CN-122021202-A - Powder feeding type additive manufacturing temperature field high-flux prediction method based on matrix operator

CN122021202ACN 122021202 ACN122021202 ACN 122021202ACN-122021202-A

Abstract

The invention discloses a powder feeding type additive manufacturing temperature field high-flux prediction method based on a matrix operator, which relates to the technical field of high-flux model calculation and comprises the steps of performing space-time dispersion on a unit volume net heat input expression by a finite difference method, constructing a three-dimensional space dispersion format and an equivalent time step length, and forming a finite difference temperature prediction model; based on the technological characteristics of additive manufacturing layer-by-layer forming, grid nodes of a current manufacturing layer and formed layers in a finite difference temperature prediction model are set as an activation calculation domain, grid nodes of other layers are subjected to shielding treatment, a matrix operator is constructed based on heat conduction rules in the finite difference temperature prediction model, a matrixing calculation flow is formed, and temperature field distribution is updated. The calculation method of the matrix operator is constructed, so that the calculation efficiency of the prediction model is greatly improved, and meanwhile, the super-parameter weighing method of the prediction model is provided for the constructed model, so that the weighing between the calculation efficiency and the prediction precision of the prediction model is realized.

Inventors

ZHANG ZHIHUI
HUANG JIANGENG
LIU RUIJIA
KANG ZHONGXIONG
TENG JINZE
WANG HAIPENG
REN LUQUAN

Assignees

吉林大学

Dates

Publication Date: 20260512
Application Date: 20260414

Claims (10)

1. The powder feeding type additive manufacturing temperature field high-throughput prediction method based on the matrix operator is characterized by comprising the following steps of, Constructing a powder feeding type metal additive manufacturing thermal field prediction theoretical model, and setting a laser Gaussian heat source item, a convection heat exchange item and a radiation heat dissipation item according to a heat conduction control equation to obtain a net heat input expression of unit volume; performing space-time dispersion on the net heat input expression in unit volume by a finite difference method, constructing a three-dimensional space dispersion format and an equivalent time step, and forming a finite difference temperature prediction model; Based on the process characteristics of additive manufacturing layer-by-layer forming, grid nodes of a current manufacturing layer and a formed layer in the finite difference temperature prediction model are set as an activation calculation domain, and grid nodes of other layers are subjected to shielding treatment; Constructing a matrix operator based on a heat conduction rule in the finite difference temperature prediction model, forming a matrixing calculation flow, and updating the temperature field distribution; And adjusting the spatial resolution of the finite difference temperature prediction model and the number of network node activation layers, establishing an accuracy-efficiency balance objective function, and minimizing through a parameter optimizing algorithm to obtain the optimal balance configuration.
2. The method for predicting the high flux of the powder feeding type additive manufacturing temperature field based on the matrix operator according to claim 1, wherein the powder feeding type metal additive manufacturing thermal field prediction theoretical model comprises temperature conduction inside a component, a laser Gaussian heat source item, thermal convection between the component and the environment and thermal radiation of the component.
3. The method for predicting the high flux of the powder feeding type additive manufacturing temperature field based on the matrix operator according to claim 1, wherein the unit volume net heat input expression is composed of a laser Gaussian heat source item, a convection heat exchange item between a component and the environment and a component heat radiation item, wherein the laser heat source item is a positive item, and the convection heat exchange item and the radiation heat exchange item are negative items.
4. The method for predicting the high flux of the powder feeding type additive manufacturing temperature field based on the matrix operator is characterized in that the method is used for constructing a three-dimensional space discrete format and an equivalent time step length to form a finite difference temperature prediction model, wherein the method is used for dividing components in space through gridding nodes to construct the three-dimensional space discrete format, and the method is used for constructing the equivalent time step length through equivalent heating times of unit volume in time to form the finite difference temperature prediction model.
5. The method for predicting the high flux of the powder feeding type additive manufacturing temperature field based on the matrix operator according to claim 1, wherein the grid nodes of the current manufacturing layer and the formed layer in the finite difference temperature prediction model are set as an activation calculation domain, the grid nodes of the other layers are subjected to shielding treatment, and the method comprises the following specific steps of, Dividing the finite difference temperature prediction model into a manufacturing layer, an intermediate layer and a basal layer, activating nodes of a forming area of the manufacturing layer, shielding nodes of an unshaped area, and gradually activating corresponding nodes of the unshaped area along with the movement of a molten pool; all nodes in the middle layer are set to be in an activated state, and all nodes in the basal layer are set to be in a shielding state.
6. The method for predicting high flux of powder feeding type additive manufacturing temperature field based on matrix operators as set forth in claim 1, wherein the matrixing calculation process comprises matrixing operator construction and matrix process operation, and matrixing operator construction comprises calculating temperature difference operators of interlayer and intra-layer temperature conduction and calculating boundary condition operators of two boundary conditions of heat convection and heat radiation.
7. The powder feeding type additive manufacturing temperature field high-throughput prediction method based on matrix operators is characterized in that the intra-layer temperature conduction temperature difference operator divides network nodes in the same manufacturing layer into conduction tail direction nodes, conduction middle section nodes and conduction head direction nodes according to the conduction direction, and performs matrixing difference operation on the temperature matrix based on a filling transformation matrix, a copying transformation matrix and a reduction matrix, and calculates temperature difference between adjacent network nodes in the same layer; The temperature difference operator of interlayer temperature conduction carries out temperature transfer between different layers by carrying out matrix-form difference operation on the temperature matrixes of adjacent manufacturing layers.
8. The method for predicting the high flux of the powder feeding type additive manufacturing temperature field based on the matrix operator, as set forth in claim 6, wherein the boundary condition operator screens network nodes positioned on the outer surface of the component by constructing a screening matrix, performs thermal convection calculation and thermal radiation calculation on the network nodes based on the screening matrix, and participates the thermal convection item and the thermal radiation item in the updating operation of the temperature field in a matrix form.
9. The method for predicting the high flux of the powder feeding type additive manufacturing temperature field based on the matrix operator, which is characterized by comprising the steps of adjusting the spatial resolution of a finite difference temperature prediction model and the number of network node activation layers, setting the resolution of the finite difference temperature prediction model and the number of the network node activation layers as adjustable parameters, obtaining a model prediction precision evaluation index based on a comparison result of the prediction temperature field and a real manufacturing temperature field, and simultaneously, reasoning the time required by one complete manufacturing process reasoning by a powder feeding type metal additive manufacturing thermal field prediction theory model.
10. The method for high-throughput prediction of powder feeding type additive manufacturing temperature field based on matrix operator as set forth in claim 1, wherein the establishing precision-efficiency balance objective function and minimizing by parameter optimizing algorithm to obtain optimal balance configuration is to establish the precision-efficiency balance objective function composed of prediction error term and calculation overhead term together based on model prediction precision evaluation index and powder feeding type metal additive manufacturing thermal field prediction theory model, and minimize by parameter optimizing algorithm to obtain optimal balance configuration.

