Errors of computational simulations in design of a high-speed civil transport (HSCT) are investigated. First, discretization error from a supersonic panel code, WINGDES, is considered. Second, convergence error from a structural optimization procedure using GENESIS is considered along with the Rosenbrock test problem. A grid converge study is performed to estimate the order of the discretization error in the lift coefficient (CL) of the HSCT calculated from WINGDES. A response surface (RS) model using several mesh sizes is applied to reduce the noise magnification problem associated with the Richardson extrapolation. The RS model is shown to be more efficient than Richardson extrapolation via careful use of design of experiments. A programming error caused inaccurate optimization results for the Rosenbrock test function, while inadequate convergence criteria of the structural optimization produced error in wing structural weight of the HSCT. The Weibull distribution is successfully fit to the optimization errors of both problems. The probabilistic model enables us to estimate average errors without performing very accurate optimization runs that can be expensive, by using differences between two sets of results with different optimization control parameters such as initial design points or convergence criteria. Optimization results with large errors, outliers, produced inaccurate RS approximations. A robust regression technique, M-estimation implemented by iteratively reweighted least squares (IRLS), is used to identify the outliers, which are then repaired by higher fidelity optimizations. The IRLS procedure is applied to the results of the Rosenbrock test problem, and wing structural weight from the structural optimization of the HSCT. A nonsymmetric IRLS (NIRLS), utilizing one-sidedness of optimization errors, is more effective than IRLS in identifying outliers. Detection and repair of the outliers improve accuracy of the RS approximations. Finally, configuration optimizations of the HSCT are performed using the improved wing bending material weight RS models. / Ph. D.

Traditional single stage design optimality procedures can be used to efficiently generate data for an assumed model y = f(x^(m),b) + ε. The model assumptions include the form of f, the set of regressors, x^(m) , and the distribution of ε. The nature of the response, y, often provides information about the model form (f) and the error distribution. It is more difficult to know, apriori, the specific set of regressors which will best explain the relationship between the response and a set of design (control) variables x. Misspecification of x^(m) will result in a design which is efficient, but for the wrong model. A Bayesian two stage design approach makes it possible to efficiently design experiments when initial knowledge of x^(m) is poor. This is accomplished by using a Bayesian optimality criterion in the first stage which is robust to model uncertainty. Bayesian analysis of first stage data reduces uncertainty associated with x^(m), enabling the remaining design points (second stage design) to be chosen with greater efficiency. The second stage design is then generated from an optimality procedure which incorporates the improved model knowledge. Using this approach, numerous two stage design procedures have been developed for the normal linear model. Extending this concept, a Bayesian design augmentation procedure has been developed for the purpose of efficiently obtaining data for variance modeling, when initial knowledge of the variance model is poor. / Ph. D.

Impaired reproduction studies with Poisson responses are among a growing class of toxicity studies in the biological and medical realm. In recent years, little effort has been focused on the development of efficient experimental designs for impaired reproduction studies. This research concentrates on two areas: 1) the use of Bayesian techniques to make single regressor designs robust to parameter misspecification and 2) the extension of design optimality methods to the k-regressor model. The standard Poisson model with log link is used. Bayesian designs with priors on the parameters are explored using both the D and F-optimality criteria for the single regressor Poisson exponential model. Since these designs are found via numeric optimization techniques, Bayesian equivalence theory functions are derived to verify the optimality of these designs. Efficient Bayesian designs which provide for lack-of-fit testing are discussed. Characterizations of D, D_s, and interaction optimal designs which are factorial in nature are demonstrated for models involving interaction through k factors. The optimality of these designs is verified using equivalence theory. In addition, augmentations of these designs that result in desirable lack of fit properties are discussed. Also, a structure for fractional factorials is given in which specific points are added one at a time to the main effect design in order to gain estimability of the desired interactions. Robustness properties are addressed as well. Finally, this entire line of research is extended to industrial exponential models where different regressors work to increase and/or decrease a count data response produced by a process. / Ph. D.

