Skip to main content

About this book

​Rotating Machinery, Optical Methods & Scanning LDV Methods, Volume 6: Proceedings of the 39th IMAC, A Conference and Exposition on Structural Dynamics, 2021, the sixth volume of nine from the Conference brings together contributions to this important area of research and engineering. The collection presents early findings and case studies on fundamental and applied aspects of Structural Health Monitoring, including papers on:

Novel Techniques

Optical Methods,

Scanning LDV Methods

Photogrammetry & DIC

Rotating Machinery

Table of Contents


Chapter 1. WaveAR: A Real-Time Sensor-Based Augmented Reality Implementation for Operating Deflection Shapes

We present WaveAR, the first sensor-based augmented reality system for structural dynamic measurements of operating deflection shapes. WaveAR offers a significant simplification of the configuration effort of vibration sensor measurements by an automatically sensor tracking coupled by an efficient 3D scan generation of the structure. AR markers are used for sensor tracking which are placed on the sensor. This allows the determination of the sensor pose relative to the scanned structure. For this purpose the vibration sensors must be covered with AR markers. This allows a universal easy application also for existing sensor equipment. For 3D scanning and sensor tracking, a low-cost stereo depth camera is used. Because the AR application should be independent from data acquisition hardware, we have implemented a universal HTTP Rest-API interface for data acquisition connection. Through this, it is also supported to use wireless sensors. Based on this configuration, a real-time visualization in the form of an augmented reality application of the operating deflection shapes can be performed under real operating conditions. This enables the user to view the measurement from different perspectives by moving the camera around the object to be examined.
Daniel Herfert, Kai Henning

Chapter 2. Full-Field 3D Mode Shape Measurement Using the Multiview Spectral Optical Flow Imaging Method

In some modal testing applications, image-based techniques offer compelling advantages compared to conventional sensors, particularly in cases where mass loading is problematic or a high spatial resolution is required. Still, the limited field of view of the well-established stereo DIC method can be a limiting factor in some measurements. By employing the principles of multi-view imaging, this constraint can be alleviated. For linear, time-invariant structures, multi-view triangulation can be performed in the frequency domain. The measurement field of view can in this way be arbitrarily extended using only a single moving camera imaging system. By using the simplified optical flow method in the displacement spectra identification step, a still-frame camera can be used for image acquisition, considerably lowering the complexity and cost of the imaging system. In this work, the spectral optical flow imaging and frequency-domain multi-view triangulation approaches are combined in an effort to identify mode shapes of a simple three-dimensional structure. The aim is to develop a robust and cost-effective single still-frame camera-based modal testing method.
Domen Gorjup, Janko Slavič, Miha Boltežar

Chapter 3. Stereophotogrammetry Camera Pose Optimization

Stereophotogrammetry makes use of calibrated camera pairs to obtain three-dimensional information from two-dimensional images. The accuracy of the extracted measurements is extremely dependent on the selection and setup of the camera system. For a given test object and desired viewing orientation, there is no one “correct” stereo camera setup, but rather a range of potential setups with some approaching an optimal system with respect to maximizing the measurement resolution. The open-ended nature of this test design exercise is compounded by equipment availability and the fact that many of the setup parameters have dependent characteristics, e.g., changing focal distance will affect stand-off distance, field of view, and image projection, among others. This work describes a planning tool that utilizes projective and Euclidian geometry to iteratively estimate optimal camera poses for available equipment, determines the most efficient image size, and also performs checks for lens diffraction, minimum focal distance, and adequate depth of field. Integrating a finite element model with these calculations further extends planning capabilities by allowing (1) an accurate definition of the volume to be imaged and (2) the ability to estimate response displacements in pixels due to an arbitrary excitation applied to the test object. This latter capability is critical for pre-test determination of the chosen camera setup’s ability to successfully extract three-dimensional measurements. The theory and workflow are presented along with an experimental demonstration.
Bryan L. Witt, J. Justin Wilbanks, Brian C. Owens, Daniel P. Rohe

