Souza, Colin Francis (2022) Improved methods for characterising acoustoplasticity. PhD thesis, University of Glasgow.
Full text available as:
PDF
Download (5MB) |
Abstract
The benefits of high-power ultrasonics to industrial metal forming processes have long been demonstrated in uniaxial mechanical tests. The astonishing reductions in flow stress observed have been linked to changes to surface friction and to an interaction of the excitation with the mechanisms of plastic deformation in metals. Many advanced techniques and material models have been brought to bear on the problem of the underlying physics of acoustoplasticity, and yet all rely fundamentally on accurate force and extension data. The effects of inertia and inhomogeneity in the loading distribution on the specimen have been largely ignored, and yet are incompatible with commonly used instrumentation.
This thesis reports investigations which address the error introduced into force measurement in mechanical testing by ultrasonic excitation. After reviewing experimental mechanics techniques, it was found that the piezoelectric force transducer retained its central role in defining true flow stress reduction. An inertia-based barrier to vibration was introduced between the force transducer and test machine crosshead, to impose the rigid boundary condition desired to ensure the force transducer coincided with a displacement node. Lumped-parameter modelling indicated that the dynamic response of the piezoelectric force transducer’s structure could significantly distort the amplitude of an oscillatory force measurand. Either amplification or attenuation could result depending on the proximity of excitation frequency to natural frequency of the force transducer’s first longitudinal mode. Simple impulse experiments provided the natural frequency of the force transducer in the free-free condition, a parameter used in later finite element (FE) modelling of the ultrasonic tensile test structure.
Experimental Modal Analysis (EMA) was used to investigate the dynamic response of the ultrasonic tensile test structure, and to map the mode shape of the first longitudinal mode, the mode utilised in ultrasonic tensile testing. A finite element model was constructed of the test apparatus, and subsequently solved in an eigenvalue analysis to extract the natural frequency and mode shape of the first longitudinal mode. When the numerically predicted waveform was compared with that found from EMA, a significant difference was discovered between the horn and specimen. The compliance of the joint was adjusted until the simulated mode shape converged on its experimental counterpart.
Once experimentally calibrated, the FE model was used to predict the force experienced by the force transducer for increasing values of vibration amplitude. Comparison with experimental force measurements found good agreement. Of greatest importance to the investigation of flow stress, the FE model predicted the indicated value from the force transducer to be 1.91 times greater than the measurand at the specimen-force transducer interface.
Strain gauges were attached to the gauge section of the specimen in the ultrasonic tensile test apparatus, and the vibration varied over a range of amplitudes. By converting the oscillatory strain measurement into force on the specimen cross-section, the loading experienced by the specimen at the strain gauge location was compared to force measurements made simultaneously by the piezoelectric force transducer. The ratio of force amplitude from the force transducer over the force amplitude calculated from the specimen strain measurement was found to vary from 3.13 to 3.50, with a mean of 3.32. Repeating the experiment within the FE model calculated an amplitude ratio of 3.33, constant over all vibration amplitudes. This value was used to develop a correction factor to extrapolate force on the specimen from piezoelectric force transducer measurement. The correction was applied to an ultrasonic tensile test on a soft aluminium. Though the mean stress was reduced during the periods of excitation, no real reduction in flow stress was observed, which is consistent with the theory of stress superposition.
The evolution of plastic deformation was studied over the gauge section of an ultrasonically excited specimen, using an optical metrology system adapted for use on the ultrasonic tensile test. To eliminate oscillatory motion from images, a high-speed strobe lit the specimen in bursts of light synchronised with the ultrasonic excitation. Digital Image Correlation was used to process the image sequence to find strain and strain rate across the whole face of the specimen gauge length. It was observed that the application of ultrasonic excitation disrupted the usual distribution of plastic deformation along the specimen length, focussing deformation towards the location of peak stress amplitude. Again, observations were consistent with the theory of stress superposition.
This thesis demonstrates how the dynamic response of the structure of the specimen and force transducer in an ultrasonic tensile test can significantly distort the force measurement, crucial for accurately identifying a real reduction in flow stress. This has implications for studies of acoustoplasticity aiming at determining underlying physical mechanisms. It is found that, when the effect of inertia is accounted for, the theory of stress superposition is sufficient to explain the stress-strain relationship observed.
Item Type: | Thesis (PhD) |
---|---|
Qualification Level: | Doctoral |
Subjects: | T Technology > T Technology (General) |
Colleges/Schools: | College of Science and Engineering > School of Engineering |
Funder's Name: | Engineering and Physical Sciences Research Council (EPSRC) |
Supervisor's Name: | Lucas, Professor Margaret and Mulvihill, Dr. Daniel |
Date of Award: | 2022 |
Depositing User: | Theses Team |
Unique ID: | glathesis:2022-82952 |
Copyright: | Copyright of this thesis is held by the author. |
Date Deposited: | 14 Jun 2022 13:20 |
Last Modified: | 14 Jun 2022 13:27 |
Thesis DOI: | 10.5525/gla.thesis.82952 |
URI: | https://theses.gla.ac.uk/id/eprint/82952 |
Actions (login required)
View Item |
Downloads
Downloads per month over past year