夯實填方材料不排水反覆受剪力學模式

Abstract

夯實性填方材料由於相對密度高, 緊密性佳。受不排水反覆載重時, 其行為表現較接近緊砂或過壓密黏土受不排水反覆載重時的反覆流動性行為。本研究提出一個基於夯實性土壤的行為特徵, 可以合理描述夯實性土壤受反覆荷重的行為反應的力學模式, 試圖合理的描述夯實填方材料受不排水載重時的有效應力之力學模式。模式之原理在於令夯實性填方材料的應力應變關係受到包括應力狀態、有效應力以及應力歷史等三因素所控制, 來推導夯實性填方材料受不排水反覆載重時的應力應變行為。而夯實性填方材料受不排水反覆載重時所引發的超額孔隙水壓增量, 則可藉由應力膨脹關係中的體積增量推得到。此一模式建構於勁度控制因素與應力膨脹關係的組合, 其物理意義簡單明瞭。 本研究除推導有效應力之力學模式，並對模式中所有參數進行參數之影響探討，以釐清個別參數之角色與影響趨勢。為便於標定參數，本研究並建立了一套以基因演算法作參數最佳化標定, 省卻人為標定的時間與人為影響因素。 為闡釋本研究之力學模式的應用，本研究以新竹寶山第二水庫的築壩材料作為實際運用模擬的對象，藉由基因演算法程式自動化的幫助, 標定一組參數使本模式可以對於目標夯實填方材料在預估工作應力的範圍之下合理的描述其力學行為的參數。模擬結果, 在應變方面十分接近, 在孔隙水壓的估算雖稍有差異, 但仍在可容許範圍之內。

Compacted fills usually have low void ratio and high relative density. Their behavior under under undrained cyclic loading is close to the dynamic behavior of dense sand or overconsoildation clay. This thesis presents an effective-stress based constitutive model that reasonably represent the dynamic response of compacted fill subjected to undrained cyclic loading. Relevant factors affecting the cyclic stress-strain relation of compacted fill were taken into account in the presented model. In this model, the material’s stiffness depends on its effective stress state and stress history. The build-up of excess pore water pressure during cyclic loading was evaluated from the change in the volume strain on the basis of a stress-dilatancy relation. Parametric study was carried out to demonstrate the role of each model parameter. To improve the effectiveness of the calibration for model parameters, this study also developed an optimization scheme for parameter calibration based on genetic algorithms. Calibration by genetric algorithm appears time saving and objective in determining the required model parameters. To demonstrate the applicability of this model, the model was employed to simulate the response of the fill material of an earth dam under undrained stress controlled cyclic loading. The model parameters were calibrated from experimental data. The simulated results, in general, agree reasonably with the experimental results. The simulated strains during the cyclic loading agreed well with the experimental results. The discrepancy between the simulated and the experimental pore water was in general acceptable, with a little discrepancy.