A.J. Peyton

Development of multiple frequency electromagnetic induction systems for steel flow visualization

This paper presents recent developments in the use of electromagnetic induction tomography (EMT) for steel flow visualization. Several aspects are reported. First, results are shown from an 8-coil, single-frequency, EMT system from tests using liquid steel. The results are consistent with video recordings of an exposed section of the steel flow passing through a submerged entry nozzle, in terms of flow size and position, providing a good representation of the steel flow profile changes during trials. The second part describes the development of a system with a C-shaped sensor, which is capable of being slotted in place for practical deployment as well as being rapidly removed during nozzle changes. The effects of reducing the number of coils in this configuration were also studied. Finally, the development of a multiple-frequency system for plant use is reported. The system is designed based on a commercial data acquisition board, which can provide three sinusoidal signals with target frequencies for excitation simultaneously. This paper describes the new hardware electronics and software. Experimental results show that the system is able to identify a variety of test samples. Instead of imaging the cross-section of the steel flow profiles, the current system is developed for checking signal levels at different operation frequencies, which are of more interest for industrial use. Nevertheless, the work demonstrates a significant step forward to develop a multiple-frequency EMT system for practical use in this industrial process application.

Liquid metal experiments with swirling flow submerged entry nozzle

Abstract The influence of a swirling flow inside the submerged entry nozzle on the structure and the stability of a liquid metal flow in a physical model of a slab casting mould are investigated. For visualisation of the flow, contactless inductive flow tomography (CIFT) is applied. As expected and desired, the swirling flow leads to a stronger upward fluid motion along the walls. At the same time, however, the oscillatory character of the flow becomes stronger. These flow features obtained with CIFT are shown to be in reasonable agreement with independent measurements using ultrasonic Doppler velocimetry. Preliminary results of numerical simulations also show a similar behaviour.

Theoretical and numerical approaches to the forward problem and sensitivity calculation of a novel contactless inductive flow tomography (CIFT)

A completely contactless flow measurement technique based on the principle of EM induction measurements?contactless inductive flow tomography (CIFT)?has been previously reported by a team based at Forschungszentrum Dresden-Rossendorf (FZD). This technique is suited to the measurement of velocity fields in high conductivity liquids, and the possible applications range from monitoring metal casting and silicon crystal growth in industry to gaining insights into the working of the geodynamo. The forward problem, i.e. calculating the induced magnetic field from a known velocity profile, can be described as a linear relationship when the magnetic Reynolds number is small. Previously, an integral equation method was used to formulate the forward problem; however, although the sensitivity matrices were calculated, they were not explicitly expressed and computation involved the solution of an ill-conditioned system of equations using a so-called deflation method. In this paper, we present the derivation of the sensitivity matrix directly from electromagnetic field theory and the results are expressed very concisely as the cross product of two field vectors. A numerical method based on a finite difference method has also been developed to verify the formulation. It is believed that this approach provides a simple yet fast route to the forward solution of CIFT. Furthermore, a method for sensor design selection based on eigenvalue analysis is presented.

Monitoring microstructure changes in rod online by using induction spectroscopy

Abstract This paper describes the application of a new multifrequency sensor system to an industrial rod mill. The system operates by analysing the response of the rod to an excitation signal which contains a range of frequencies. The sensor system is able to successfully monitor the transformation of the material. Sensor configurations that could be deployed online have been considered and a prototype system has been built and tested in the Stelmor section of the Scunthorpe Rod Mill fitted in the enhanced cooling zone. Initial results are shown, demonstrating that the phase spectra is essentially unaffected by variations in lift-off and the lay pattern of the rod on the run-out table. The system was also able to detect different phase spectra from different grades of steel and also to detect features of the process which are known to affect microstructure.

Non-Contact Characterisation of Carbon-Fibre-Reinforced Plastics (CFRP) Using Multi-frequency Eddy Current Sensors

The characterization of carbon-fiber-reinforced plastics (CFRP) using multi-frequency eddy current sensors are presented in this paper. Three sensors are designed for bulk conductivity measurements, directionality characterization, fault detection and imaging of CFRP samples. Models describing the interaction of the sensors with the CFRP plate samples are developed to provide explanation and physical insights. A signal processing method is developed to compensate for the variation in lift-off during the measurements.

Impedance spectroscopy for remote analysis of steel microstructures

Abstract The current state of impedance spectroscopy as an online technique in hot strip mills to determine the progress of phase transformations in steels is reviewed. Electromagnetic sensors have the advantage of being non-contact probes and the use of multi-frequency techniques allows the variation of phase composition with depth to be profiled.

Measurement of electromagnetic properties of power station steels

Accurate measurement of electromagnetic properties of steels is of significant importance as EM properties are indicative of microstructure and hence properties of the material. In this paper, we present the measurement of cylindrical power station steel samples in different microstructural states representative of their initial condition and after service exposure. Cylindrical air cored sensors were used. Analytical and numerical methods (Finite Elements Methods) were employed to calculate the sensor response of these samples. Experimental results were obtained for a range of samples and their electromagnetic properties inferred by fitting finite element models to the measured results. In addition, sensitivity and error analysis were carried out to evaluate the accuracy of the method.

Imaging Internal Structure with Electromagnetic Induction Tomography

A particularly difficult case in electromagnetic induction tomography (EMT) is to image internal structure at the centre of the object space when the material at the outer regions is conductive. This is because the outer material acts as an electromagnetic screen, which partially excludes the magnetic field from the interior space and hence reduces the sensitivity at the centre. In this paper, we propose a methodology to image the conductivity distribution of an annular object when internal conductive objects are present. Finite element simulations were carried out to investigate how sensor coil outputs change with factors such as the size of internal object with regards to the external one. Linear and non-linear image reconstruction methods are applied to the tomographic data that are collected from a newly developed EMT system and image results are given in the paper

Electromagnetic visualisation of steel flow in continuous casting nozzles

Abstract Control of steel flow to the mould during continuous casting is critical to ensure steel cleanness and surface quality. Existing methods of assessing steel flow in the submerged entry nozzle (SEN) and mould are based on mathematical and physical modelling; there is no means of determining actual conditions within the SEN, which is fundamentally an opaque tube. Difficulties are experienced when studying significant process variations such as stopper disturbances, air entrainment, asymmetry of flow and clogging in the tundish nozzle. Work to design and test an electromagnetic technique to visualise the steel flow pattern in the tundish pouring nozzle is reported. Initial development work has been extended by pilot plant and production plant trials at Corus.

Modelling and experimental study of magnetic induction spectroscopy for rail decarburisation measurement

This paper presents a modeling and experimental study, based on a magnetic induction spectroscopy (MIS) technique, for non-destructive evaluation of rail decarburization. The inductive sensor contained an H-shaped ferrite core, which was excited with a multi-frequency waveform over the range of approximately 1 kHz to 100 kHz. Finite-element (FE) simulation was carried out to understand the link between EM sensor output and the level of decarburization. Rail samples with different levels of decarburisation, due to different bloom reheat times (85 min-412 min) were tested in the lab. It was found that the zero-crossing frequency in the MIS response is linearly proportional to the decarburisation level of the rails by FE simulation, theoretic analysis and experiment. This finding is helpful in understanding the response of the EM sensor to rail decarburisation and may lead to a non-contact, non destructive method for use during rail manufacturing.