Description

Powder feeding type additive manufacturing temperature field high-flux prediction method based on matrix operator Technical Field The invention relates to the technical field of high-flux model calculation, in particular to a powder feeding type additive manufacturing temperature field high-flux prediction method based on a matrix operator. Background The powder feeding type metal additive manufacturing is widely used for defect repair, surface coating strengthening and near-net forming, has remarkable advantages in the high value-added fields of aerospace, energy equipment, molds, medical implantation and the like, can realize the material feeding according to the requirements of complex curved surfaces and support the manufacturing of complex configuration components such as internal runners, lightweight topological structures, porous structures and the like, and because the high-energy laser heat source continuously melts metal powder, huge thermal gradients and huge temperature accumulation cannot be avoided in the manufacturing process, the problems of residual stress and deformation, pore and defect formation, tissue and performance anisotropy and the like caused by the fact that the problems restrict the powder feeding type metal additive manufacturing, and has remarkable significance in improving the performance and the forming quality of the powder feeding type metal additive component for reasoning and analysis of the thermal history in the manufacturing process. The current powder-feeding type metal additive manufacturing thermal process is mainly evolved into two methods through long-term research, the thermal field evolution is analyzed through a theoretical formula based on a numerical method, such as commercial simulation software of Ansys, COMSOL and the like, the thermal field evolution is predicted through data driving based on machine learning or deep learning, the thermal field distribution of a component in the manufacturing process is obtained mainly by means of an energy conservation framework of a thermal conduction equation and phase change heat, combining with moving heat source, convection and radiation boundary conditions and layer-by-layer and channel deposition geometric updating, the calculation amount is huge, engineering applications such as online simulation and offline optimization are difficult to realize, the thermal field fast prediction and online analysis are realized by taking data driving as a core through a mapping relation between learning process parameters, geometric paths and the thermal field evolution, and the thermal history prediction method based on data driving relies on a large amount of data to train a machine learning prediction model, and when the data amount is insufficient, the thermal history prediction model based on data driving faces a cold starting problem. Disclosure of Invention The present invention has been made in view of the above-described problems occurring in the prior art. Therefore, the invention provides a powder feeding type additive manufacturing temperature field high-throughput prediction method based on a matrix operator, which solves the problem that the traditional finite difference model is insufficient in calculation efficiency in node-by-node index and difference update. In order to solve the technical problems, the invention provides the following technical scheme: The invention provides a powder feeding type additive manufacturing temperature field high-throughput prediction method based on a matrix operator, which comprises the steps of constructing a powder feeding type metal additive manufacturing thermal field prediction theoretical model, setting a laser Gaussian heat source item, a convection heat source item and a radiation heat dissipation item according to a heat conduction control equation to obtain a unit volume net heat input expression, performing space-time dispersion on the unit volume net heat input expression through a finite difference method to construct a three-dimensional space discrete format and an equivalent time step to form a finite difference temperature prediction model, setting grid nodes of a current manufacturing layer and a formed layer in the finite difference temperature prediction model as an activation calculation domain based on process characteristics of additive manufacturing layer-by-layer, performing shielding treatment on grid nodes of other layers, constructing a matrix operator based on a heat conduction rule in the finite difference temperature prediction model, forming a matrixing calculation flow, updating temperature field distribution, adjusting spatial resolution and network node activation layers of the finite difference temperature prediction model, establishing a precision-efficiency objective function, and performing minimization through a parameter optimizing algorithm to obtain optimal balance configuration. The matrix operator-based powder feeding type additive man