Generation of sine on random vibrations for multi-axial fatigue tests
Introduction
Vibration fatigue analysis for structures and mechanical components subjected to complex dynamic environments is a problem of general concern for the engineering community. It is obvious that the vibration environments are crucial to fatigue life and failure results. Currently, many vibration fatigue tests, as an effective means of investigating durability and reliability of products, are performed by a single exciter to provide a uni-axial base excitation environment [1], [2], [3], [4]. However, by virtue of the facts that many measured vibrations are multi-axial and that coupling relationships exist between different axes manifested as phase and coherence coefficients [5], [6], [7], [8]. Some attempts have been made to conduct multi-axial vibration fatigue tests, which shows the different fatigue damage results compared to uni-axial or sequential axial vibrations [9], [10], [11]. It is therefore necessary to simulate complex multi-axial vibrations for dynamic fatigue test purposes.
To achieve a prescribed multi-axial vibration environment, advanced multi-exciter multi-axis vibration control tests can be conducted. Compared to uni-axial tests, multi-axial loading inputs can provide a proper energy distribution throughout the test structure and excite more vibration modes, which may result in different multi-axial stress/strain distributions and states. It should be noted that the multi-axial concept involved in multi-exciter multi-axis dynamic fatigue tests have two meanings. One is that the vibration directions of test structures are multi-axial, and the other is that the stress states of test structures are multi-axial. There are both differences and connections between the multi-axial vibrations and multi-axial stress states. On the one hand, the stress states caused by multi-axial vibration are generally multi-axial, whereas single degree of freedom excitation may also lead to multi-axial stress states. However, problems which are specifically introduced by multi-axial vibrations are the coherence and phase relationships between the different axial vibrations that may lead to a more complex combination of stress states. This is the key to the multi-exciter multi-axis vibration fatigue tests which is different from the traditional single exciter vibration fatigue problems.
Multi-exciter multi-axis sine on random vibration control is more difficult than a simple form of multi-axial vibrations, such as onefold random, sine or shock. In nature, many vibrations are characterized by sinusoidal signals superimposed on a background random signal with a wideband frequency range, typical applications with simultaneous sine and random environments are helicopters and rocket engines. An example of sine on random vibrations is illustrated in Fig. 1. A sine signal can be determined by the amplitude, frequency and phase in the time domain. But for random signals, the traditional power spectral density in the frequency domain cannot completely describe its time-domain characteristics, such as skewness and kurtosis which may have a non-negligible impact on fatigue life results of structures. In previous works by the authors, the sequential phase modification method is used to obtain the desired multi-channel non-Gaussian random signals with specified skewnesses and kurtoses [12], [13]. This paper presents a nonlinear transformation method based on a Hermite polynomial model to control non-Gaussian characteristics. Different from the sequential phase modification method, the advantage of the nonlinear transformation method is that it can realize multi-channel non-Gaussian random signals transformation simultaneously.
The key issue in sine on random mixed vibration control is to separate the sine and random components from mixed signals. Underwood proposed a tracking filter method to extract the amplitudes and phase angles of the sweeping response sine signals [14]. A Vold-Kalman filter with order tracking is used to accurately separate the sine and random environments from each other and from noise [15]. A correlation integral method was put forward by Zhang with the advantage of high precision and computation speed, but without control of phase differences of sine signals [16]. For the multi-exciter sine on random vibration control, not only the amplitudes of the sine component need to be controlled, but also the phase differences should be taken into consideration. The control of phase differences is far more significant in practice than that of their actual phase angles. This is because that the phase relationships are sometimes very important to the structural dynamic response performances such as stress distributions.
The objective of this paper is to present a closed loop multi-axial sine on random mixed vibration control method for dynamic fatigue tests. An improved correlation integral method is proposed to extract the amplitudes and phase differences of the response sine component. The nonlinear transformation method based on the Hermite polynomial model is employed to generate multi-output non-Gaussian random signals. The drive signals are updated by the control algorithms to produce the specified reference vibration environments within prescribed tolerances. The remainder of the paper is organized as follows. Section 2 elaborates the control method for multi-axial sine on random mixed vibration control test. Discussion about non-Gaussian random component is presented in Section 3. The results from tri-axial sine on random mixed vibration control test are provided in Section 4. Conclusions are finally given in Section 5.
Section snippets
Control method
This section explains the closed loop control method for multi-exciter multi-axis sine on random vibration testing, as shown in Fig. 2. The control objectives are the time-frequency characteristics of response signals that are comprised of response power spectral densities, skewnesses and kurtoses of random component and response amplitudes and phase differences of sine component. The separation of sine on random signals is implemented by an improved correlation integral method, and then a
Discussion
The advantage of the Hermite polynomial model transformation method is that the analytic solutions of skewness and kurtosis of a transformed random signal are provided and it is therefore direct and efficient to control desired skewness and kurtosis. However, the range of application has a limitation which is manifested that when the reference skewness and kurtosis are large, the accuracy of transformed results is low. Fig. 4 shows the kurtosis of before and after transformation by using the Eqs.
