本研究以連心維(2019)設計WERAFT之筏基變形差分程式為基礎,代入雙曲線模式分析力與位移之關係在筏基板上掌握所有節點之土壤彈簧係數且隨著力量而改變呈現非線性,並將主要影響雙曲線模式之兩個參數K與Qult進行一維正規化整理,開發WERAFT非線性分析功能,稱為WERAFT_S3,並與三維有限元素分析軟體Midas/GTS NX 分析結果驗證,並探討外力、剪力波速、筏基尺寸、筏基厚度等設計參數對筏式基礎變形之影響。本研究統整黏土地盤尺寸16~34 (m)、剪力波速120~180 (s/m)、厚度1~2 (m)之筏基直到破壞為止版上每個節點的力與位移之二次函數,供後人參考借鑒。 研究成果顯示:1. 發現雙曲線模式中之參數初始切線斜率K範圍在1.2-17.6MN/m^3、最大壓應力Qmax之範圍在203-953 kN/m^2,兩參數影響取決於土壤勁度與筏基尺寸。 2. WERAFT_S3分析結果在筏基尺寸16m×16m與26m×26m時與三維有限元素分析結果相互驗證,其結果相當適用,且當筏基尺寸變小時,發現其結果相比較為保守。3. 結果比對發現,對於尺寸較大時的筏基,WERAFT_S3預測結果與Midas/GTS NX驗證在力與位移圖上隨著力量提升誤差會逐漸擴大。4. 本研究所開發之非線性分析結果顯示是非常可以適用且提供極大之效率性,供工程師用於初步預測。
This study was made to model the nonlinear load-displacement relationships for raft foundation subjected to vertical structure load. The hyperbolic model was used to model the soils underneath the raft. Model parameters were calibrated using the ones appeared for the soils at the center of foundation with the normalized function expanded along the cross section of the foundation, The two dimensional expansion of the soil springs underneath the foundation are thus obtained. The nonlinear soil model was implemented into the WERAFT FD analysis and the corresponding results were found compatible to the 3D FE analysis. The soil spring model can be used for the variations of soil stiffness (Vs=120~180m/s) , the foundation width (16~34m), and the thickness of the raft slab (1-2m). The results can be found that: 1. The hyperbolic model parameter for initial spring constant k was found in the range of 1.2-17.6 MN/m3, whereas the maximum load pressure Qmax was found in the range of 203-953 kN/m2. They are dependents of the soil stiffness and the foundation size. 2. The proposed model can provide rational solutions in comparison with the FE analysis in the foundations with raft width at 16m and 26m. More conservation results were found as the foundation size became smaller. 3. The results showed that for foundation with larger dimensions, the load-displacement curves from the FD predictions will enlarge its deviations from the results from FE analysis. 4. The nonlinear analysis from the proposed method is very effective and efficient which can be used for preliminary prediction.