Frequency Domain Methods

Obstructive sleep apnea (OSA) is common in peripheral arterial disease (PAD) and associates with high mortality after surgery. Since abnormal heart rate variability (HRV) is predictive of postoperative complications, we investigated the relations of HRV with PAD, OSA and major adverse cardiovascular and cerebrovascular events (MACCE). Seventy-five patients (67±9 years) scheduled for sub-inguinal revascularization and 15 controls (63±6 years) underwent polysomnography and HRV analyses. OSA with an apneahypopnea index (AHI) !20/hour was considered significant. HRV was measured during wakefulness, S2, S3-4 and rapid eye movement (REM) sleep with time and frequency domain methods including beat-to-beat variability, low frequency (LF) and high frequency (HF) power, and detrended fluctuation analysis (DFA). MACCE was defined as cardiac death, myocardial infarction, coronary revascularization, hospitalized angina pectoris and stroke. Thirty-six patients (48%) had AHI!20/hour. During follow-up (median 52 months), 22 patients (29%) suffered a MACCE. Compared to controls, fractal correlation of HRV (scaling exponent alpha 1 measured with DFA) was weaker during S2 and evening wakefulness in all subgroups (+/-AHI!20/hour, +/-MACCE) but only in patients with AHI!20/hour during

Investigating gear damages using vibration signal is a subject of a high interest, because gears vibration signals are complex and difficult to understand. A failure diagnosis of gearbox based on Fourier analysis of the vibration produced by speed reducers has shown its limits in terms of spectral resolution. In the present paper, a comparative study of the performances of various different methods of fault diagnosis of helicopter gearbox gear is carried out. The results are highlighted on the basis of real data recorded during a helicopter flight and have showed that cepstral analysis is most effective technique in detecting gearbox gear faults. Keywords Time domain analysis Á Fast Fourier transforms (FFT) Á Amplitude modulation analysis Á Cepstral analysis Á Synchronous averaging technique Á Helicopter gears

This paper will review practical techniques and procedures employed to quiet gearboxes and transmission units. The author prefers solving the gear noise problem at the very source to introduce an enclosure as a means to reduce radiated noise, which seems to be easy but its effect on the sound pressure level is small. The gearbox noise problem solution is focused on the improvement of gear design; on the verification of its effect on the radiated noise and the determination of the gears' contribution to the truck's or car's overall noise levels and on the analytical and/or numerical computer-based tools needed to perform the signal processing and diagnostics of geared axis systems. All of the analytical methods are based on the time and frequency domain approach. Special care is addressed to the smoothness of the drive resulting from the transmission error variation during a mesh cycle. This paper will review the progress in technique of the gear angular vibration analysi...

In this work, comparison of high contact ratio gears and low contact ratio gears has been done for the same load taking different parameters like contact stresses (equivalent stresses), deformation, strain, frictional stresses and pressure on each tooth into consideration. results shown that the performance of HCR is much better when compared to the LCR gear set. total deformation, stress, strain, frictional stress, and pressure values r found to be much lower for HCR gear set compared to LCR gear set. This is due to the load on each pair of tooth in HCR is less than the same in LCR gear set

Vehicle NVH (Noise Vibration & Harshness) is of continued concern to customers in this increasingly competitive market and driveline NVH performance is a key factor in overall vehicle quality. A typical way to increase this quality is the use of axle end of line test stands that utilize NVH signal analysis methods to offer pass/fail criteria. In the manufacturing environment there are high costs associated with axle assemblies that are rejected due to the amount of time for NVH analysis to determine root cause for the failure. Of more interest to both product development and manufacturing activities is the ability to understand the root cause of the failures from the axle end of line test stand. This information can improve the manufacturing process by eliminating errors, streamlining rebuild activities, aiding in product design improvements, and in turn reducing cost. This paper describes the activities used to identify specific bearing defects for rear axle application through signal analysis of end of line test stand data. Measurements of both nominal and intentionally mis-built (head, tail, & differential) bearings, with known defect characteristics, were used to develop the analysis techniques. Various bearing fault frequencies and their harmonics of the NVH signatures were used to identify the specific failure mode for each of the misbuilt (head, tail, & differential) bearings, which included rolling element pass defects, and cage spin defects.