Experimental results
A tri-axial sine on random mixed vibration control test is carried out and the results are presented in this section. The test system is shown in Fig. 6 where a computer is used to control the data acquisition and transmission VXI system. The vibrations of the three directions x, y and z of shake table are regarded as the control targets.
The reference spectral densities for the random component are set as described in Tables 1 and 2. Only values of the power spectral densities at the break
Conclusions
The paper presents a control method for multi-axial sine on random mixed vibration test. An improved correlation integral method is used to separate sine on random mixed signals with high precision and computational efficiency. The non-Gaussian characteristics are considered in random vibration component.
First, the Gaussian random signals are generated by reference spectra including auto-power spectral densities, phase and coherence coefficients, and then the Hermite polynomial model
Funding
This work was supported by the China Scholarship Council. (No. 201806830049).
References (21)
- et al.
The effort of the dynamic simulation on the fatigue damage evaluation of flexible mechanical systems loaded by non-Gaussian and non-stationary loads
Int. J. Fatigue
(2017) - et al.
Non-Stationarity index in vibration fatigue: theoretical and experimental research
Int. J. Fatigue
(2017) - et al.
Synthesis of Sine-on-Random vibration profiles for accelerated life tests based on fatigue damage spectrum equivalence
Mech. Syst. Sig. Process.
(2018) - et al.
A minimum drives automatic target definition procedure for multi-axis random control testing
Mech. Syst. Sig. Process.
(2018) - et al.
Matrix Power Control algorithm for multi-input multi-output random vibration test
Chin. J. Aeronaut.
(2011) - et al.
Vibration fatigue analysis and multi-axial effect in testing of aerospace structures
Mech. Syst. Sig. Process.
(2009) - et al.
Multiaxial vibration fatigue–a theoretical and experimental comparison
Mech. Syst. Sig. Process.
(2016) - et al.
Control method for multi-input multi-output non-Gaussian random vibration test with cross spectra consideration
Chin. J. Aeronaut.
(2017) - et al.
Multi-exciter stationary non-Gaussian random vibration test with time domain randomization
Mech. Syst. Sig. Process..
(2019) - et al.
Vibration fatigue analysis of a cantilever beam using different fatigue theories
Cited by (29)
Non-random vibration analysis of rotate vector reducer
2023, Journal of Sound and VibrationControl strategy for multi-axial swept sine on random mixed vibration testing
2022, Journal of Sound and VibrationCitation Excerpt :Therefore, the SoR vibration signals are the superposition of the deterministic sine and broadband random. Zheng et al. [18] proposed an improved correlation integral method to separate the deterministic sine and random signals, and it is applied to the MIMO sine on random vibration control test. When the rotor is in the process of acceleration or deceleration, the frequencies of the rotating machinery are varying with time.
Power spectrum and kurtosis separation method for multi-shaker non-Gaussian random vibration control
2022, Mechanical Systems and Signal ProcessingCitation Excerpt :To obtain a reliable test result, the control apparatus should be capable of simulating multi-dimensional non-Gaussian random vibrations. Currently, there is an increasing interest in non-Gaussian vibration fatigue [9,10] and non-Gaussian vibration control [11,12]. The non-Gaussian random vibration test differs from the traditional Gaussian case because time-domain characteristics of responses should be controlled simultaneously with frequency-domain characteristics.
Multi-shaker half sine shock on random mixed vibration control
2021, Journal of Sound and VibrationCitation Excerpt :The missile needs intermittent ignition to achieve an orbital maneuver during the flight which is a typical shock on random vibration environment. Zheng et al. provided a control strategy for the multi-shaker sine on random vibration test in which an improved correlation integral method is employed [6]. A signal separator using the partial coherence technique is designed to decouple the mixed shock on random vibration by Zou et al [7].
On the statistical distribution of the maxima of sine on random process
2021, Mechanical Systems and Signal ProcessingSpectral analysis of sine-sweep vibration: A fatigue damage estimation method
2021, Mechanical Systems and Signal ProcessingCitation Excerpt :According to the category of the product, different qualification tests may be mandatory [1,2,10]. For example, sine-on-random or sine sweep test may be imposed [11,12]. The latter, mostly typical in automotive and helicopter sectors, foresees exciting the components with an harmonic signal with frequency content and amplitude varying according to predefined laws [13].