Chapter 4. Simplified Finite Element Models of Pyramidal Truss Sandwich Panels with Welded Joints for Dynamic Analysis and Their Experimental Validation

Pyramidal truss sandwich panels (PTSPs) are widely used in engineering structures, and their face sheets and core parts are generally bonded by the welding process. A large number of solid elements are usually required in the finite element model of a PTSP with welded joints to obtain its accurate modal parameters. Ignoring welded joints of the PTSP can save many degrees of freedom (DOFs) but significantly change its natural frequencies. This study aims to accurately determine modal parameters of a PTSP with welded joints with much fewer DOFs than its solid element model and to obtain its operational modal analysis results by avoiding missing its modes. Two novel methods that consider welded joints as equivalent stiffness are proposed to create beam-shell element models of the PTSP. The main step is to match stiffnesses of beam and shell elements of a welded joint with those of its solid elements. Compared with the solid element model of the PTSP, its proposed models provide almost the same levels of accuracy for natural frequencies and mode shapes for the first 20 elastic modes while reducing DOFs by about 98% for the whole structure and 99% for each welded joint. The first 14 elastic modes of a PTSP specimen that were measured without missing any modes by synchronously capturing its two-faced vibrations through use of a three-dimensional scanning laser vibrometer and a mirror experimentally validate its beam-shell element models created by the two proposed methods.
Ke Yuan, Weidong Zhu

Chapter 5. Operational Modal Analysis of Rotating Structures Under Ambient Excitation Using Tracking Continuously Scanning Laser Doppler Vibrometry

A continuously scanning laser Doppler vibrometer (CSLDV) system is capable of rapidly obtaining spatially dense vibration measurement by continuously sweeping its laser spot along a path on a structure surface. This paper presents a new operational modal analysis (OMA) method for a rotating structure based on a rigorous rotating beam vibration theory, an image processing method, and a data processing method called the lifting method. A novel tracking CSLDV (TCSLDV) system was developed in this work to track and scan a rotating structure, and the real-time position of the rotating structure can be determined by image processing so that the TCSLDV system is capable of tracking a time-varying scan path on the rotating structure. The lifting method can transform raw TCSLDV measurement into measurements at multiple virtual measurement points as if they were measured by transducers attached to these measurement points. Modal parameters of the rotating structure with a constant speed, including damped natural frequencies, undamped mode shapes, and modal damping ratios, and operating deflection shapes (ODSs) of the structure with a constant or prescribed time-varying rotation speed can be determined by calculating and analyzing correlation functions with non-negative time delays among measurements at virtual measurement points. Experimental investigation is conducted using the TCSLDV system to study the OMA method with which modal parameters and an ODS of a rotating fan blade with different constant speeds, as well as an ODS of the rotating fan blade with a non-constant speed, are successfully estimated.
L. F. Lyu, W. D. Zhu

Chapter 6. Delamination Detection in Fiber Metal Laminates Using Ultrasonic Wavefield Imaging

