核设备抗震鉴定试验中功率谱密度检验方法分析评估

为满足核设备抗震鉴定试验中输入运动的功率谱密度(PSD)要求,基于对规范背景和目标PSD算法的调研以及典型算例的对比分析,对PSD的检验方法进行分析评估。结果表明,检验PSD最为直观的方法即对比输入运动PSD与目标PSD;根据各类目标PSD算法的结果精度、保守性及其规范依据,推荐使用2014版美国核管理委员会标准审查大纲(SRP)3.7.1节附录B中人工合成时程的方式来计算目标PSD:虽然该算法通常适用于核电厂的厂址设计反应谱,但对于设备抗震鉴定反应谱,仅需将人工合成时程的目标反应谱替换为鉴定反应谱即可;采用本文推荐方法计算目标PSD时,设备抗震鉴定输入运动的PSD检验应与SRP 3.7.1保持一致,即在0.3 Hz到目标反应谱的最高截断频率范围内包络目标PSD的70%。     

核电厂设备抗震设计标准功率谱密度的生成方法

对于核电厂设备抗震设计的输入地震波,通常要求其同时包络目标反应谱（RRS）和标准功率谱密度（PSD）,然而目前国内外对标准PSD缺少统一的算法。在美国核管会标准审查大纲（SRP）3.7.1建议的标准PSD生成方法基础上,优化了迭代过程,提出了一个改进的标准PSD合成方法,并在2个核电设备RRS算例上实现了该方法。结果显示改进的标准PSD生成方法与RRS匹配程度较高,同时计算快速、简便,收敛精度与基于随机振动理论方法计算的结果相似,此法可以作为核电厂设备抗震设计输入人工地震波的标准PSD检验依据。      

核电厂楼面反应谱的人工合成地震动时程方法对比分析

为满足核电厂核岛内部结构、系统和部件(SSCs)的抗震分析需求,采用2种典型算法(频域调幅法和时域叠加法)来合成匹配核岛内部楼面反应谱的人工时程。通过算例分析,对比2种算法在反应谱匹配精度、时程非平稳特性、相位、基线漂移等方面的优缺点及其与美国核管理委员会(NRC)颁布的标准审查大纲(SRP)第3.7.1节要求的符合性。结果表明:在反应谱匹配精度(特别是多阻尼反应谱)、时程非平稳特性的改变、基线漂移的控制上,时域叠加法均优于频域调幅法;但频域调幅法可较完整地保留种子时程的相位信息,适合于SRP 3.7.1对时程相位有特殊要求的分析。     

利用HHT产生与多阻尼楼板谱匹配的地震波

美国原子能管理委员会（USNRC）规范规定了用于核电厂抗震分析和设计的地震波要求。在抗震分析和设计中，采用的地震波可与多阻尼目标反应谱匹配，也可与单阻尼目标反应谱匹配。然而，在对核电设备和部件进行动力时程分析时，则需要与多阻尼目标楼板谱匹配的地震波。基于此问题，提出利用希尔伯特-黄变换（HHT）方法，通过修改种子地震波的频率和振幅信息，使之与多阻尼目标楼板谱匹配，且完全符合USNRC规范的匹配标准，从而为核电设备和部件的地震安全评估提供合适的地震激励。      

同时满足反应谱匹配和功率谱密度包络要求的人工时程生成算法

人工时程是核电站反应堆厂房楼层反应谱计算的关键输入数据.采用单组人工时程分析法不仅需要匹配目标反应谱,而且要满足功率谱密度的包络要求.本文提出同时满足目标反应谱匹配和功率谱密度包络要求的人工时程生成算法.该方法在分别与目标反应谱和功率谱密度对应的有效频率带宽内调整傅立叶幅值谱,并通过迭代达到要求.数值算例表明,该算法得到的人工时程在对目标反应谱拟合精度较高的同时,满足对功率谱密度的包络要求.     

AP1000设计地震反应谱在具体厂址评价中的应用

分析AP1000设计地震反应谱（CSDRS）与各相关导则中定义的反应谱的对应关系,指出在特定厂址评价中,应基于同一标高比较厂址特定设计反应谱（SRS）和AP1000 CSDRS。基于5种设计场地模型将AP1000 CSDRS反演至设计基岩处和核岛结构基础底部,计算得到设计基岩处和结构基础底部的AP1000设计谱。计算结果表明,AP1000 CSDRS不能包络已有核电厂核岛结构抗震设计采用的0.2g标定的RG1.60的设计反应谱;若在非硬质基岩场地建造AP1000核岛结构,应进行AP1000 CSDRS的保守性分析。     

核设备抗震鉴定试验目标PSD生成方法研究

为避免抗震鉴定试验用人工时程能谱特性失真,需满足功率谱密度80%包络的要求。从楼层反应谱、傅里叶幅值谱和功率谱密度之间的互相关系入手,开展楼层目标功率谱密度生成方法的研究。分别采用Kaul公式法和SRP3.7.1附录A快速傅里叶变换法,将楼层反应谱和人工加速度时程转换为功率谱密度。通过对比分析不同方法的基本假设和不确定性以及获得功率谱密度的形状、峰值的差异,推荐用生成楼层反应谱的原始响应时程进行快速傅里叶变换的方法作为目标功率谱密度的生成方法。     

