| Summary: | 碩士 === 國立成功大學 === 土木工程學系碩博士班 === 94 === Segregation refers to the areas of non-uniform distribution of coarse and fine aggregate particles in hot-mix asphalt pavements. In general, segregation is caused by mix design, improper handling of asphalt mix and paving operation. Furthermore, occurrence of segregation reduces pavement service life. Segregation could be confirmed by performing gradation tests on aggregate extracted from cores. However, the extensive time required for lab testing would prohibit immediate adjustment of the paving operation. The identification of segregation has been typically based on visual observations in Taiwan. Hence, one of the objectives of this study is to develop a quantitative model to detect presence or severity of segregation.
Field and laboratory tests were conducted. At each test section, segregation patterns were identified and field density tests were performed using non-nuclear density gauge. After field tests, cores were taken and sent to laboratory to determine texture depth, indirect tensile strength, percent air voids, asphalt content and gradation curve.
It is found that “continuous” and “end of the load” are the most frequently observed segregation types in Taiwan. The non-nuclear density gauge, PaveTracker, is non-destructive and can be used to identify segregation in medium and heavy segregated areas. The density difference between segregated area and control area increases with increasing severity of segregation. The correlation coefficient R is 0.74 based on the result of linear regression analysis between the calibrated PaveTracker density and core density.
The results of laboratory tests indicated that the texture depth increases when the segregation level increases. The texture ratio of heavy segregated area and control area is 2.9. Indirect tensile strength in medium and heavy segregated area is 8-9% lower than that in the control area. The air voids in medium or heavy segregated area increase by 20% compared to the air voids in the control area. Results of statistical comparisons between segregated and control areas showed that if a statistical difference in PaveTracker density exists, the chance of aggregate gradation difference in percent passing No.8 sieve for medium and heavy segregation is 83% and 77%, respectively.
A multiple linear regression model was developed to quantify segregation. The correlation coefficient of the statistical model is 0.88. The dependent variable is calibrated PaveTracker density value. The independent variables are texture depth, air voids, the difference in percent passing No.4 sieve, asphalt content, and core density.
|
|---|