This work presents a novel, non-destructive evaluation (NDE) method for detecting delaminations in fiber metal laminate (FML) plate-like structures. FMLs are rapidly replacing other materials in many aerospace applications because of their superior mechanical properties, including improved tolerance to fatigue, corrosion, and impact damage. However, delaminations can occur deep in the plate, and since access is limited to the composite face during most operations, the ability of traditional NDE techniques to discern these defects is limited. Many researchers have proposed using ultrasonic guided waves to image defects, but the anisotropic nature of wave propagation in FMLs and the subtlety of defects between metal and fiber-reinforced composite layers necessitate a new approach. In contrast to repeated transient excitations proposed in other literature, the method proposed here utilizes the full-field, steady-state response of an FML plate to ultrasonic excitation. Thus, the inspection time is shortened as the delay between measurements is removed, and a higher-energy input improves the signal-to-noise ratio. A 2D scanning laser Doppler vibrometer (LDV) is used to record the measurements at discrete points, while a piezoelectric transducer supplies the ultrasonic excitation. The steady-state response is processed to visualize defects on a pixel-by-pixel basis and locate potential regions of delamination in the FML plate. In this study, the one-dimensional response of a plate-like T800 graphite composite Ti-6Al-4V FML specimen with known areas of delamination is simulated. Two defect-detection features, based on simulated physical phenomena—detrended Hilbert envelope magnitude (DHEM) and low-pass local phase derivative (LLPD)—are subsequently evaluated over a wide range of excitation frequencies, to determine an optimal input for increased precision. Results from these simulations suggest potential guidelines to achieve a rapid and reliable NDE method for delamination detection in FML structures.
Casey Gardner, Young Ko, Michael Koutoumbas, Eric Flynn, Ian Cummings, Phil Cornwell

Chapter 7. One-Dimensional Convolutional Neural Networks for Real-Time Damage Detection of Rotating Machinery

This paper presents a novel real-time rotating machinery damage monitoring system. The system detects, quantifies, and localizes damage in ball bearings in a fast and accurate way using one-dimensional convolutional neural networks (1D-CNNs). The proposed method has been validated with experimental work not only for single damage but also for multiple damage cases introduced onto ball bearings in laboratory environment. The two 1D-CNNs (one set for the interior bearing ring and another set for the exterior bearing ring) were trained and tested under the same conditions for torque and speed. It is observed that the proposed system showed excellent performance even with the severe additive noise. The proposed method can be implemented in practical use for online defect detection, monitoring, and condition assessment of ball bearings and other rotatory machine elements.
Onur Avci, Osama Abdeljaber, Serkan Kiranyaz, Sadok Sassi, Abdelrahman Ibrahim, Moncef Gabbouj

Chapter 8. A Practical Guide to Motion Magnification

Numerous algorithms and software packages have been developed to detect and magnify small motions in video recordings with applications to structural vibration analysis. While some techniques follow naturally from intuition or mathematical duality, others seem overly complex at face value. Due to this, it can be difficult to assess the best tool for analysis, especially in terms of an intuitive understanding of how the results came to be. To facilitate the development of this intuition, input videos with different test articles and vibration characteristics were processed through selected magnification schemes. For each input, the outputs of the different methods were compared in order to assess their differences, amplification potentials, and limitations. Beginning with single regions of interest, the investigation then moved into more complicated scenarios involving multiple areas of localized motion. Variations such as coherence and scale for objects in the frame were also examined. Unlike previous review papers, these magnification tools are introduced in a framework that is more geared toward the working acoustician; in particular, intensity-, Fourier-, multi-resolution-, and Hilbert-based methods are evaluated in a structural-acoustic context to garner a better understanding of the underlying principles and idiosyncrasies. The main result is a thorough evaluation of each method’s performance in order to develop a guide for choosing the best algorithm in a given scenario.
Sean Collier, Tyler Dare

Chapter 9. Squeeze Film Damper Experimental and Numerical Correlation: Test Setup Description and Parameter Identification of Dry System

The process of characterizing a damper involves both experimental techniques and model simulation. Experiments characterize dynamic behavior of a system through empirical modal analysis using impact and shaker testing. The layout of the test setup is crucial for a low-mass, nonlinear system. Physical arrangement, orientation of the system, and instrumentation affect the results that are discussed in this paper. These results can be used to validate a numerical simulation of the same system. With a validated model, important physical input parameters and modeling techniques can be determined and implemented in future designs. In other words, this modeling can be used with confidence to design alternative dampers that may be difficult, costly, and time-consuming to characterize experimentally.
Jason Cook, Jay Basinger, Thomas Hazelwood, Claire Luttrell, Blake Van Hoy, Adolfo Delgado

