In cold rolling the interface between the roll and strip, known as the roll bite, is key to the finish and geometry of rolled products. This interface is complex and operates in the mixed regime with part asperity contact and part fluid film. The factors which affect this are of interest to the cold rolling community. The inaccessibility of the interface has made direct measurement of its condition difficult. In this thesis ultrasonic approaches have been developed to measure the state of the roll bite, in-situ and in real time. The thesis starts by introducing the background theory underpinning modern cold rolling. The theory behind the proposed ultrasonic measurements is developed. It is proposed that the proportion of longitudinal and shear waves reflected by the asperity and lubricant film components of the mixed mode interface are dependent on their respective stiffnesses. From these the lubricant layer thickness can be calculated. It is also shown how time-of-flight measurements can be used to measure strip thickness, roll stress, roll deflection and roll material properties. Ultrasonic sensor systems were incorporated into pilot and semi-industrial cold-rolling mills. Measurements were taken as steel was rolled under a range of lubrication conditions, with rolling velocities from 25 m/min to 1200 m/min and elongations from 5% to 50%. Stiffness and lubricant film thickness measurements were demonstrated for films from 0.3 μm to ~6 μm. Roll bite contact lengths of 9.13mm to 15.34mm were recorded for elongations from 9.7% to 40%. For these same elongations average radial roll stresses of 180 MPa to 340 MPa and roll deflections of 30 μm to 55 μm at the roll bite centre were measured. Time-of-flight measurements yielded thickness profiles of strip reduced from 2.8mm to 2.154mm. Validation was provided by multiple numerical models which showed good agreement with the ultrasonic results.

No description available.

Metrology is integral to all manufacturing operations and component generation, as it is used to prove conformance to design specifications. An inspection process that is not designed or executed properly can lead to delayed deliveries and costly repairs. The combination of Mixed Reality (MR) and metrology could lead to immersive metrology (IM), which has the potential to radically change how part inspection is undertaken. By doing so, it may enhance the value adding capability of being able to dynamically inspect a part, in situ. The aim of this research is to explore the application of MR in metrology-enabled assembly for increasing the availability of knowledge at the time of assembly and inspection. This research presents a case for IM within the context of assembly inspection. A system termed Immersive System for Large Volume Metrology (ISLVM) is proposed to enable the integration of the critical elements of MR and metrology for dimensional inspection of large volume assemblies. ISLVM was tested with a participant based study in which 72 volunteers with no previous experience with metrology hardware and software were guided through an inspection process with 3 different media. These were a paper manual, a digital manual accessed through a laptop, and a MR headset. The guides explained how to use metrology instruments and software to complete an inspection process on an assembly. The results obtained were analysed, showing that there is a statistically significant difference between all three interfaces. The interface in which the participants committed the least amount of errors with was the MR interface. The tests performed in this research demonstrated that MR technology can be integrated with current metrology hardware and software. This integration enables the creation of an ISLVM that has the capability of guiding inexperienced users through complex inspection tasks, while committing fewer errors than the current state-of-the-art tools and methods used in industry. Furthermore, this research produced a generic methodology that enables an immersive system to be used in different applications and industries such as manufacturing and assembly processes.

Today, high levels of precision and accuracy are needed in manufacturing to meet the increased complexities in product designs. Most products consist of multiple assembled parts, and fitting these parts together can present a major challenge, especially for complex products. Thus manufacturing with high precision is particularly required, and CNC machining is typically used as a machining process to reduce the risk of parts not fitting together in the assembly process, especially for automatic assembly. Thereby improving quality control and reducing scrap in high-value and low volume production. Over the last 60 years, NC and CNC machines have been used to improve product quality due to their increased accuracy. However, even with today’s more sophisticated machine tools, errors still occur during machining. The literature shows that there are numerous sources of error in machining processes. Additionally, different methods are being used to define and subsequently correct these errors. The methods used to compensate these errors typically depend on offline error compensation. A gap in the existing research methods has been identified as a lack of online error compensation methods to enable parts to be manufactured to specification and corrected during the machining process. The major contribution of this research is the design and implementation of a method for production of right-first-time parts based on an online error compensation. The proposed framework, CLosEd loop MAchining and inspecTIon System (CLeMatiS), is considered to be an important approach for industry to improve the machining and measuring accuracy for high-cost parts. A computational model has been developed, where an algorithm within this model can handle different types of feature relationships and is able to update feature positions based on on-machine measurements. This research shows that the proposed method for compensating the machining errors in order to machine parts right-first-time provides advantages over traditional methods. The method thus improves the positional accuracy of machined features while maintain the relationships between them, compared to the traditional machining.

No description available.

No description available.

