Essential difference and design application of boundary effect model and size effect model

Junfeng Guan, Xiaozhi Hu, Qingbin Li, Zhimin Wu

    Research output: Contribution to journalArticlepeer-review

    13 Citations (Scopus)

    Abstract

    A comparative study of the essential difference, design, and application of boundary effect model (BEM) and size effect model (SEM) was conducted. SEM applies the four different multi-parameter empirical equations to the different ratio of initial crack length and specimen depth α. SEM is limited to a data-fitting function. BEM model involves the use of a unique analytical solution. On one hand, BEM can determine the material parameters-fracture toughness KIC and tensile strength ft through the fitting of experimental data. On the other hand, it can establish a complete designed curve that describes the overall fracture of structure based on the identified KIC and ft. The effect on three fracture controlled model (ft-controlled, KIC-controlled and quasi-brittle fracture) were systematically calculated and analyzed from specimen to structure (size W=25-25000 mm). The results show that linear elastic fracture mechanics (LEFM) applies for very long crack and very large structures, and there is no further need for size effect study. The size effect model is only relevant when the crack tip is close to either the front or back boundaries. The so-called "size effect" is only a special case of boundary effect. Based on the results of fracture tests (the specimen size W is 40, 93, 215 and 500 mm, and the ratio of initial crack length and specimen depth α is 0, 0.02, 0.075, 0.15 and 0.3), the different combination conditions of tested data that BEM systematically studied are adopted to determine the influence law of KIC and ft. The results are shown that, values are basically consistent to KIC and ft, which have been determined by BEM, when the test data reaches a certain amount and one or two groups of data are removed from the overall data, either the specimens have the different W with same α=a0/W, or the same W but different α=a0/W. The studied results of this paper gave a new ideal to solve the problem that the material parameters cannot be accurately obtained using small size specimen, as well as the problem on predicting the actual structure fracture by using the material parameters.

    Original languageEnglish
    Pages (from-to)955-967
    Number of pages13
    JournalShuili Xuebao/Journal of Hydraulic Engineering
    Volume48
    Issue number8
    DOIs
    Publication statusPublished - 1 Aug 2017

    Fingerprint

    Dive into the research topics of 'Essential difference and design application of boundary effect model and size effect model'. Together they form a unique fingerprint.

    Cite this