Chapter 10. Full-Field Modal Analysis by Using Digital Image Correlation Technique

At the state of art, the design validation of mechanical products in automotive, aerospace, and other advanced manufacturing industries is made by using vibration measurements and experimental modal analysis. The structures’ dynamic behavior is investigated mostly using contact transducers (accelerometers, strain gauges, load cells, …). Optical methods such as Digital Image Correlation (DIC) have recently received special attention in the structural dynamics field because it can be used to obtain full-field measurements. However, DIC can only measure vibrations in the camera field of view, which causes problems for hidden parts or in case the structure is too large to fit into the field of view. Therefore, it is of paramount importance to combine DIC with other traditional sensors such as accelerometers and strain gauges. Complex 3D test articles and structures need to be measured from many different directions. The knowledge gained by developing a heterogeneous sensor measurement system can significantly reduce the time and cost associated with pointwise sensor instrumentation. In this paper, DIC technique is used to get the full-field displacement. The sensor merging, DIC + pointwise sensors, is investigated. This information is then used to derive the modal characteristic of the structure (e.g. natural frequencies, damping ratios, and full-field mode shapes) that can be used for a more reliable FE model validation.
Davide Mastrodicasa, Emilio Di Lorenzo, Simone Manzato, Bart Peeters, Patrick Guillaume

Chapter 11. Validating Complex Models Accurately and Without Contact Using Scanning Laser Doppler Vibrometry (SLDV)

The need to validate simulation models of complex mechanical structures has grown in importance for efficiency in the design process. This is especially true for non-linear structures (such as composite panels and jointed components) where it is critical to use an accurate full-field measurement method. This tutorial covers a how-to-guide and use cases of scanning laser Doppler vibrometry (SLDV) as a non-invasive technology to efficiently characterize critical mechanical structures.
Jerome Eichenberger, Joerg Sauer

Chapter 12. Effect of Different Test Setup Configurations on the Identification of Modal Parameters from Digital Image Correlation

Over the past few years, the application fields for digital image correlation (DIC) have significantly broadened and are now extensively used not only for material characterization and deformation analysis but also for vibration and dynamic testing. The key advantage over more traditional technologies such as accelerometer- or laser-based measurement is the fact that all points of interest can be measured at once or over a limited number of repetitions without mass loading the structure or having to deal with complex cabling scenarios. While qualitative analysis, such as operational deflection shapes, proved to be very successful and straightforward, performing classical modal analysis on frequency response functions between a load cell signal and the deformations obtained by applying digital image correlation still requires extra care when setting up the experiments. In this paper, the effect of different cameras, camera settings, excitation levels, and speckle pattern on the modal parameters of a simple plate structure will be investigated and compared with classical acceleration measurement. The results will also be compared with numerical simulations where the accuracy of the mode shapes can be easily and quantitatively assessed.
L. Marchetti, D. Mastrodicasa, E. Di Lorenzo, S. Manzato, L. Bregant, B. Peeters, P. Lava

Chapter 13. WaveImage – Order ODS for Rotating Machineries

Operational deflection shape (ODS) analysis is a method to determine the vibration pattern of a structure influenced by its own operating forces and can be applied to the time and frequency domain. For rotating machinery with rapidly changing numbers of revolutions, ODS analysis provides insufficient results in the frequency domain. This work introduces the novel order ODS from the software WaveImage. This approach applies order analysis (OA) techniques to the ODS analysis.
Classical OA deals with the analysis of structures which contain rotating elements such as engines or turbines. Each rotating element generates its own noise and vibration pattern which contributes to the overall vibration pattern. With OA these individual patterns can be identified and analyzed for each rotating element. During a run-up or run-down, relevant frequencies of the structure shift in relation to the current rotational speed and are therefore more difficult to evaluate. Furthermore, in OA each measuring channel is evaluated individually. The visualization of a complex geometry together with the simultaneous measurement of several measuring points is currently not applied in the field of OA.
To overcome these disadvantages, a new domain was introduced to the ODS analysis. The advantage is the possibility of visualizing the ODS for a certain rotational speed. In OA, digital resampling takes place in which the original time signal is transformed into the angular range. After that the result is transferred to the target domain by Fourier transformation. To carry out the order ODS, the RPM signal must be measured in addition to the response measurement.
For the validation of the order ODS approach, a structure was measured using acceleration sensors and a slow-motion video sequence. The acceleration data was used to perform a frequency ODS. To calculate the order ODS, the optical flow is used to determine the acceleration data from the recorded video. Finally the result of the frequency ODS was compared to the order ODS.
Matthias Urban, Daniel Herfert, Maik Gollnick