This work proposes a novel approach to extend blind identification methods to signals recorded in speed-varying conditions. The excitation/transfer separation is useful for many reasons. First numerical models require knowledge of physical parameters often estimated experimentally. Moreover, for example in the case of gearboxes, the extraction of "noisy" excitation components (gear whine) enables the access to other interesting excitations. Classical blind identification methods are developed for stationary signals. In our case, the emphasis of the work is on vibration response of a mechanical system in non-stationary conditions, which makes the originality of this paper. Our technique aims to separate the excitation from the transfer in four steps by exploiting both time and angular domains. The interest is on gearbox vibration signals, modelled with a convolution between an impulse response and an excitation composed of a deterministic part depending on instantaneous "angle" frequency and a broadband random part. In the Campbell diagram in time domain (represented in speed vs. frequency in Hz) order harmonics increase proportionally to speed and structural resonances are located at constant frequencies. After angular resampling, the Campbell diagram is presented in speed vs. "angle" frequency in events/revolution. Order harmonics are then located at constant "angle" frequencies and structural resonances appear as hyperbolas. In this method the characteristics of both domains are exploited : • In the first step the Campbell diagram is considered in angular domain and a comb filter is applied to first extract the orders. • In a second step blind identification methods are applied in time domain to extract the impulse response, supposing that excitation is white only (harmonics being removed in the first step). • In the third step the signal is whitened to remove the influence of the impulse response. The excitation can then be extracted in angular domain by separating deterministic and random parts (using an angle comb filter). The validity and the limits of this approach are analysed on a gearbox signal model, with a discussion on the interest of first extracting the deterministic part of the excitation. The method is then applied on real data measured on a gearbox.

The Gear Whine Noise (GWN), as well as the gear rattle noise, is one of the main vibro-acoustic phenomena of gearbox. For that reason GWN is an important problem to solve or to reduce or to "manage" at least in an appropriate way. Running conditions of the gear-box were reproduced by means of an innovative test-bed, which allows to perform the test without using the engine. This means that the noise and vibration due to the engine were not introduced into the system. From this point of view we say that the present study was performed in optical Noise, Vibration and Harshness.

16 www.ijeas.org AbstractIn this paper the study of dynamic performance analysis of Involute helical gear with defects is presented. Analysis is carried out to solve noise and vibration problems. It improves reliability and safety of the gear. This paper proposes study of the common failures of gearbox and vibration analysis related to gearbox defects and its diagnosis, distinct vibration pattern related to gearing defect and noise level of gearbox and its pattern due to defects in gear during its working. For this, we designed and developed small gearbox and experimental setup for study of vibration analysis then analysed the vibration signature of a healthy gearbox. Created different gear defects (Defects selected for this is pitting, tooth breakage in steps from 25% of one tooth breakage up to complete one tooth breakage and scoring.) on gear that normally occurs in a gearbox during power transmission and analysed each defect by vibration signal analysis, compared a faulty vibrat...

The Gear Whine Noise (GWN), as well as the gear rattle noise, is one of the main vibro-acoustic phenomena of gearbox. For that reason GWN is an important problem to solve or to reduce or to "manage" at least in an appropriate way. Running conditions of the gear-box were reproduced by means of an innovative test-bed, which allows to perform the test without using the engine. This means that the noise and vibration due to the engine were not introduced into the system. From this point of view we say that the present study was performed in optical Noise, Vibration and Harshness.

This article reports the results of several tests, carried out in optical gear whine noise to evaluate the difference between the vibro-acoustic emission of two gearboxes of the same type, one, in particular, showing a micro-geometrical error located on the side of tooth and characterized by lack of material at the base of the tooth due to a slight drift of the production process. The tests were carried out in fourth gear. The user of the vehicle, in fact, seems to be particularly attentive to the vibro-acoustic emissions. For that reason the major car manufacturers devote significant resources to the Noise, Vibration and Harshness sector, which deals with noise and vibration of the vehicle to the vehicle driving comfort.

In general, for the diagnostics of a faulty machine several diagnostic methods can be considered. In this paper, the benefits and limitations of the acoustic diagnostic methods are shown, based on a noise problem of a real car with a 6-speed manual transmission. The possible diagnostic methods will be pointed out, and an analysis will be performed with noise and vibration measurements on the car. Based on the analysis conclusions regarding the benefits and limits of the used diagnostic method will be drawn.

