دانلود مقاله بررسی تجربی ممان خمشی گذرا شمع‌ در جریان روان‌ گرایی لرزه‌ای

چکیده

1. مقدمه

2. مدل سازی فیزیکی برهمکنش گذرا توده-خاک

3. نتایج تجربی

4. بحث و نتیجه گیری

ضمیمه - A. مقیاس بندی مبتنی بر مکانیک

ضمیمه B.

منابع

Abstract

1. Introduction

2. Physical modelling of transient pile-soil interaction

3. Experimental results

4. Discussion and conclusion

Appendix – A. Mechanics based scaling

APPENDIX – B.

References

چکیده

     هدف این مقاله بررسی رفتار سازه‌های پشتیبان شمع در خاک‌های روان‌پذیر است، به‌ویژه زمانی که خاک اطراف شمع از حالت بدون روان‌گرایی به حالت روان‌گرایی کامل منتقل می‌شود. مجموعه‌ای از آزمایش‌های میز تکان‌دهنده بر روی چهار سازه پشتیبان شمع انجام شد که در معرض حرکات ورودی مختلف و در نتیجه زمان‌های متفاوت برای رسیدن به حالت روان‌گرایی کامل قرار گرفتند. گشتاور خمشی شمع ها در طول فاز گذرا با موارد پیش بینی شده در مراحل قبل و بعد از مایع سازی مقایسه می شود. نتایج تجربی نشان داد که حداکثر گشتاور خمشی ممکن است در مرحله گذرا (یعنی در حین ایجاد فشار منفذی اضافی قبل از اینکه خاک کاملاً مایع شود) رخ دهد. احتمالاً، تقویت مشاهده شده در لنگر خمشی ناشی از اثر تنظیم بین فرکانس غالب حرکت ورودی و فرکانس ساختار متکی به شمع است که به تدریج در طول فرآیند روان‌گرایی کاهش می‌یابد. نتایج با استفاده از یک چارچوب غیر بعدی ارائه می‌شوند که پارامترهای آن از مکانیک حاکم مشتق شده‌اند. یک پارامتر جدید TAF (ضریب تقویت گذرا) برای پیش بینی گشتاور خمشی طراحی در طول فاز گذرا تعریف شده است. نشان داده شده است که گشتاور خمشی گذرا را می توان از پارامتر تازه معرفی شده TAF و مقادیر حداکثر گشتاورهای خمشی را در مراحل قبل و بعد از مایع سازی بدست آورد. مشخص شد که TAF تابعی از دو پارامتر به راحتی قابل حصول است: (الف) زمان صرف شده برای رسیدن به مایع شدن کامل، این می تواند از طریق تجزیه و تحلیل پاسخ سایت به دست آید. ب) طویل شدن دوره طبیعی ارتعاش که به صورت نسبت دوره زمانی سازه در حالت روانگرایی کامل به دوره زمانی در حالت روانگرایی صفر بیان می شود. در نهایت، مفاهیم عملی یافته های اصلی مورد بحث قرار می گیرد.

توجه! این متن ترجمه ماشینی بوده و توسط مترجمین ای ترجمه، ترجمه نشده است.

Abstract

     The purpose of this paper is to study the behavior of pile-supported structures in liquefiable soils, specifically when the soil surrounding the pile transits from no-liquefaction to full-liquefaction. A series of shaking table tests were performed on four pile-supported structures subjected to different input motions and, as a result, with different times to reach full-liquefaction. The bending moment of the piles during the transient phase is compared with those predicted in pre- and post-liquefaction stages. The experimental results showed that the maximum bending moment may occur during the transient phase (i.e. during the development of excess pore pressure before the soil is fully liquefied). Arguably, the observed amplification in bending moment is caused by the the tuning effect between the predominant frequency of the input motion and the frequency of the pile-supported structure, which is progressively decreasing during the liquefaction process. Results are presented using a non-dimensional framework whose parameters are derived from the governing mechanics. A new parameter TAF (Transient Amplification Factor) is defined to predict the design bending moment during the transient phase. It is shown that the transient bending moment can be obtained from the newly introduced parameter TAF, and the values of maximum bending moments in the pre- and post-liquefaction stages. It is found that TAF is a function of two easily obtainable parameters: (a) time taken to reach full liquefaction, this can be obtained through site response analysis; (b) elongation of natural period of vibration, expressed as ratio of the time period of the structure at full liquefaction to the time period at zero-liquefaction. Finally, practical implications of the main findings are discussed.