Chapter 14. Multi-Level Damage Detection Using Octree Partitioning Algorithm

This research proposes a damage detection methodology that takes advantages of full-field sensing of structural elements via 3D digital image correlation (3D-DIC) within a new optimization algorithm. To get a unique solution for a finite element model updating problem in detecting internal properties of a structure by topology optimization method, a novel strategy is proposed termed as octree partitioning algorithm (OPA). Damage detection is attained using minimizing a defined objective function of residuals between the observed and corresponding mimicked responses, resulting in the identification of the underlying material distribution. This brief paper explores this concept through a case study on a series of small structural test specimens with fabricated damage on the specimens. The results show that the proposed OPA has capability of tuning the initial model to find the ideal partition configuration for the proposed model updating process.
Mehrdad S. Dizaji, Zhu Mao

Chapter 15. Photogrammetry-Based Experimental Modal Analysis for Plate Structures

A photogrammetry-based vibration measurement method was first developed and applied to beam structures for modal parameter estimation and structural damage identification. In this work, the photogrammetry-based method is experimentally investigated for its application to modal parameter estimation of plate structures. A single-point laser equipped with a diffractive optical element projected a grid consisting of 100 laser spots to a plate structure in a fully non-contact, non-destructive manner. Deformations of the plate were measured by tracking the projected spots that served as external optical features to be tracked in the photogrammetry-based method. A scanning laser Doppler vibrometer was also used to measure deformations of the plate for comparison purposes. The accuracy of the photogrammetry-based method and its three-dimensional kinematic relation for plate structures were studied by comparing its measurements and modal parameter estimation results with those by the scanning laser Doppler vibrometer. The three-dimensional kinematic relation between a point on a field-of-view plane of the plate and that on an image plane of a camera was experimentally verified. Both rigid body and elastic modes of the plate were identified, and their modal parameters were estimated. Mode shapes of the plate estimated using the photogrammetry-based method were compared with those from a finite element model, and their modal assurance criterion values were all over 90%, which verified the accuracy of the EMA method.
J. S. Kim, Y. F. Xu

Chapter 16. An Optical Mode Shape-Based Damage Detection Using Convolutional Neural Networks

One of the most widely used methods for structural health monitoring (SHM) is vibration-based, in which changes of the dynamics of a structure are collected and associated to damages. While the change of the natural frequencies may be influenced by many different factors, mode shapes provide a more informative feature set in identifying damages. However, most of the sensing hardware, such as accelerometers, and even some of the noncontact techniques, such as laser Doppler vibrometer, are only able to take measurements from specific locations, leading to a limited spatial resolution of deformation field. The use of camera-based sensing allows the extraction of information at much denser locations on a structure. Among the current techniques for motion sensing, phase-based motion estimation (PME) and phase-based motion magnification (PMM) aim to extract and amplify subtle motions, allowing the extraction of full-field mode shapes with an enhanced visibility. Convolutional neural network (CNN) is applied to metamodel the amplified mode shapes extracted from both undamaged baseline and damaged conditions. A lab-scale testbed is employed to provide data to validate the damage identification approach via the CNN algorithm under different operational conditions.
Celso T. do Cabo, Zhu Mao

