CN-121982155-A - Method for generating microscopic mechanical property distribution diagram of solid waste composite cement-based cementing material

CN121982155ACN 121982155 ACN121982155 ACN 121982155ACN-121982155-A

Abstract

The invention discloses a method for generating a microscopic mechanical property distribution diagram of a solid waste composite cement-based cementing material, which belongs to the technical field of solid waste composite cement-based cementing material property prediction and comprises the steps of obtaining BSE, EDS and nano indentation data of a sample; identifying the indentation areas in the nano indentation mark graph, generating corresponding masks of the areas, counting the masks of the areas to obtain element content data of the indentation areas, establishing a prediction model based on the element content and the elasticity data of the areas by adopting machine learning, performing superpixel segmentation on the BSE image to be matched with the scale of the indentation areas, calculating element characteristics of the superpixel areas, predicting the elasticity modulus of each superpixel area, and mapping the predicted elasticity modulus to the corresponding areas to generate an elasticity modulus distribution map. The method can realize the prediction of the two-dimensional spatial distribution of the internal elastic modulus of the solid waste composite cement-based cementing material, and provides data and visual support for the research of material heterogeneity and multi-scale modeling.

Inventors

YAN CHANGAN
WU HAIXIA
XU RUIHAN
YANG ZIYUE
LI LIHUI
WANG XIANGQUAN
MEI XIANGYANG
YANG YONGCHEN
GUAN HUAI
Duan Dianrong
CHEN DONGNI

Assignees

云南省生态环境科学研究院

Dates

Publication Date: 20260505
Application Date: 20260407

Claims (9)

1. The method for generating the microscopic mechanical property distribution map of the solid waste composite cement-based cementing material is characterized by comprising the following steps of: S1, acquiring a back scattering electron image, an energy spectrum analysis element surface spectrum image and nano indentation test data of a corresponding area of a solid waste composite cement-based cementing material sample; S2, based on the nano indentation mark image, identifying each nano indentation area by utilizing image processing software, and generating an area mask image corresponding to each nano indentation point, so that the discrete nano indentation test points are converted into standardized space analysis areas; S3, based on the regional mask image in the step S2, carrying out regional statistics on the spectral images of the energy spectrum analysis elements, and calculating the average gray value of each element in each region to obtain quantitative data of the element content of each region; S4, taking the element content of the region as an input characteristic, taking the experimental measured elastic modulus of the corresponding region as an output target, establishing a mapping model between the element content of the region and the elastic modulus by adopting a machine learning method, and storing a trained elastic modulus prediction model; S5, performing super-pixel segmentation on a microstructure of the material based on the back scattering electron image, and dividing the whole microscopic image into a plurality of small regions with continuous space, so that the size of the small regions is matched with the scale of the nano indentation region; S6, calculating corresponding element characteristics of each super-pixel region, inputting the element characteristics into an elastic modulus prediction model trained in the step S4, and predicting the predicted elastic modulus of each super-pixel region; And S7, mapping the predicted elastic modulus of each super pixel region to a corresponding pixel region, and enabling pixels in the same super pixel region to be endowed with the same elastic modulus value, so that an elastic modulus distribution diagram under the microscopic scale of the material is generated.
2. The method for generating the microscopic mechanical property distribution map of the solid waste composite cement-based cementing material according to claim 1, wherein the S1 step is characterized in that the back scattering electron image acquisition condition is that the accelerating voltage is 15kV, the working distance is 11mm, the energy spectrum analysis element surface spectrum image acquisition condition is that the processing time is set to be 4, the dead time is not more than 20%, the scanning parameter per frame is 256 mu S/pixel, and the scanning frame number is 4 frames.
3. The method for generating the microscopic mechanical property distribution map of the solid waste composite cement-based cementing material according to claim 1, wherein in the step S2, an edge recognition tool of image processing software is used for adjusting tolerance parameters to accurately recognize the reaction edge of the target cementing particles.
4. The method for generating a microscopic mechanical property distribution map of a solid waste composite cement-based cementing material according to claim 1, wherein in the step S3, the element content is calculated in each strip mask area by aiming at each element surface map, so as to finally obtain a quantitative curve of element content changing along with strips, and the quantitative curve is stored for CSV.
5. The method for generating the microscopic mechanical property distribution map of the solid waste composite cement-based cementing material according to claim 1, wherein in the step S4, the machine learning method is to build a nonlinear mapping relation between the regional element content and the elastic modulus by adopting a regression model, and the regression model is an integrated learning model based on a decision tree.
6. The method for generating a microscopic mechanical property distribution map of a solid waste composite cement-based cementing material according to claim 1, wherein in the step S5, the super-pixel segmentation is performed based on a principle of pixel gray similarity and spatial continuity, and the average pixel number of the single super-pixel region after segmentation is made to be similar to the average pixel number of the nano-indentation region by adjusting the segmentation number parameter.
7. The method for generating a microscopic mechanical property distribution map of a solid waste composite cement-based cementing material according to claim 1, wherein in step S6, element characteristics are calculated for each super-pixel region in a manner consistent with the nanoindentation region.
8. The method for generating a micromechanics property distribution map of a solid waste composite cement-based binder according to claim 1, wherein in the step S7, in the predicted elastic modulus map, the predicted elastic modulus value of each super-pixel region is directly assigned to all pixels in the region, and the elastic modulus between adjacent regions is not interpolated or smoothed, so that region boundary information is retained.
9. The method for generating the microscopic mechanical property distribution map of the solid waste composite cement-based cementing material according to claim 1, wherein the solid waste composite cement-based cementing material is pure silicate cement, a fly ash composite silicate cementing system, a slag composite silicate cementing system or a metakaolin composite silicate cementing system.

