Papers by Keyword: Sensitivity Analysis

Abstract: In order to reduce the variation of suspension parameters during wheel runout, andimprove the handling stability of the whole vehicle, a method for optimal design of Torsional beamsuspension structural parameters based on Multi-island genetic algorithm is proposed. Study on theMacpherson Front suspension of a car, based on the Euler's four-element method of multi-bodydynamics and the spatial analytic geometric transformation formula, the mathematical model ofTorsional beam suspension is derived in this paper, the reliability of the mathematical model isverified by the simulation analysis of the corresponding Adams model. On this basis, the degree ofsuspension performance affected by structure parameters is obtained by Sensitivity analysis. Finally,the optimization of suspension performance is realized by modifying the parameters by usingMulti-island genetic algorithm, which plays an important role in the stability and safety of thewhole vehicle.

Abstract: Thanks to their high damage detection sensitivity and low requested power consumption, guided-waves (Lamb waves) have been increasingly used in the last years to monitor the structural integrity in primary and secondary composite structures. The monitoring of the structural health through the propagation of Lamb waves in composite structures is notoriously complex and, for this reason, the development of a prediction model can be a helpful tool for the improvement of Structural Health Monitoring (SHM) systems. Finite Element Method (FE) appears to be the best candidate for such type of simulation. However, since Lamb waves propagation depends strictly on the local material properties of the medium they propagate through, their numerical characterization is a thorny phase. Real composite components are usually affected by the presence of a large number of voids and defects, which cannot be reproduced in numerical models; this leads to a variability of the mechanical properties of materials, with particular reference to elastic moduli and density. These aspects get really ambitious the development of a well-established FE model. In this paper, a design of experiment (DOE) has been carried out to numerically investigate on the effects of the material properties variability on guided-waves time of flight.

Abstract: In the process of studying the curing deformation behavior of fiber metal laminates (FMLs) subjected to thermal stress, there are some discrepancies between analytical models and experimental results for the room-temperature stable configuration. The main reason is the uncertainty of the parameters in the analytical calculation. In present paper, a parametric sensitivity analysis of the influence of Young’s modulus, Poisson’s ratio, thermal expansion coefficients, ply thickness, ply angle and temperature difference on the cured stable configuration is carried out to identify the most sensitive properties. Accurate characterization of these sensitive properties and precise control of manufacture process can reduce the discrepancies and obtain a more accurate prediction model. It is believed that the present study can be helpful of guiding the design and manufacture of the FMLs in the engineer application.

Abstract: This contribution concerns the reproducibility of energy certificates. To examine the impact of different input data assumptions on the results of energy certificates, sensitivity analysis was performed. Conducting such sensitivity analysis manually is not only time consuming and error-prone, but is also typically limited in view of input-data/result combinations. This can be a problem, as a number of input data combinations can have contradictory influences on corresponding KPI (key performance indicator) results. For instance, transparent building components increase not only solar gains, but also transmittance losses. As an alternative to manual modification of input data, the parametric algorithm-editor, Grasshopper (an Add-on to the CAD/CAM-environment Rhino) was utilized. The calculation standards of the Austrian energy certificate procedure were implemented in this environment. Input data values can be modified via a graphical user interface (rulers), and the impact of the change is instantly generated by the environment. Past research regarding planners' usage of energy certificate tools identified usability and overall user acceptance of such calculation tools as a weak spot. Thus, the implementation of calculation routines within a widely used drafting/parametrization tool could result in increased acceptance by planners. In a first methodological approach, the impact of input assumptions regarding building typology and window properties on solar gains, transmittance losses, and heating demand were considered. The contribution illustrates the method and the results of a number of case studies, which show that the concept can be usefully deployed in energy evaluation of buildings. Thereby, the approach can accommodate two requirements: On the one hand, it can result in a tool to evaluate building concepts in early design stages. On the other hand, it can serve as a conceptual vehicle to explore and evaluate calculation methods, such as the standardized energy certification procedure.

Abstract: The behavior of silicon carbide power MOSFETs is analyzed using TCAD device simulations with respect to conduction and switching losses. Device designs with varying breakdown voltages are simulated. The contributions to the on-state resistance are shown at room and elevated temperature. Whereas channel and substrate resistance dominate at low breakdown voltages, drift and JFET resistance dominate at high breakdown voltages. With increasing temperature, the channel resistance decreases and thus the drift resistance is the main contributor already at medium breakdown voltages. Manufacturing processes of a device can have a high influence on its losses. Variations in interface mobility, drift doping, and p-body doping can lead to a significant change of on-resistance, internal capacitances, and reverse recovery charge. For higher voltage classes the drift layer properties should be of major interest as it influences on-resistance and reverse recovery charge.