In aero engines a thermally sprayed abradable material is applied to the inside of the casing in order to provide a seal around the blade tips of the compressor blades. During running and handling the blades cut a channel into the abradable material. This rub is poorly understood and often results in blade wear and poor sealing particularly in the later stages of the compressor. This thesis aims to investigate and characterise these rubs and provide recommendations for changes to the running and handling procedure, the abradable or the blade tip that will improve sealing and reduce blade wear. The AlSi polyester vs Ti(6Al 4V) and NiCrAl bentonite vs Inconel 718 rubs have been investigated through high speed rub testing. The effect of incursion rate, blade speed and abradable hardness have been investigated. Wear mechanisms have been proposed and clarified by further high speed wear testing and microstructure modelling. In addition several possible tip modifications have been tested and a successful modification has been extensively characterised. The wear mechanism was found to depend on the materials and the incursion rate while the abradable batch and, in the case of the NiCrAl bentonite abradable, the blade speed affected the severity of the mechanism. The AlSi based abradable cut at high incursion rates but wore by adhesion on to the blade an abrasion at low incursion rates. The NiCrAl bentonite abradable wore by local compaction and subsurface damage. At low incursion rate material was removed fast enough to accommodate the incursion but at higher incursion rates blade wear and compaction of the abradable were observed.

The work focuses on the design of the Tuned Mass Damper (TMD), targeted multi-mode, multi-directional vibrations of mechanical structures occurring over a wide temperature ranges. Extension of the target frequency range is achieved by making the devices adjustable, using components with nonlinear load-deflection behaviour. Two nonlinear components that are new in TMD design are studied, namely elastomeric O-rings and Tangled Metal Wire (TMW) particles. Evaluation of the performance of these devices on a typical engineering structure is carried out, and the feasibility of the proposed devices demonstrated. For the O-ring TMD, analytical models are developed to describe the load- deflection behaviour of the O-ring. An existing model for axial compression is improved while new models are established for shear and rocking deformations. Validation of the models is carried out using a specifically designed vibration test. Numerical models, aiming to estimate the elasticity of the O-rings with irregular cross-sectional shape, are developed and validated by comparison with the experimental results. The TMW particles seeks to address high temperature applications. The strong compression-dependent stiffness of these particles provides the basis for an adjustable TMD. Although there is some variation in the stiffness and damping for different collections of particles with similar physical properties, uniformity in- creases after several test runs. According to the assumptions of the equivalence of the TMW materials and the hyperelastic solid, a semi-empirical analytical model is developed and validated using experimental results. A novel design optimisation algorithm, based on the complex power approach, developed to provide an alternative route for the TMD involving nonlinear elements. The proposed route, involving the use of the a numerical, evolutionary search method, is finally applied to the design of a nonlinear TMD.

Nonlinear dynamic behaviour has become increasingly significant due to the performance demands on modern mechanical structures that are increasingly lightweight and flexible, e.g. the geometric nonlinearity caused by the large detection. Also, numerous traditional mechanical applications are found to be able to achieve better performance when nonlinear characteristics are exploited. However, the application of any traditional linear analysis to nonlinear systems can only provide, at best, suboptimal solutions as the well-established linear techniques fail to capture the unique nonlinear features, e.g. modal interactions and bifurcations. This thesis aims to improve the theoretical understanding of the smooth nonlinear dynamic behaviours of mechanical systems and apply the findings to develop innovative approaches for practical use. Backbone curve analysis is employed throughout the thesis as a tool to develop this understanding. The resonant interactions only involving two modes of a three-lumped-mass nonlinear oscillator are investigated. It is demonstrated that the backbone curves of this example system can provide an interpretation of the underlying nonlinear dynamic behaviours, including stability and bifurcations. Then we consider two kinds of triple-mode resonant interactions in other 3-DoF systems, including 1:1:1 and 1:2:3 modal interactions. The effects of these multi-mode resonant interactions, e.g. the non-existent of single- and double-mode responses and the resonance between ‘non-resonant’ modes after involving extra modes, are demonstrated, and the mechanism is explored using backbone curves. A nonlinear dynamic phenomenon, resonant frequency shift, is also considered. The power spectrum density results of a thin plate under multi-mode-multi-frequency excitations are used to demonstrate this nonlinear behaviour, which shows that the frequency shift can be caused by an interaction between any non-resonant modes. Based on a nonlinear reduced-order model, backbone curves are used to explain the mechanism of the non- resonant modal interaction, which is caused by the unconditionally resonant mixed-mode nonlinear terms. The understanding of the non-resonant modal interaction is then used to develop a practical approach for nonlinear system identification which employs the backbone curves as the parametric model. The proposed identification approach is applied to the example plate to demonstrate its accuracy and advantages.

No description available.