Publications

In this paper, the possibility of designing a five-phase Surface-mounted Permanent Magnet (SPM) machine with 20 slots and 8 poles for a low power marine propulsion system is examined. Due to its particular winding and surface magnet design, the machine inherently offers an electronic pole changing effect from 3×4 pole pairs at low speed to 4 pole pairs at high speed. At high speed, in the constant power range, according to Finite Element Analysis, the Maximum Torque Per Ampere strategy appears not to be the right solution to minimize the whole machine losses (copper, iron and magnets). In particular, a strategy that favors the 4-pole rotating field at high speed allows to mitigate the magnet losses, thus limiting the risk of magnet overheating.

Decision support tools are essential to help in the management of industrial systems at different levels: strategic, to design the system; tactical to plan activities or assign resources; operational to schedule activities. In this paper, we present a generic and modular decision support tool to solve different planning, assignment, scheduling or lot-sizing problems. To the best of our knowledge, such generic tool does not exist. The methodology is illustrated by solving a real world lot-sizing and scheduling problem from a plastic injection company.

Damage-tolerant active control (DTAC) is a recent research area that encompasses control design methodologies resulting from the application of fault-tolerant control methods to vibration control of structures subject to damage. The possibility of damage occurrence is not usually considered in the active vibration control design requirements. Damage changes the structure dynamics, which may produce unexpected modal behavior of the closed-loop system, usually not anticipated by the controller design approaches. A modal H∞H∞ norm and a respective robust controller design framework were recently introduced, and this method is here extended to face a new DTAC strategy implementation. Considering that damage affects each vibration mode differently, this paper adopts the modal H∞H∞ norm to include damage as a design requirement. The basic idea is to create an appropriate energy distribution over the frequency range of interest and respective vibration modes, guaranteeing robustness, damage tolerance, and adequate overall performance, taking into account that it is common to have previous knowledge of the structure regions where damage may occur during its operational life. For this purpose, a structural health monitoring technique is applied to evaluate modal modifications caused by damage. This information is used to create modal weighing matrices, conducting to the modal H∞H∞ controller design. Finite element models are adopted for a case study structure, including different damage severities, in order to validate the proposed control strategy. Results show the effectiveness of the proposed methodology with respect to damage tolerance.

The presented study introduces the development of the multi-reflection grazing-incidence X-ray diffraction method (MGIXD) for residual stress determination. The proposed new methodology is aimed at obtaining more reliable experimental data and increasing the depth of non-destructive stress determination below the sample surface. To verify proposed method measurements were performed on a classical X-ray diffractometer (Cu Kα radiation) and using synchrotron radiation (three different wavelengths: λ = 1.2527 Å, λ = 1.5419 Å and λ = 1.7512 Å). The Al2017 alloy subjected to three different surface treatments was investigated in this study. The obtained results showed that the proposed development of MGIXD method, in which not only different incident angles but also different wavelengths of X-ray are used, can be successfully applied for residual stress determination, especially when stress gradients are present in the sample.

The goal of this work was the determination of lattice strains distribution in two phases of duplex steel during plastic necking. Subsequently, the stress heterogeneity in the neck was studied in order to determine the reason for the damage initiation and to verify the hypothesis that the damage begins in the ferritic phase. To do this, X-ray synchrotron radiation was used to scan the ‘in situ’ variation of the interplanar spacings along the necking zone for samples subjected to tensile loading. A self-consistent model and FEM simulation were applied for the experimental data interpretation. It was found that for advanced necking the phase lattice strains, especially those measured at some distance from the neck centre, show a large inversion of the loads localised in both phases compared to the undamaged state (the lattice strains in the ferrite become smaller than in the austenite). This effect indicates stress relaxation in the ferrite which is connected with the damage phenomenon. Correlation of the experimental results with the modelling shows that the value of von Mises stress is responsible for the initiation of the ferritic phase softening.

