| CPC G06F 30/20 (2020.01) [G06F 2111/08 (2020.01); G06F 2111/10 (2020.01)] | 3 Claims |
|
1. A method for numerical reconstruction and heat transfer characteristics evaluation of a microstructure of thermal barrier coatings (TBCs) containing microcracks before application of TBCs on surface of an object, comprising:
S1: determining a simulation area and size settings, comprising settings of a physical model and a simulation calculation grid; and generating random microcracks with different morphological characteristics based on a Monte Carlo simulation method and a quartet structure generation set (QSGS) method;
S2: changing inclination directions of different microcracks by a coordinate transformation method according to a certain statistical law and placing them in the simulation area, and determining whether a space occupied by the microcracks reaches a porosity ratio of preset microcracks;
S3: building a general pore model of thermal barrier coatings (TBCs) based on the QSGS method, wherein a solid skeleton in the TBCs is taken as a growth phase, pores are non-growth phases, settings of the porosity controls a pore volume fraction of the coating, and settings of a generation probability of a nucleation center of the solid skeleton controls a number of the pores; traversing the initially randomly distributed growth nucleus nodes, making each growth nucleus node randomly grow in a three-dimensional space, controlling the generation of the TBCs with different structures according to the growth probabilities of growth nucleus in 26 directions under three-dimensional coordinates, wherein the growth probability of the growth nucleus of a lamellar structure coating in a horizontal direction is far greater than that of the columnar structure coating, and the columnar structure coating is opposite;
S4: superposing a general pore structure into the simulation area containing microcracks to generate a coating pore model coupled with microcrack defects, reconstructing true mesoscopic morphologies of the TBCs, and determining whether the preset volume fraction has been reached;
S5: building a heat transfer analysis model based on a thermal Lattice Boltzmann method to calculate heat insulation performance parameters such as temperature distribution, and thermal conductivity, and using the Monte Carlo simulation method in the step S1 to generate a microcrack morphology under a certain statistical distribution law:
S1.1 generating random numbers uniformly distributed over an interval of [0,1]:
an iterative formula of a linear congruential method for generating the random numbers is:
xn=(axn-1+c)(mod M) (1)
in the formula, xn is a current pseudo random number, xn-1 is a previous pseudo random number, a is a non-negative multiplier; c is a non-negative increment; M is a modulus; and (mod M) represents a remainder after being divided by m;
random numbers R uniformly distributed over the interval of [0,1] are thus generated:
 S1.2 using uniform random number calculation to generate random numbers with other specified statistical distribution law:
the uniform random numbers over the interval of [0,1] generated in the (1) above are used to generate random numbers that obey other distribution types;
the probability density function of a uniform distribution is:
 wherein a and b represent end points of an interval,
a method for calculating random variables is as follows:
xf=(b−a)R+a
in the formula, xf is random numbers required; and R represents random numbers uniformly distributed over the interval [0,1];
S1.3 carrying out translation and rotational transformation of coordinates to realize changes in the orientation of microcracks:
there are a base coordinate system 0 and a moving coordinate system I, their postures are the same, and their origins are not coincident; for any particle P in a space that is fixedly connected to the coordinate system I, its coordinates in the coordinate system 0 are expressed as P0=[p0x p0y p0z]T, and in the coordinate system I are expressed as P1=[p1x p1y p1z]T, and coordinates of the origin O1 of the coordinate system I are D01=[d0x d0y d0z]T; there exists P0=P1+D01 according to the rules for vector operations; when an attitude change matrix of the coordinate system I relative to the coordinate system 0 is taken into account, relative relationship between them is considered as a combination of translational motion between the origins and pure rotation around the origins, there exists P0=Rot01P1+D01; similarly P1=Rot10P0+D10; wherein Rot represents basic rotation, which is written in a matrix in three-dimensional coordinates in the form of:
 the above formulae identify rotation by angles α, ϕ and θ around x, y and z axes, respectively; and
S1.4 generating microcracks, and using the QSGS method to control the growth probability ratio in between the length and width direction to 100 to 1 to ensure that the microcracks are wedge-shaped; controlling the thickness direction of a single microcrack to be two layers of grids, so as to ensure that the microcrack is approximately a lamellar structure; and using the Monte Carlo simulation method to generate such morphology parameters as nucleation center point (x0, y0, z0), inclination angle φ, inclination angle θ, length, and the number according to a certain statistical distribution law, wherein the nucleation center point (x0, y0, z0) serves as a growth nucleation center of microcracks under QSGS method, the length of the microcracks controls the growth probabilities of the microcracks in different directions, and the inclination angle φ and the inclination angle θ serve as bases for rotational transformation of coordinates, such that microcracks with different shapes, sizes, directions, and the like, distributed according to a certain statistical law are generated, and the algorithm in the step S3 is a circular traversal of the initial growth nucleus nodes and the growth nucleus nodes after growth; further, the growth phase avoids the growth of microcracks in the simulation area and does not overlap the microcracks, and the sum of the porosities of the microcracks and the general pores is equal to the preset total porosity; the three-dimensional coordinates have 26 directions in total, comprising 6 main directions, 12 face opposite angle directions and 8 body opposite angle directions, the growth nucleus node is taken as the growth origin, and the skeleton regenerates random numbers to non-porous adjacent points in the 6 main directions; when the generated random number is smaller than the growth probability of growth nucleus, the point grows to be a growth phase node; and repeating this step until the growth phase meets the preset volume fraction;
S1.5 preparing TBCs based on the results of the numerical reconstruction and heat transfer characteristics evaluation.
|