Chapter 17. Full-Field 3D Experimental Modal Analysis from Dynamic Point Clouds Measured Using a Time-of-Flight Imager

The ability to measure static, high-resolution 3D point cloud data has existed for multiple decades and has been used to great benefit in both civil and mechanical engineering applications. Recently, time-of-flight imagers have emerged that are capable of measuring 3D dynamic point clouds at rates as high as 30 point cloud captures per second with resolutions approaching the millimeter scale. Conventional modal analysis utilizes contact measurements that are captured in the Lagrangian (i.e., material) coordinate system. Imager measurements such as used for DIC are captured in what is approximately an Eulerian frame of reference. However, oftentimes the imager measurements are captured in a small-motion, sub-pixel regime and can be assumed to be captured in a Lagrangian reference frame. As a result, most experimental modal identification algorithms are designed to operate on data captured in a Lagrangian reference frame. Measurements of 3D point clouds of vibrating structures do not necessarily fit into either an Eulerian or Lagrangian framework, particularly in the case where motion of the structure is large. An additional feature of these measurements is that the number of points measured on the structure can change significantly through time as a result of occlusions, the change in angle of the structure, or simply noise in the measurement. This feature of point clouds is significantly different from imagers and contact sensors in which the dimensionality of the measurements does not change through time. In this work we present the first known technique for extracting structural dynamics information from dynamic point clouds.
Moisés Silva, Andre Green, John Morales, Peter Meyerhofer, Yongchao Yang, Eloi Figueiredo, David Mascareñas

Chapter 18. Application of a U-Net Convolutional Neural Network to Ultrasonic Wavefield Measurements for Defect Characterization

Recent advances in nondestructive evaluation (NDE) techniques have sought to improve the testing speed and accuracy of automatic flaw detection algorithms to minimize the costly downtime of removing in-service parts and components for testing and maintenance. Acoustic wavenumber spectroscopy (AWS) is a rapid NDE technique that utilizes steady-state ultrasonic excitation and laser Doppler vibrometer (LDV) measurements to identify component flaws orders of magnitude faster than traditional time-of-flight ultrasonic NDE techniques. However, current AWS technology is limited when applied to larger domains, such as rooms and larger structures, due to increased processing needs, and it is limited in accuracy and spatial resolution when applied to smaller defects on the order of one wavelength in size as well as defects on the edges of the structure. This paper presents the novel application of a U-Net style convolutional neural network (CNN) to improve the processing speed and spatial resolution of current AWS technology by performing semantic segmentation on simulated ultrasonic wavefield images of a steady-state, single-tone excitation of an aluminum plate. The well-adopted ResNet architecture, which was pre-trained on the large and openly available ImageNet dataset, was trained by transfer learning on the augmented wavefield dataset for the purpose of defect localization and characterization in aluminum plates. Finally, the performance of the CNN processing time and spatial resolution accuracy were shown to improve upon the processing methods of current AWS technology.
Joshua D. Eckels, Isabel F. Fernandez, Kelly Ho, Nikolaos Dervilis, Erica M. Jacobson, Adam J. Wachtor

Chapter 19. Application of the RASTAR Method to Continuous Scanning LDV Measurements

This paper presents the application of RASTAR method to diagnostics performed by continuous scanning spectral sidebands. In a previous investigation, RASTAR proved to detect damage condition of a cantilever beam which was continuously scanned along a line by a laser beam. The objective of this paper is to extend the method to two-dimensional scan rather than one. As the method is based on spectral sidebands, the access to spectral lines for deflections in both directions in space should enhance the detection of damage for structures which are lookalike panels. The research is focused on numerical simulations where the LDV was synthesized using continuous scanning approach. The results are encouraging and show that two-dimensional scans can be more useful for RASTAR to detect damage.
D. Di Maio, S. Bruinsma
Additional information

Premium Partners

    Image Credits