Introduction

     The dynamic behaviour of pile-supported structures founded in liquefiable deposits is still an area of active research [[1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30]] due to the poor performance of piled foundations as observed in most of the recent earthquakes [9,31]. The current methods of pile design - see for Example codes of practices such as JRA [32,33] NEHRP [34], IS-1893 [35], focus on avoiding bending, buckling and settlement failures. The effect of dynamic soil-structure-interaction (SSI) effects is taken into account by means of empirical correlations.

     Fig. 1 shows the two main stages of loading in a typical pile-supported structure during an earthquake. Fig. 1 (a) shows the stage at the start of the shaking, before any build-up of the excess pore water pressure takes place. In this stage, piles are mostly subjected to inertia loads induced by the ground shaking and oscillation of the superstructure. It is known that inertia loads tend to generate high bending moments at relatively shallow depths along the pile. Fig. 2 (a) shows the “Beam on Non-Linear Winkler Foundation" model, in which the effect due to soil-structure interaction is modelled by a set of springs distributed at discrete locations along the length of the pile. Each spring is defined by a relationship between the soil pressure (p) and pile deflection (y) referred to as p-y curve. This method is conventionally used to compute the internal forces (i.e. bending moment and shear force) and pile's deflection. The typical shape of p-y curves for sand in non-liquefied condition is shown in Fig. 2 (b), and further details can be found in API [36]. However, in saturated loose to medium dense sand,the ground shaking induce a gradual increase in pore water pressure, resulting in the soil to progressively lose its strength and stiffness. When the excess pore pressure equalises the overburden pressure, the soil is said to be in a full-liquefaction condition; at full liquefaction the foundation loses the support from its sorrounding soil and the piles tend to act as an unsupported column over the liquefied layer, see Refs. [37,38]. Fig. 1c shows the time history of a real earthquake together with excess pore water pressure profile, which can be either measured in experiments or computed by means of site response analyses. From the figure it can be noted that the onset of liquefaction occurs in a finite amount of time, whose duration depends mostly on the input motion characteristics and the density of the soil; the latter is typically expressed in terms of relative density. Typical values of time to reach liquefaction range between 6 and 15s, however, the actual time can be estimated by means of nonlinear finite element analysis. 

Experimental results

3.1. Frequency domain analysis

     To quantify the change in vibration characteristics of the models, the experimental data was first analysed in frequency domain. Before the start of each test, the natural frequencies of the models were estimated from free vibration tests performed by exciting the structure with an impact hammer. The frequency response functions FRFs obtained from the free vibration tests are depicted in darker lines (denoted by before shaking) in Fig. 4, Fig. 5, Fig. 6. In the same figures, the power spectral density (PSD) of the input motion and FRF estimated after liquefaction is shown for comparison. It may be observed that after liquefaction the natural frequency of the models reduced significantly, which may be attributed to the development of excess pore pressure, and in some cases formation of plastic hinges in the piles. From Table 3 it is interesting to note that the four models failed in tests CH1 to CH4. (see Fig. 4 for frequency response). Other tests listed in Table 3 were carried out to investigate the sole effect of liquefaction on the variation in natural frequency of the models and transient bending moments. In these tests, formation of plastic hinges was deliberately avoided with lower pile-head mass. The results shown in Fig. 5, Fig. 6 indicate that the natural frequency of the models reduced to about half due to subsurface liquefaction. As the model transited from higher to lower frequencies it was likely that the seismic response amplified due to a temporarily matching between natural frequency of the models and predominant frequency of the input motion. This aspect is further investigated in the next section by considering the bending moment time histories along the piles.