樁腳式平台揚壓力之實驗研究

Abstract

波浪衝擊平台甲板為一極複雜的流體,剛體互制作用,不易由波浪理論求得解析解, 經過許多簡化,假設的數值計算方法,亦難以獲得真貌.故本文藉由模型實驗,希冀了解進行 波對海洋平台結構的基本揚壓性質. 本研究設計了光滑平板(flat slab)與加勁平板 (rib-attached slab)二種型式模型,在等水深固定平板長度條件下,於二維斷面水槽造出 各種不同條件的中間水深規則波,探討影響揚壓力的因子(波高,水深,波長)等彼此的關係, 並與Wang(1970),Goda(1970)過達(1988)等曾提出相關理論與經驗公式比較.且從機率分佈 的角度分析揚壓機率分佈,發現衝擊揚壓並不純然為一隨機現象. 從無因次揚壓參數 (P/rH)隨相對超高(hc/H)變化的分析結果,在相對水深大於0.15時,無因次揚壓參數之最大 值發生在相對超高為0.3時,不同於部份重波的揚壓特性,如以此相對超高為分界,在其上下 會呈良好的指數函數曲線變化.本研究以相對超高為變數經非線性迴歸所得之經驗迴歸式, 與往昔理論與經驗公式比較,證實有解釋性良好的優點. Wave impacting on platform was a extreme complex interaction betwe en fluid and rigid body. It was not easy to get an analytic solution b y wave theory. Even by many simplified and assumed numerical methods, it w as hard to get the whole fact. This paper used several designed models t o study the basic properties of uplift pressuredue to progressing wa ves acting on ocean platform. In this research, we designed two types of model : one was flat slab, and another was rib-attached slab. Under the condition of uniform water depth and fixing platform length, regular int ermediate depth water waves were produced by a wave maker of piston type s in two-dimensional flume. With thiskind of wave, we discussded the rel ationship about several factors thatmight influence the uplift pressure i ncluding wave height, wave length, and water depth. And we also compared th ese theoritical and experiential formulae with which proposed by Wang(1970), Goda(1970), and Kuo(1988).By the way, the result show that impact uplift pres sure was not a naturalrandom event which analyzed the uplift pressure from t he aspect of probability distribution. In the result that analyzed be tween the dimensionless uplift pressure factor(P/rH) and hc/H ratio, the m aximum of P/rH occured as the relative water depth over 0.15 and the hc/H rat io was 0.3, and this was a special character of uplift pressure which differ from partial clapotis. If we divided by this hc/H ratio, a fair exponential curve could be obtained. This research studied the hc/H ratio calculated by n onlinear regression, and we compared theresult with the previous theries and e xponential formulae, then the result showed an advantage of well explanation.