Electro-fusion and casting is used to produce large refractory bricks (∼250 kg) containing a high amount of ZrO2. These bricks are used in glass-making furnaces where good mechanical performance is required at very high temperatures (>1500 °C). During the manufacturing procedure, they develop large residual stresses as a result of the cooling conditions and structural phase transformations they underwent. This leads to stress concentration and crack formation at different length scales. In order to characterize these phenomena, a ‘multi-scale’ analysis approach is under development, where different internal strain measurement methods are combined. In this approach we benefit from different gauge volumes provided by various diffraction methods, ranging from a few hundred nanometres to a few tens of millimetres. In the present paper, the results of neutron diffraction measurements on large ZrO2 blocks are given. These results show the level of internal strains at the millimetre scale, based on <img height="17" border="0" style="vertical-align:bottom" width="39" alt="View the MathML source" title="View the MathML source" src="http://origin-ars.els-cdn.com/content/image/1-s2.0-S0955221917300225-si4.gif">(3¯11) reflection of the monoclinic ZrO2. Overall, a range of 0.025% tensile to 0.1% compressive strain was observed. Clear strain gradients were also visible, as larger values in the interior of the block were encountered.

Laser assisted joining of AA5754 aluminum alloy to T40 titanium with use of Al-Si filler wires was carried out. Continuous Yb:YAG laser beam was shaped into double spot tandem and defocalized to cover larger interaction zone in V shaped groove. Experimental design method was applied to study the influence of operational parameters on the tensile properties of the joints. Microstructure examination and fractography study were carried out to understand the relation between local phase content and fracture mode. Within defined window of operational parameters, statistically important factors that influenced the strength of T40 to AA5754 joints in V groove configuration were Si content in the filler metal and groove opening angle on T40 side. The best quality joint showed joint coefficient of 90% (or 200 MPa of apparent UTS). Tensile strength of the joints was found to be determined by the proportion between well-developed and under-developed reaction zones of T40/melted zone interface. The formation of 2–25 μm thick Si-rich interlayers composed by Ti5Si3 and τ2 proved to enhance the strength of brazed interface. The creation of very thin (<0.5 μm) Si-rich layers at the bottom of the groove was found not sufficient to establish mechanical continuity of the joint and thus should be avoided.

Thermal oxidation of three epoxy resins differing by the nature of prepolymer (bisphenol A diglycidyl ether and 1,4-butanediol diglycidyl ether) and hardener (isophorone diamine and 4,7,10-Trioxa-1,13-tridecanediamine) was studied by monitoring changes in glass transition temperature using DSC. Results were discussed using the DiMarzio's approach in which parameters are estimated from an additive group contribution. This theory allowed a fair assessment of Tg values for unaged networks. During oxidation, epoxy networks were shown to undergo chain scissions occurring in great part in hydroxypropyl ether and isophorone groups. However, the exploitation of Tg changes showed the coexistence and even the predominance of crosslinking in materials having linear aliphatic segments. The DiMarzio's approach was used to discuss the possibility of intramolecular cyclization or intermolecular crosslinks which were shown to predominate. Crosslinks were tentatively justified from a mechanistic point of view and quantified depending on experimental conditions.

In order to control laser-induced shock processes, two main points of interest must be fully understood: the laser–matter interaction generating a pressure loading from a given laser intensity profile and the propagation of induced shock waves within the target. This work aims to build a predictive model for laser shock-wave experiments with two grades of aluminum at low to middle intensities (50 to 500  GW/cm 2 500  GW/cm2 ) using the hydrodynamic Esther code. This one-dimensional Lagrangian code manages both laser–matter interaction and shocks propagation. The numerical results are compared to recent experiments conducted on the transportable laser shocks generator facility. The results of this work motivate a discussion on the shock behavior dependence to elastoplasticity and fracturation models. Numerical results of the rear surface velocity show a good agreement with the experimental results, and it appears that the response of the material to the propagating shock is well predicted. The Esther code associated to this developed model can therefore be considered as a reliable predictive code for laser ablation and shock-wave experiments with pure aluminum and 6061 aluminum in the mentioned range of parameters. The pressure–intensity relationship generated by the Esther code is compared to previously established relationships.

The presented work is about optimization of the hospital system. An existing solution is the pooling of resources within the same territory. This may involve different forms of cooperation between several hospitals. Problems of sizing, planning and scheduling may be considered. We define the problem of activities planning with resource assignment. To solve this problem, we propose a hybridization between a metaheuristic and a list algorithm. Single based metaheuristics are used. This proposition requires a new encoding inspired by permutation problems. This method is easy to apply: it combines already known methods. With the proposed hybridization, the constraints to be considered only need to be integrated into the list algorithm. For big instances, the solver used as a reference returns only lower and upper bounds. The results of our method are very promising. It is possible to adapt our method on more complex issues through integration into the list algorithm of the constraints. It would be particularly interesting to test these methods on real hospital authorities to assess their significance.