The topic of this paper is focused at noise and vibration of car and truck transmission units for the diversity of their operational condition in loading and rotational speed. This paper discusses the assessment of noise and vibration sources, measurement methods, and diagnostics. The analysis tools for gearbox inspection are based on the frequency and time domain methods. Noise and vibration problems in gearing are mainly concerned with the smoothness of the drive. The oscillation of the tooth contact stiffness causes the self excited angular vibration and consequently noise. The paper is dealing with the problem of the transmission error measurement as a part of the experimental study of the tooth contact dynamics. To keep the radiated noise under control, the effect of load, the gear contact ratio and the tooth surface modification on noise and vibration are illustrated by measurement examples giving an idea how to reduce transmission noise.

The paper deals with the dynamic transmission error (TE) measurements of a simple gear set at gearbox operational conditions that means under load and during rotation. The analysis method is focused on the processing of impulse signals generated by incremental rotary encoders attached to both the gears in mesh. The analysis technique benefits from demodulation of a phase-modulated signal. The theory is illustrated by experiments with a car gearbox and measurement errors are discussed.

This paper aims to present causality tests between financial system and economic growth in the frequency domain for Brazil, India, France, Japan, United States and Korea. This approach allows to capture nonlinearities in the direction of the causality from short to longer terms. Indeed the results show variation in causality tests depending on the frequency of the cycles considered. There is evidence of that the financial system is a causal factor for long-term economic growth in Brazil. For developed countries (United States and France) that causality is not as important once controlled by capital accumulation.

Este trabalho tem como objetivo apresentar testes de causalidade entre sistema fi nanceiro e crescimento econômico no domínio da frequência, para séries temporais brasileiras. Esta abordagem permite não-linearidades associadas a mudanças de direção na causalidade do curto para o longo prazo. Para avaliar esta relação no Brasil, o artigo emprega séries temporais para o PIB per capita e de crédito bancário entre 1960 e 2010. Os resultados mostram variação nos testes de causalidade dependendo da frequência dos ciclos considerados, havendo evidências de que o sistema fi nanceiro seja um fator causal para o crescimento econômico no Brasil, mas apenas no longo prazo. Palavras-chave crescimento econômico; sistema fi nanceiro; séries de tempo no domínio da frequência.

O estudo de séries temporais com características de longa dependência iniciou-se com trabalhos de Hurst (1951). Esta característica ocorre, por exemplo, em modelos ARFIMA (p,d, q), quando o parâmetro de diferenciação d pertence ao intervalo (0;0.5). Por esta razão, estes modelos são também referenciados como "fracionariamente integrados". O processo ARFIMA (0,d,0) (onde p e q, os graus dos polinômios autoregressivo e médias móveis, são ambos zero) é representado por (1-β) d X t = ε t , para todo t ∈ Z. Nesta expressão o processo {ε t } t∈Ζ é um ruído branco, isto é, é um conjunto de variáveis aleatórias com uma distribuição fixada, com média constante E (ε t) = µ ε , usualmente assumida zero, variância constante Var(ε t)=σ 2 e a função de auto-covariância γ ∈ k = Cov (ε t, ε t+k) = 0, para todo k≠0. E β é o operador de defasagem, isto é, β (X t) = (X t-1). Para a estimação do d usaremos o método de Geweke e Porter-Hudak (GPH) que é um método de regressão linear usando a função periodograma. Nos processos bivariados fracionariamente integrados o parâmetro de diferenciação d será agora um vetor d = (d 1, d 2) bi-dimesional. O objetivo deste trabalho consiste em estudar o comportamento do método de estimação GPH para o parâmetro d de um processo bivariado fracionariamente integrado, agora um vetor bi-dimensional. Queremos analisar também o efeito do método de estimação GPH para d quando utilizamos as correlações canônicas dos dois processos. (BIC/UFRGS).

Neste artigo consideramos modelos de espectro misto da forma p Zt= L A;cos(w;t+</>;) +{t, tEZ, j=l onde p não é necessariamente conhecido c, para cada 1 :::; j � p, A j é constante desconhecida, Wj é freqüência desconhecida em (-71", 7rJ e a fase 4>j é uma variável aleatória uniformemente distribuída em (-1\",7r] independente cada uma entre si e do processo de ruído. O objetivo é estimar as freqüências da parte discreta do espectro. Esta estimativa é obtida através de um método recursivo cujos parâmetros são atualizados a cada passo. Os cosenos das freqüências são obtidos como pontos fixos atratores de uma determinada aplicação.