airfoil.

8. The aerodynamic topology optimization system for airfoil design according to claim 7, characterized in that: The areas of the wing airfoil that need to be removed include: Two self-intersecting airfoil regions and two self-contained airfoil regions in the wing airfoil are detected; For multiple intersection areas in two self-intersecting airfoil regions, the largest intersection area is retained and the remaining intersection areas are removed; For two self-contained airfoil regions, the contained airfoil region is removed.

9. The aerodynamic topology optimization system for airfoil design according to claim 8, characterized in that: The two self-intersecting airfoil regions and the two self-contained airfoil regions in the wing airfoil are detected using a graphics detection algorithm.

10. The aerodynamic topology optimization system for airfoil design according to claim 6, characterized in that: Subdomain optimization of the proxy model includes: after the subdomain converges, moving and reducing the subdomain, reaching the maximum number of iterations or overall convergence, and obtaining an optimized penalized objective function.

Images