Complex mathematical models are often built to describe a physical process that would otherwise be extremely difficult, too costly or sometimes impossible to analyze. Generally, these models require solutions to many partial differential equations. As a result, the computer codes may take a considerable amount of time to complete a single evaluation. A time tested method of analysis for such models is Monte Carlo simulation. These simulations, however, often require many model evaluations, making this approach too computationally expensive. To limit the number of experimental runs, it is common practice to model the departure as a Gaussian stochastic process (GaSP) to develop an emulator of the computer model. One advantage for using an emulator is that once a GaSP is fit to realized outcomes, the computer model is easy to predict in unsampled regions of the input space. This is an attempt to 'characterize' the overall model of the computer code. Most of the historical work on design and analysis of computer experiments focus on the characterization of the computer model over a large region of interest. However, many practitioners seek other objectives, such as input screening (Welch et al., 1992), mapping a response surface, or optimization (Jones et al., 1998). Only recently have researchers begun to consider these topics in the design and analysis of computer experiments. In this dissertation, we explore a more traditional response surface approach (Myers, Montgomery and Anderson-Cook, 2009) in conjunction with traditional computer experiment methods to search for the optimum response of a process. For global optimization, Jones, Schonlau, and Welch's (1998) Efficient Global Optimization (EGO) algorithm remains a benchmark for subsequent research of computer experiments. We compare the proposed method in this paper to this leading benchmark. Our goal is to show that response surface methods can be effective means towards estimating an optimum response in the computer experiment framework. / Ph. D.

Within the realms of biological and medical research, toxicity studies which measure impaired reproduction are becoming more and more common, yet methods for efficiently designing experiments for these studies have received little attention. In this research, response surface design criteria are applied to four models for impaired reproduction data. The important role of control observations in impairment studies is discussed, and for one model, a normal error linear model, a design criterion is introduced for allocating a portion of the sample to the control. Special attention is focused on issues surrounding optimal design of experiments for two of the models, a Poisson exponential model and a Poisson linear model. As most of the optimal designs for these models are obtained via numerical methods rather than directly from criteria, equivalence theory is used to prove analytically that the numerically obtained designs are truly optimal. A further complication associated with designing experiments for Poisson regression is the need to know parameter values in order to implement the optimal designs. Thus, two stage design of experiments is investigated as one solution to this problem. Finally, since researchers frequently do not know the appropriate model for their data a priori, the optimal designs for these two different models are compared, and designs which are robust to model misspecification are highlighted. / Ph. D.

Yes / This paper investigates the coupled effect of functionalized multiwall carbon nanotubes (MWCNTs-COOH), nanomontmorillonite (NM), and sodium dodecyl benzene sulfonate (SDBS) anionic surfactant on compressive and flexural strengths of cement paste. The response surface methodology (RSM) was used to optimize the content of the two nanomaterials and surfactant, and to analyze the effect of their interactions on mechanical properties and microstructural characteristics of the paste. Test results indicate that the simultaneous use of NM and MWCNT can lead to 30% gain in compressive strength and 40% increase in flexural strength. Using analysis of variance, it was possible to predict the optimal weight percentage of nanomaterials. Atomic Force Microscope observations showed that the use of NM and MWCNT can reduce the surface roughness of cement paste and refine porosity, thus reducing the risk of cracking at the cement matrix and improving the homogeneity of the microstructure.

This dissertation explored how a change in the work/home boundary stemming from a mandatory switch to full-time telework influenced employee well-being. Organizational scholars have called for more investigations into how crisis events impact employees, and the COVID-19 pandemic presented an opportunity to examine a change in employees' work and home domains as it unfolded. Additionally, as full-time telework becomes a more common way of work, understanding how this once rare work arrangement affects employee well-being holistically is important. Using boundary theory, I hypothesized that a switch to full-time telework would increase the level of integration between employees' work and home domains, and that a greater change in integration level would associate with worse daily well-being outcomes. To explain this association, I turned to recovery theorizing and proposed daily work-related rumination and lack of psychological detachment as linking mechanisms. Additionally, I expected that teleworkers whose current level of integration was closer to their preferred level would experience better well-being. Using multilevel response surface analysis (MRSA), which enabled illustration of these complex associations in a more nuanced manner than is possible via either change scores or moderation analyses, I found that maintaining higher work/home integration both before and after telework co-varied with worse holistic well-being through work-related rumination and lack of psychological detachment. I also found that having higher integration than preferred and even high integration when preferred associated with worse well-being through work-related rumination and lack of psychological detachment. Based on these results, I point to boundary work and its facilitation of segmentation as a potential means of protecting employee well-being in the event of a future crisis that moves work into the home. / Doctor of Philosophy / This dissertation examined the influence of the COVID-19-induced abrupt and mandatory switch to telework on employees' well-being. More understanding is needed regarding how crisis events impact employees, according to organizational scholars, and the COVID-19 pandemic presented an opportunity to conduct an investigation of change in employees' work situations in real time. As employees experienced change in numerous ways due to the pandemic and stay-at-home orders, I expected that experiencing greater changes in the separation between one's work and home would correlate with well-being impairments. I expected that less separation (i.e., more integration) between work and home would associate with the tendency to ruminate about work during non-work time (work-related rumination) and an inability to detach from work (lack of psychological detachment), which would in turn relate to worse well-being. I also anticipated that employees whose preference in level of work/home integration more closely matched their current situation would enjoy better well-being, potentially due to less work-related rumination and better psychological detachment. Instead, I found that maintaining old habits in how closely integrated employees keep their work and home lives from before to during telework associated with worse well-being. Employees who had more work/home integration pre-telework and maintained more integration during telework showed worse well-being through more work-related rumination and less psychological detachment. Preferring more integration did not appear to protect one's well-being, as those teleworkers who both preferred and enacted more work/home integration had worse well-being through more work-related rumination and less psychological detachment, as did those who had more integration than preferred. Based on these findings, I recommend that in crisis situations and abrupt, mandatory transitions to full-time telework, teleworkers protect their ability to recover from the workday's often strenuous demands by creating separation between work and home. Employees, organizations, and managers can all facilitate boundary work, wherein the teleworker performs actions to create greater separation between work and home, even when fully working from home.