Description

Method for generating microscopic mechanical property distribution diagram of solid waste composite cement-based cementing material Technical Field The invention belongs to the technical field of solid waste composite cement-based cementing material performance prediction, and particularly relates to a method for generating a microscopic mechanical property distribution map of a solid waste composite cement-based cementing material. Background The macroscopic mechanical properties and long-term durability of cement-based cements are fundamentally dependent on the spatial distribution of the phases and pore structures on their microscopic scale, and thus on the microscopic mechanical properties. Therefore, the realization of high-resolution characterization of the spatial distribution of the micromechanics (such as elastic modulus) is a key link for establishing a microstructure-macroscopic performance relationship chain. The nanoindentation technology is widely applied in the micromechanics characterization field as a technical means capable of directly measuring the mechanical properties of the microcells, but is essentially point-by-point measurement, so that the problems of low data acquisition efficiency and insufficient space coverage rate exist, and the structure around the microbending is possibly adversely affected in the test process, so that the large-area mechanical property spatial distribution map is difficult to efficiently and quickly acquire. The above problems are more pronounced with regard to solid waste composite cement-based cementitious materials. Because of the complex chemical nature and mineral composition of the solid waste admixture (fly ash, slag, metakaolin and the like), the reaction products are more various, and the spatial heterogeneity of the microstructure is more obvious. Therefore, the spatial distribution of the micromechanics of the solid waste composite cement-based cementing material is accurately represented and predicted, and the method has important theoretical and engineering values for regulating and controlling the reaction process and scientific design of the proportion. On the other hand, the analysis technology combining the back scattering electron image and the energy dispersion X-ray spectrum (energy spectrum analysis) surface scanning can efficiently acquire the atomic number degree and element distribution information of the material, and an effective analysis means is provided for phase identification and quantification. Studies have shown that there is A clear correlation between the elastic modulus of phases in cement-based materials, such as C-S-H gels, C-A-S-H gels, unhydrated cement clinker, fly ash, etc., and their chemical composition (e.g. calcium to silicon ratio, aluminum to silicon ratio) and density. This shows that the mapping relationship exists between the characteristics of the backscattering electron, the phase category obtained by energy spectrum analysis, the element content and the like and the micromechanics property. However, the existing characterization method lacks effective fusion that the back scattering electrons and the energy spectrum analysis can rapidly provide large-area component and phase distribution information, but mechanical properties cannot be directly obtained, and the nano indentation can accurately measure the mechanical properties of local micro-areas, but large-area coverage is difficult to realize. The Chinese patent publication No. CN105241904A discloses a fly ash phase analysis method based on energy dispersion X-ray spectrum, which carries out superposition treatment on a plurality of element energy spectrum surface distribution images of the same test area to judge the phase types of each pixel in the test area, however, the accuracy of the method depends on the effectiveness of an image processing algorithm, and the quantitative analysis capability is limited, especially for complex phases, the practical application is limited. Therefore, how to effectively integrate the back scattering electrons, the energy spectrum analysis and the nanoindentation technology, and to take limited but accurate nanoindentation points as connection coordinates, establish a machine learning prediction model between micromechanics and chemical components and phase contents thereof, and further extrapolate and predict micromechanics feature distribution of the whole imaging area is a feasible thinking for solving the current characterization problem. Aiming at a solid waste composite system with more complex components and structures, the method for fusing multi-modal data and predicting machine learning is developed, and is particularly important for efficiently and accurately evaluating the spatial distribution of micromechanics performance. Disclosure of Invention Aiming at the problems that in the traditional micromechanics research, the nanoindentation test is mainly based on discrete measuring points, the space cover