我国核电厂抗震设计反应谱和RG 1.60设计反应谱的比较分析

设计反应谱对评价核电厂在地震作用下的安全性极为重要。本文从统计核电厂抗震设计标准反应谱时选取的强震数据及统计方法两个方面,分析比较了美国RG 1.60设计反应谱和我国核电厂抗震设计规范反应谱的异同。通过对比分析,深入理解核电厂抗震设计反应谱的提出需考虑的关键因素,为核电厂抗震设计和审评工作提供参考。     

拟合目标反应谱的高精度人工加速度时程生成算法及应用

加速度时程是核电厂结构和部件开展抗震分析和试验的重要输入。本文基于改进的窄带时程构造形式,综合应用频域法和时域法,提出一种拟合目标反应谱的高精度人工加速度时程生成算法,通过RG1.60标准反应谱和两个复杂度不同的楼层反应谱拟合测试以验证算法的有效性和稳健性。结果表明：对目标反应谱的拟合相对偏差最大不超过10%;迭代次数少,拟合收敛性好。本文提出的算法可生成高精度的符合法规要求的人工加速度时程,对于核电工程抗震设计具有重要的应用价值。     

横掠管束的两相流湍流激振力的包络谱研究

蒸汽发生器等换热器中存在大量受两相流冲刷的管束结构。计算两相流湍流导致的管束振动响应时,需先确定流体激振力的功率谱密度（PSD）。由于缺乏有公认合理的两相流激振力的无量纲归一化方法,从而导致没有可供工程应用的两相流激振力的PSD包络谱。本文通过修改de Langre无量纲归一化方法中的混合物流速定义,获得1组新的两相流激振力PSD的包络谱。通过将本文包络谱与基于单相流的包络谱和de Langre包络谱的对比,并针对非均匀横向两相流作用下的蒸汽发生器U型传热管,计算了3种包络谱作用下的湍流激振响应。结果表明,基于单相流激振力PSD包络谱作为输入计算两相流湍流激振响应并不保守,而本文包络谱在保证安全性的前提下适当降低了de Langre包络谱过高的保守性。     

Artificial ground motion compatible with specified peak ground displacement and target multi-damping response spectra

With respect to the design ground motion of nuclear power plant (NPP), the Regular Guide 1.60 of the US not only defined the standard multi-damping response spectra, i.e. the RG1.60 spectra, but also definitely prescribed the peak ground displacement (PGD) value corresponding to the standard spectra. However, in the engineering practice of generating multi-damping-spectra-compatible artificial ground motion for the seismic design of NPP, the PGD value had been neglected. Addressing this issue, this paper proposed a synthesizing method which generates the artificial ground motion compatible with not only the target multi-damping response spectra but also the specified PGD value. Firstly, by the transfer formula between the power spectrum and the response spectrum, an initial uniformly modulated acceleration time history is synthesized by multiplying the stationary Gaussian process with the prescribed intensity envelope to simulate the amplitude-non-stationarity of earthquake ground motion. And then by superimposing a series of narrow-band time histories in the time domain, the initial time history is modified in the iterative manner to match the target PGD as well as the target multi-damping spectra with the pre-specified matching precisions. Numerical examples are provided to demonstrate the matching precisions of the proposed method to the target values.     

A method for the generation of artificial earthquake accelerograms

The design spectra are the main specifications for the seismic design of nuclear power plants. As more and more consideration is given to non-linearities in structural analysis, there is a need for the generation of time histories compatible with these spectra.The presented method uses a probabilistic model to generate the power spectral density of an equivalent stationary Gaussian process having the design spectrum as expected maximum response. This process is then shaped to simulate a real earthquake. A subsequent iterative process refines the solution to match the target spectrum.Some additional features are also assured: imposed peak acceleration, zero final velocity and displacement, low correlation between successive time histories.The resulting computer programme THGE is quite economical to use; this results from an appropriate spectrum computation technique and the use of the Fast Fourier Transform to connect the time and frequency domains.     

Automated generation of spectrum-compatible artificial time histories

Criteria for the seismic design of nuclear power plants are usually defined in the form of response spectra, and it is often necessary to generate artificial time histories of ground motions to ‘match’ these spectra. A recently developed automated iterative procedure employing a combination of frequency- and time-domain techniques in any desired sequence for the development of spectrum-compatible artificial time histories is presented. The procedure employs four basic steps: (1) generation of an initial (staring) time history using either sinusoidal superposition with an envelope function or specification of a real time history of a recorded ground motion; (2) manipulation of the amplitude and phase of the Fourier transform representation of this time history, and generation of successive time histories which have response spectra converging to the target design spectrum; (3) manipulation of the amplitude and phase of the Fourier transform representation only locally in areas where the peaks of the computed spectrum have larger magnitudes of deviations than desirable; and (4) manipulation of local areas of the latest time history. Computer program EDAC/SEQGEN, which completely automates the above steps, is described. Results are presented for the development of two artificial time histories with corresponding spectra matching the target USNRC Regulatory Guide 1.60 response spectra within about 5–7%.     

Effects of ductility on seismic response of piping systems and their implication on design and qualification

This study is concerned with the inelastic seismic response of nuclear power plant piping systems. Two systems are examined. The first one is an idealized four-equal-span pipe run and the second one consists of two configurations modified from an existing pipe run. Detailed finite element seismic time history analyses are performed using the computer program. By varying the various geometrical and physical parameters, calculations are made for a total of 76 cases. The results show that ductility generally contributes to reducing the response of piping systems. An empirical relation between the support load reduction factor and support ductility demand is given and a chart and simple procedures are suggested for the design and qualification of piping supports taking ductility into consideration.