Search results

Defects can be caused by a number of factors, such as maintenance damage, ground handling and foreign objects thrown up from runways during an in-service use of composite aerospace structures. Sandwich structures are capable of absorbing large amounts of energy under impact loads, resulting in high structural crashworthiness. This situation is one of the many reasons why sandwich structures are extensively used in many aerospace applications nowadays. Their non-destructive inspection is often more complex. Hence, the choice of a suitable non-destructive testing (NDT) method can play a key role in successful damage detection. The paper aims to discuss these issues.

A comparison of detection capabilities of selected C-scan NDT methods applicable for inspections of sandwich structures was performed using water-squirt, air-coupled and pitch-catch (PC) ultrasonic techniques, supplemented by laser shearography (LS).

Test results showed that the water-squirt and PC techniques are the most suitable methods for core damage evaluation. Meanwhile, the air-coupled method showed lower sensitivity for the detection of several artificial defects and impact damage in honeycomb sandwiches when unfocussed transducers were used. LS can detect most of the defects in the panels, but it has lower sensitivity and resolution for honeycomb core-type sandwiches.

This study quantitatively compared the damage size indication capabilities of sandwich structures by using various NDT techniques. Results of the realised tests can be used for successful selection of a suitable NDT method. Combinations of the presented methods revealed most defects.

There are structural elements on the aircraft that may be exposed to high-intensity sound levels. One of them is an air inlet duct of the jet engine. To prepare data for the air duct damage tolerance analysis, flat panels were tested under acoustic loading. The paper aims to discuss this issue.

The acoustic fatigue test equipment for grazing wave’s incidence was designed based on the FE analyses. Flat composite panels were designed and manufactured using the Hexply 8552/AGP193-PW prepreg with the simulation of production imperfections or operational damage. The dynamic behaviour of panels has been tested using three regimes of acoustic loading: white noise spectrum, engine noise spectrum and discrete harmonic frequencies. The panel deflection was monitored along its longitudinal axis, and the ultrasonic NDT instruments were used for the monitoring of relevant delamination increments. The FE model of the panel was created in Abaqus to study panel dynamic characteristics.

No delamination progress was observed by NDT testing even if dynamic characteristics, especially modal frequency, of the panel changed during the fatigue test. Rayleigh damping coefficients were evaluated for their use in FE models. Significant differences were found between the measured and computed panel deflection curves near the edge of the panel.

The research results underscored the signification of the FE model boundary conditions and the element type selections when the panel works like a membrane rather than a plate because of their low bending stiffness.

This paper aims to Test the capabilities/accuracies of four deep learning pre trained convolutional neural network (CNN) models to detect and classify types of highway cracks, as well as developing a new CNN model to maximize the accuracy at different learning rates.

A sample of 4,663 images of highway cracks were collected and classified into three categories of cracks, namely, “vertical cracks,” “horizontal and vertical cracks” and “diagonal cracks,” subsequently, using “Matlab” to classify the sample to training (70%) and testing (30%) to apply the four deep learning CNN models and compute their accuracies. After that, developing a new deep learning CNN model to maximize the accuracy of detecting and classifying highway cracks and testing the accuracy using three optimization algorithms at different learning rates.

The accuracies result of the four deep learning pre-trained models are above the averages between top-1 and top-5 and the accuracy of classifying and detecting the samples exceeded the top-5 accuracy for the pre-trained AlexNet model around 3% and by 0.2% for the GoogleNet model. The accurate model here is the GoogleNet model as the accuracy is 89.08% and it is higher than AlexNet by 1.26%. While the computed accuracy for the new created deep learning CNN model exceeded all pre-trained models by achieving 97.62% at a learning rate of 0.001 using Adam’s optimization algorithm.

The created deep learning CNN model will enable users (e.g. highway agencies) to scan a long highway and detect types of cracks accurately in a very short time compared to traditional approaches.

A new deep learning CNN-based highway cracks detection was developed based on testing four pre-trained CNN models and analyze the capabilities of each model to maximize the accuracy of the proposed CNN.

The purpose of this paper is to present a study on the application of four damage factors to several single and multiple damage scenarios of aluminium beams. Each one of these damage factors is defined by the information given by modal curvatures of the beams.

The methodology consisted of a first experimental stage in which the modal rotations were measured with shearography and a subsequent numerical analysis in order to obtain the modal curvatures. To this end, three finite difference formulae were applied. The modal curvatures were then used to calculate the damage factors.

It was found that the profile of the damage factors varies according to the finite difference formula used. In view of the findings, the differences among the damage factors analysed are highlighted and some final recommendations to improve damage identifications via modal curvature-based are presented.

To the best of the authors’ knowledge, the application and comparison of several finite difference formulae and corresponding optimal sampling has not been carried out before. With the proposed approach, it is possible to identify multiple damages, which is still a great challenge. The post-processing of shearography measurements with a numerical method, which is inherently a multidisciplinary approach, is also a substantial improvement upon other type of approaches found in the literature.

The purpose of this paper is to provide an insight into the techniques used for the non‐destructive testing (NDT) of non‐metallic structural materials, notably polymer and ceramic composites.

Following a short introduction, this paper first considers methods for testing carbon fibre‐ and glass fibre‐reinforced polymer composites. It then discusses the role of NDT in wind and wave power systems and some of the techniques used to test ceramics and ceramic composites. Brief conclusions are drawn.

This shows that the growing use of non‐metallic engineering materials in critical applications has highlighted the need for a range of advanced NDT methods. While some traditional techniques can be adapted to test these materials, in several instances novel methods are required. These include a range of thermal, ultrasonic, electromagnetic, radiographic and laser‐based technologies.

The paper provides a review of the techniques used and being developed for the non‐destructive testing of non‐metallic engineering materials.

Large structural objects, primarily concrete bridges, can be reinforced by gluing to their stretched surface tapes of fiber-reinforced polymer (FRP). The condition for this technology to work requires the quality of the bonding of FRP and the concrete to be perfect. Possible defects may arise in the phase of construction but also as a result of long-term fatigue loads. These defects having different forms of voids and discontinuities in the bonding layer are difficult to detect by optical inspection. This paper aims to describe the development of a rapid and nondestructive method for quantitative assessment of the debonding between materials.

The applied technique belongs to the wide class of active infrared (IR) thermography, the principle of which is to heat (or cool) the investigated object, and determine the properties of interest from the recorded, by an IR camera, temperature field. The methodology implemented in this work is to uniformly heat for a few seconds, using a set of halogen lamps, the FRP surface attached to the concrete. The parameter of interest is the thermal resistance of the layer separating the polymer tape and the concrete. The presence of voids and debonding will result in large values of this resistance. Its value is retrieved by solving an inverse transient heat conduction problem. This is accomplished by minimizing, in the sense of least squares, the difference between the recorded and simulated temperatures. The latter is defined as a solution of a 1D transient heat conduction problem with the already mentioned thermal resistance treated as the only decision variable.

A general method has been developed, which detects debonding of the FRP tapes from the concrete. The method is rapid and nondestructive. Owing to a special selection of the compared dimensionless measured and simulated temperatures, the method is not sensitive to the surface quality (roughness and emissivity). Measurements and calculation may be executed within seconds. The efficiency of the technique has been shown at a sample, where the defects have been artificially introduced in a controlled manner.

A quantitative assessment procedure which can be used to determine the extent of the debonding has been developed. The procedure uses inverse technique whose result is the unknown thermal resistance between the member and the FRP strip.