Crystallisation is gaining a lot of interest in pharmaceutical industry to help designing active ingredients with tailored physicochemical properties. Many factors have been found to affect the crystallisation process, including process parameters and material attributes. Several studies in the literature have discussed the role of these parameters in the crystallisation process. A comprehensive study is still missing in this field where all the significant terms are taken into consideration, including the square effect and the interaction terms between different parameters. In this study, a thorough investigation into the main factors affecting crystallisation of a polymeric system, processed via injection moulding, was presented and a sample of response optimisation was introduced which can be mimicked to suite a specific need. Three grades of pure polyethylene oxide; 20K, 200K and 2M, were first characterised using differential scanning calorimetry (DSC), thermogravimetric analysis (TGA), powder X-ray diffraction (PXRD) and shear rheometry. The onset of degradation and the rate varied according to molecular weight of polyethylene oxide (PEO). The peak melting temperature and the difference in enthalpy between melting and crystallisation were both in a direct proportion with PEO molecular weight. PEO200K and PEO2M struggle to recrystallise to the same extent of the original state at the tested cooling rates, while PEO20K can retain up to a similar crystallinity degree when cooled at 1 °C/min. Onset of crystallisation temperature (Tc1) was high for PEO2M and the difference between the 20K and 200K were pronounced at low cooling rate (20K is higher than 200K). The rheometer study showed that PEO2M has a solid-like structure around melting point which explains the difficulty in processing this grade at a low temperature via IM. PEO20K was almost stable within the strain values studied (Newtonian behaviour). For higher grades, PEO showed a shear thinning behaviour. The complex viscosity for PEO2M is characterised by a steeper slope compared to PEO200K, which indicates higher shear thinning sensitivity due to higher entanglement of the longer chains. For binary blends of PEO, the enthalpy of crystallisation studied by DSC was in direct proportion to the lowest molecular weight PEO content (PEOL %) in PEO20K/200K and PEO20K/2M blends. The effect of PEOL% on Tc1 became slightly pronounced for PEO20K-2M blends where Tc1 exhibited slight inverse proportionality to PEOL% and it became more significant for PEO200K-2M blends. It was interesting to find that Tc1 for the blends did not necessarily lie between the values of the homopolymers. In all binary blends, Tc1 was inversely proportional to cooling rate for the set of cooling rates tested. Thermal analysis using hot stage polarised light microscopy yields different behaviours of various PEO grades against the first detection of crystals especially where the lowest grade showed highest detection temperature. Visual observation of PEO binary blends caplets processed at various conditions via injection moulding (IM) showed the low-quality caplets processed at mould temperature above Tc1 of the sample. The factors affecting crystallisation of injection moulded caplets were studied using response surface methodology for two responses; peak melting temperature (Tm) and relative change in crystallinity (∆Xc%) compared to an unprocessed sample. Mould temperature (Tmould) was the most significant factor in all binary blend models. The relationship between Tmould and the two responses was positive non-linear at the Tmould ˂ Tc1. Injection speed was also a significant factor for both responses in PEO20K-200K blends. For Tm, the injection speed had a positive linear relationship while the opposite trend was found for ∆Xc%. The interaction term found in the RSM study for all models was only between the injection speed and the PEOL % which shows the couple effect between these two factors. Molecular effect was considered a significant factor in all ∆Xc% models across the three binary blends. The order of ∆Xc% sensitivity to the change in PEOL% was 3, 5 and 7 % for 20K-200K, 200K-2M and 20K-2M.

No description available.

No description available.