Untitled

Johannesson, Pär; Thomas, Jean-Jacques

doi:10.1023/a:1015277305308

Cited by 51 publications

(19 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In EVE, the load spectrum made up of extreme values is extrapolated. Johannesson and Thomas [17] and Johannesson [40] divided EVE into two branches: the extrapolation in the time domain (EVET, based on the extraction methods of block maxima method and peak over threshold [41]) and in the rainflow domain (EVER, based on extraction method of level upcrossings [10]). So the process of EVE is divided into two steps: extract extreme values from a load time history and fit the extracted data.…”

Section: Extreme-value Extrapolation Methods (Eve)mentioning

confidence: 99%

“…Level upcrossings (LU): According to Johannesson and Thomas [17], LU is proposed to obtain the maxima and minima of the load cycles, then determine the limiting shape of the rainflow matrix (RFM) and estimate the limiting RFM G [17]:…”

Section: A Review Of the Extrapolation Methods In Load Spectrum Compilingmentioning

confidence: 99%

“…This approach is based on an asymptotic theory for the crossings of extreme (high and low) levels. First, obtain a measured RFM F [17]:…”

Section: A Review Of the Extrapolation Methods In Load Spectrum Compilingmentioning

confidence: 99%

“…The traditional load spectrum compiling method multiplies a short-term load spectrum with a constant proportionality coefficient [15] to [17]. Since only the data measured in a finite time is repeated, the extreme loads that cannot be measured and have a greater impact on damage are ignored.…”

Section: Introductionmentioning

confidence: 99%

“…In practice, a long-term load spectrum contains the complete load information, but it is difficult to be directly measured due to the restrictions of testing technology, as well as time and cost. Therefore, it is necessary to obtain a long-term load spectrum based on a short-term one.The traditional load spectrum compiling method multiplies a short-term load spectrum with a constant proportionality coefficient [15] to [17]. Since only the data measured in a finite time is repeated, the extreme loads that cannot be measured and have a greater impact on damage are ignored.…”

mentioning

confidence: 99%

See 4 more Smart Citations

A Review of the Extrapolation Method in Load Spectrum Compiling

Wang¹,

Hong-bin²,

Li³

et al. 2016

SV-JME

View full text Add to dashboard Cite

60 INTRODUCTIONIn engineering, many mechanical structures and components are subjected to complex and random loads, which determine the fatigue reliability and life of the machinery [1] and [2]. Thus, it is indispensable to conduct fatigue analysis and life prediction of the structures and components based on a load spectrum [3] and [4]. Currently, a load spectrum is widely used in the fields of aerospace [5] . In practice, a long-term load spectrum contains the complete load information, but it is difficult to be directly measured due to the restrictions of testing technology, as well as time and cost. Therefore, it is necessary to obtain a long-term load spectrum based on a short-term one.The traditional load spectrum compiling method multiplies a short-term load spectrum with a constant proportionality coefficient [15] to [17]. Since only the data measured in a finite time is repeated, the extreme loads that cannot be measured and have a greater impact on damage are ignored. Load extrapolation methods can overcome the above limitation of the traditional method. With the development of statistics and computer software, new methods have been applied to load extrapolation. In load spectrum compiling, results may vary from each other with different extrapolation methods. Therefore, selecting an appropriate load extrapolation method is very important, but that is difficult in practice. For a better understanding of the methods and to provide selection guidance, several commonly used extrapolation methods are reviewed and summarized based on the literature and illustrations in this paper.The extrapolation methods are classified as the parametric extrapolation method (PE), nonparametric extrapolation method (NPE) and quantile extrapolation method (QE). In PE, sample data is supposed to obey a known distribution, and the parameters in the function are estimated according to the load sample. In NPE, an extrapolated result is obtained because the density distribution with an arbitrary shape can be received based on a nonparametric density estimation. When the sample data has different load characteristics due to different working conditions and different operating behaviors in the testing process, QE can break the data into a series of clusters and computes the damage of each rainflow matrix. The literature and illustrations are presented to evaluate the extrapolation methods and the characteristics of various extrapolation methods, such as the critical factors, the advantages and disadvantages, and the application ranges, are summarized. Some potential research prospects are

show abstract