Abstract: Thermal processes occurring in soft tissues are subjected to laser irradiation are analyzed. The transient bioheat transfer is described by the generalized dual-phase lag model. This model consists of two coupled equations concerning the tissue and blood temperatures supplemented by the appropriate boundary and initial conditions. The efficiency of the internal heat source connected to the laser irradiation results from the solution of the diffusion equation. This approach is acceptable when the scattering dominates over the absorption for wavelengths between 650 and 1300 nm, and just such a situation occurs in the case of soft tissues. Sensitivity analysis with respect to the parameters occurring in the mathematical model is done using the direct approach (differentiation of the basic equations and the boundary-initial conditions with respect to the parameter considered), especially the absorption coefficient and scattering coefficient of the soft tissue are considered. At the stage of numerical modeling the basic problem and additional problems connected with the sensitivity functions are solved using the finite difference method. In the final part the conclusions and examples of computations are presented.

Abstract: In the paper the numerical solution concerning the skin tissue heating in the case of uncertain thermophysical parameters is discussed. The solutions of this type of problems presented previously are based on the application of interval arithmetic. In particular, the parameters appearing in the governing equations and boundary-initial conditions are treated as the interval numbers. Here, the authors propose another approach using for this purpose the methods of sensitivity analysis. The mathematical model of the process concerns the heterogeneous tissue domain subjected to an external heat source. At the stage of numerical modeling both the basic model and the sensitivity ones are solved using the finite difference method. In the final part of the paper, a computational example is presented.

Abstract: Reactive distillation, being an intensified process of combining reaction and distillation in a single vessel, is an ongoing research. This work considered the use of this novel process to investigate how the purity of a fatty acid methyl ester (oleic acid methyl ester – methyl oleate), which is an alternative fuel that has a potential economic bedrock, vary with variations in reflux ratio and reboiler duty with the aid of Aspen Plus model of the process. The column employed in developing the model of the process was a RadFrac type having 29 stages, which was divided into five sections, namely condenser (top), rectifying, reaction, stripping and reboiler (bottom) sections. After the development, the model was simulated using UNIversal QUAsiChemical model (UNIQUAC) base method. In order to investigate the sensitivity of the system, reflux ratio and reboiler duty were varied from 2.0-5.5 and 1350-1800 W, respectively. The results obtained from the steady-state simulation of the process revealed that the developed Aspen Plus model of the system was table because it was able to converge when simulated. Furthermore, it was discovered from the sensitivity analysis carried out that a methyl oleate having a mole fraction of approximately 0.7627 could be obtained from the process when the reflux ratio and the reboiler duty were 2.0 and 1800 W respectively. Also, it was observed that the mole fraction of methyl oleate was sensitive to both reflux ratio and reboiler duty because the variations in them (reflux ratio and reboiler duty) resulted in corresponding variation in the methyl oleate mole fraction obtained from the system.

Abstract: A typical multiscale simulation consists of numerous fine scale models, usually one for each computational point of a coarse scale model. One of possible ways of limiting computing power requirements is replacing fine scale models with some simplified and speeded up ersatz ones. In this paper, the authors attempt to develop a metamodel, replacing direct thermodynamic computations of precipitation kinetic with an advanced approximating model. MatCalc simulator has been used for thermodynamic modelling of precipitation kinetic. Typical heat treatment of P91 steel grade was examined. Selected variables were chosen to be modelled with approximating models. Several attempts with various approximation variants (interpolation algorithms and Artificial Neural Networks) have been investigated and its comparison is included in the paper.

Abstract: Load monitoring and damage identification are important tasks in the field of Structural Health Monitoring and are necessary for assessing the structural integrity and predicting the remaining useful life time. Reconstructing unknown force inputs or system parameters usually involves the solution of an inverse problem which is mostly ill-posed and therefore needs regularization. Using prior information about the desired values is advisable for obtaining meaningful solutions. Damages like for example cracks can often be interpreted as spatial singularities, which cause local stiffness reductions of the observed structures. Damage identification is the task of localizingand quantifying these stiffness reductions. On the other hand, unknown structure excitation usually has also some specia lcharacteristics which can be assumed as known apriori, e.g. spatial concentration for singular forces, short time duration for impact loads or narrow frequency bands for harmonic loads. In this case force reconstruction becomes also a localization and magnitude estimation problem. Thischaracteristic information is used to transform the inverse problem into a sparse recovery task. Inthe last years sparsity constrained regularization of inverse problem has attracted a lot of attention inapplied mathematics, especially in the context of compressive sensing.In this contribution it is shown how sparse solution techniques can be applied in monitoring sys-tems and how this will improve the reconstruction results and additionally reduce the number of required sensors.