Measuring high-frequency wave propagation in railroad tracks by joint time–frequency analysis

Scalea, Francesco Lanza di; McNamara, John

doi:10.1016/s0022-460x(03)00563-7

Cited by 46 publications

(19 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The most obvious examples are seen in composite materials where, for example, Kinra et al [17] have measured the dispersion curves for a layered periodic particulate composite using longitudinal wave transducers on specimens of two different lengths. High frequency wave propagation in railway tracks was also found to be dispersive [11], which is again driven by the microstructure if one considers the spacing between sleepers on a rail track to be the micro-scale (compared to the length of the track). Tyas and Watson [24,25] report wave dispersion in elastic bar impact experiments that occurs due to the wavelengths of propagation approaching the diameter of the bar, and thus radial inertial effects becoming significant.…”

mentioning

confidence: 95%

Finite element modelling of wave dispersion with dynamically consistent gradient elasticity

Bennett

Askes

2008

Comput Mech

View full text Add to dashboard Cite

The discretisation aspects are investigated of a recently developed gradient elasticity model that is capable of describing wave dispersion and microstructural effects. The model includes higher-order stiffness terms alongside higher-order inertia terms. To avoid the need for C 1 -continuous interpolations in a numerical implementation, the field equations are rewritten using a recently developed operator split. The equations are discretised in space using finite elements and discretised in time using the Newmark time integrator. Firstly, the critical time step is derived for use with conditionally stable time integrators. Next, the dispersion behaviour of the continuum is compared with the dispersion behaviour of the discretised medium, which allows the formulation of rules on how to select the element size and the time step size. These rules are then verified in a onedimensional bar example and in a two-dimensional example of wave propagation. It follows that the element size must be chosen roughly equal to the smallest of the intrinsic length scales, while the optimal time step follows from the ratio of element size and wave velocity and from the ratio of the length scales.

show abstract

mentioning

confidence: 95%

Finite element modelling of wave dispersion with dynamically consistent gradient elasticity

Bennett

Askes

2008

Comput Mech

View full text Add to dashboard Cite

show abstract

“…The guided wave showed a non-dispersive behaviour with the increase of the frequency as for the guided wave showed in (Lanza di Scalea and McNamara, 2004). Unfortunately, a direct comparison with the values found in literature (Lanza di Scalea and McNamara, 2004) could not be made, since the maximum frequency at which the guided wave velocity evaluated in literature (8000 Hz) is well below the minimum frequency investigated in this paper. In Fig.…”

Section: Guided Wave Velocity Evaluationmentioning

confidence: 51%

“…21, although the top frequency is well above the maximum meaningful frequency (83.5 kHz) evaluated in Eq. (17), the guided wave velocity behaviour resembles the physical conduct of experimental and numerically derived guided wave velocities (Cawley et al, 2002;Lanza di Scalea and McNamara, 2004).…”

Section: Guided Wave Velocity Evaluationmentioning

confidence: 83%

See 1 more Smart Citation

A new damage detection technique based on wave propagation for rails

Zumpano

Meo

2006

International Journal of Solids and Structures

View full text Add to dashboard Cite

This paper presents a novel damage detection technique, tailored at the identification of structural surface damage on rail structures. The damage detection, proposed in this paper, exploits the wave propagation phenomena (P, S, Rayleigh and guided wave velocities) by identifying discrepancies, due to damage presence, in the dynamic behaviour of the structure. The uncorrelations are generated by waves reflected back to the sensor locations by the flaw surfaces. The peculiarity of the presented approach is the use of a time frequency coherence function for the identification of the arrivals of guided wave reflected back to the sensors by the damage surfaces.The damage detection methodology developed was divided into three steps. In the first step, the presence of the damage on the structure was assessed. In the second step, the arrival time of the reflected wave (or echo) was estimated using the continuous wavelet transform. Then, the detection algorithm was able, through a ray-tracing algorithm, to estimate the location of damage.A numerical investigation of two single damages was carried out. The damage was introduced on the railhead surface to simulate rolling contact fatigue defects. The results showed that the proposed methodology can be used successfully to localise the damage location, however, as expected, the localisation is strongly affected by the frequency range used. The results suggested that the separation and the characterisation of single modes are crucial for the identification of different types of rail defects. Further work is needed to establish the damage severity by relating the magnitude of the changes of the time frequency coherence to reflection and attenuation coefficients of each guided wave used and on the selection of the best range of frequency according to the type of damage to be identified.

show abstract

“…Such a remark naturally leads to use a time-frequency or a time-scale analysis. These methods are largely used in the railway transports field to detect the rails defects [14], [15] and also in the industrial sectors to perform a diagnosis in the revolving machines and the internal combustion engine [16]. In the present study, one hopes to detect defects which range from ten centimeters down to a millimeter, that is over 2 orders of magnitude.…”

Section: A Acquisition Systemmentioning

confidence: 97%

Combining learning methods and time-scale analysis for defect diagnosis of a tramway guiding system

Mamar¹,

Chainais²,

Aussem

2008

2008 16th Mediterranean Conference on Control and Automation

View full text Add to dashboard Cite

This paper presents a diagnosis system for detecting tramway rollers defects. First, the continuous wavelet transform is applied on vibration signals measured by specific accelerometers. Then, the Singular Values Decomposition (SVD) is applied on the time-scale representations to extract a set of singular values as classification features. The resulting multi-class classification problem is decomposed into several 2-class sub-problems. The predicted probabilities are coupled using a pairwise coupling method. Empirical results demonstrate the efficiency and robustness of the overall diagnosis system on measurement data.

show abstract

Measuring high-frequency wave propagation in railroad tracks by joint time–frequency analysis

Cited by 46 publications

References 18 publications

Finite element modelling of wave dispersion with dynamically consistent gradient elasticity

Finite element modelling of wave dispersion with dynamically consistent gradient elasticity

A new damage detection technique based on wave propagation for rails

Combining learning methods and time-scale analysis for defect diagnosis of a tramway guiding system

Contact Info

Product

Resources

About