延續Hong等人(2003b)對於直徑9 mm模型土釘的試驗，本研究進一步針對另三種較大直徑(12、16及19 mm)之模型土釘分別進行單釘及雙釘拉出試驗，並加入陳威志(2004)之單釘試驗，以充分之單釘試驗結果重新詮釋土釘表面粗糙因子，並由雙釘拉出試驗結果歸納探討不同影響參數對於雙釘拉出行為之影響。 經由試驗結果得到以下結論：(1)尖峰及殘餘視摩擦係數皆隨螺紋牙距增大而增加，粗糙土釘視摩擦係數為光滑土釘之4~7倍；粗糙土釘尖峰及殘餘視摩擦係數皆隨直徑增加而遞減，遞減趨勢於直徑12 mm至16 mm趨於平緩，光滑土釘則不受直徑影響。(2)尖峰及殘餘視摩擦係數受覆土壓力影響皆不明顯。(3)雙釘試驗於密間距下，應力有重疊及相互干擾之現象，造成拉出力降低；土釘埋設間距對於尖峰與殘餘拉出力影響顯著，當群釘效率小於100 %時，群釘效率與水平間距約呈線性正比關係。(4)雙釘間互制效應及影響範圍受粗糙因子及土釘直徑影響極為明顯，尖峰狀態下，達到100%群釘效率所需正規化間距S/D與粗糙因子成正比關係，與土釘直徑則成反比關係；但殘餘狀態下之群釘效率在目前試驗條件下幾乎皆無法達到100%，表示殘餘狀態下雙釘影響範圍較尖峰時為大。 透過充分之單釘試驗結果，重新詮釋一粗糙因子R''，由視摩擦係數與R''關係預測單釘尖峰拉出力，整體誤差約6%，較陳威志(2004)經驗式預測精度為高，且於推估粗糙因子隨土釘直徑變化之趨勢上較為合理。由一系列雙釘試驗結果透過多元迴歸分析所得之經驗模式，於預測S/D時精確性極高，期可供實務界參考。 According to the previously laboratory experiment results that the diameter of model nail was 9 mm (Hong et al. 2003b), the purpose of this study is to investigate the pullout behavior of larger-diameter nails (12, 16 and 19 mm) through the tests of single- and double nails. The conclusions via the tests are summarized: (1) The peak and residual apparent friction coefficients at the soil-nail interface are all increased with increase of the screw pitch, and the apparent friction coefficients of rough nails are 4-7 times greater than that of smooth nails. The peak and residual apparent friction coefficients of rough nails are decreased with increase of the nail diameter and reaches a constant value, but such of smooth nails are not influenced by the nail diameter. (2) Influence of surcharge stress on the nail for the peak and residual apparent friction coefficients can be neglected. (3) Because the stress influence around the nail has the phenomenon of interference and superposition in the experiment of double-nails under dense spacing, the pullout force of nail will reduce. When the group efficiency is smaller than 100%, there would be nearly a linear relation between the group efficiency and the horizontal spacing. (4) The diameter-spacing ratio S/D as peak group efficiency reaches 100% is linear proportion to the surface roughness factor, but it is an inverse proportion to the nail diameter. In present condition of apparatus, residual group efficiency cannot reach 100%. This study re-describes the surface roughness factor R'' through sufficient tests, and to predict the pullout forces of single-nail by the relation between the apparent friction coefficients and R''. The predict values are 6% smaller than the value of the laboratory experiment. In addition, the experience model of double-nail obtained by multiple-variables regression could accurately predict the diameter-spacing ratio S/D as peak group efficiency reaches 100%.
