The aim of the thesis is to rationalize the water heaters welding area in Tatramat company - ohrievače s.r.o. The rationalization consists of three parts. The first part deals with the arrangement of the workplaces, the second part re-evaluates the monthly production planning. The third part describes in detail the production of one type, which based on simulations created in the simulation software Witness, compares the recorded time in company informartion system Orfert to the real production time in the operation. Each individual part offers optimization proposals and merging them into one unit, will create an efficient, transparent and economically value adding rationalization of the water heaters welding area.

Natural or man-made disasters result in unfortunate events around the nation every year. Such extreme events necessitate the short-notice or no-notice evacuation of a large population from the stricken area. This research presents an optimization modeling technique to develop an evacuation plan for transit-dependent residents during no-notice disaster situation. The proposed plan relies on the application of existing public transit system of an urban area. The public transit routing plan (PTRP) problem is formulated as a mixed integer linear program. The PTRP identifies the optimal routes for transit vehicles to move evacuees from the danger zone to designated safe destinations. A heuristic TABU search algorithm is used to find high-quality solution in a reasonable amount of time. Finally, DYNASMART-P is used to evaluate the effectiveness of the developed PTRP. Numerical experiments are conducted using the traffic network of the city of Fort Worth, TX, to illustrate the proposed modeling technique.

Doctor of Philosophy / Department of Industrial & Manufacturing Systems Engineering / Jessica L. Heier Stamm / This dissertation focuses on logistics challenges arising in the biofuels industry. Studies have found that logistics costs in the biomass-to-biofuel supply chain (BBSC) account for 35%-65% of total biofuel production cost. This is mainly due to the low density of biomass that results in high costs associated with biomass transportation, storage, and handling in the biomass-to-biofuel supply chain. Densification provides an as-yet-unexplored opportunity to reduce logistic costs associated with biomass-to-biofuel supply chains. This research advances understanding about biomass-to-biofuel supply chain management through new optimization models. As a first step, the author presents an extensive overview of densification techniques and BBSC optimization models that account for biomass densification. This literature review helps the author to recognize the gaps and future research areas in BBSC studies. These gaps direct the author toward the remaining components of the dissertation. In particular, the literature review highlights two research gaps. First, the review indicates that mobile pelleting holds promise for improved BBSC management, but that there is no mathematical optimization model that addresses this opportunity. Second, currently, there does not exist a model that explicitly accounts for farmers’ objectives and their probability to sell biomass to the bioenergy plant in BBSC optimization. To fill the first gap, the author focuses on managing the BBSC considering mobile densification units to account for chances to minimize logistics costs. A mixed integer linear programming model is proposed to manage the BBSC with different types and forms of biomass feedstock and mobile densification units. Sensitivity analysis and scenario analysis are presented to quantify conditions that make mobile densification an attractive choice. The author conducts a case study to demonstrate model applicability and type of analysis that can be drawn from this type of models. The result indicates that mobile pelleting is not an attractive choice under the current economic status. However, modest changes in pelleting cost, satellite storage location fixed cost, and/or travel distances are enough to make mobile pelleting an attractive choice. To fill the second gap, the author introduces a model that explicitly accounts for mobile densification and farmers’ probability to supply a bioenergy plant with biomass feedstock. Farmers’ probability to provide biomass to the bioenergy plant depends on contract attributes, including expected net return and services provided by the bioenergy plant. The proposed model helps the bioenergy plant to meet biofuel demand while considering farmers’ choices that satisfy their own objectives and preferences. The model makes it possible to determine most important factors that influence type of contract offered to each supplier and optimal BBSC design. A case study based on the state of Kansas is conducted to demonstrate how bioenergy plant can benefit from this type of model.

<p>Hemijskim transformacijama se iz biljnih ulja dobijaju vredni &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; derivati, poput epoksidovanih biljnih ulja, koji se<br />koriste u hemijskoj i polimernoj industriji. Kvalitet, pa<br />time i primena epoksidovanih biljnih ulja, zavise od<br />sadržaja epoksidnih grupa u derivatizovanom ulju, koji bi<br />trebalo da je &scaron;to veći. Kako epoksidne grupe nastaju<br />oksidacijom dvostrukih veza u trigliceridima kao<br />dominantnoj grupi jedinjenja u biljnim uljima, pogodna<br />sirovina za epoksidovanje su visoko nezasićena ulja,<br />kakvo je laneno.<br />Proizvodnja epoksidovanih biljnih ulja zahteva izbor<br />takvih procesnih uslova pri kojima bi se postigli &scaron;to<br />potpunija konverzija dvostrukih veza i &scaron;to veća<br />selektivnost procesa u odnosu na epoksidnu grupu. Zato je<br />kao cilj ove doktorske disertacije postavljeno određivanje<br />optimalnih vrednosti procesnih uslova epoksidovanja<br />lanenog ulja persirćetnom kiselinom formiranom in situ iz<br />sirćetne kiseline i 30% vodenog rastvora vodonik<br />peroksida u prisustvu jonoizmenjivačke smole kao<br />katalizatora. Određivanje je izvedeno primenom<br />metodologije odzivne povr&scaron;ine, kao i kori&scaron;ćenjem u ovoj<br />disertaciji predloženih kinetičkih modela ispitivanog<br />reakcionog sistema, u oba slučaja sa maksimumom<br />relativnog prinosa epoksida kao funkcijom cilja.<br />Ispitivanje uticaja procesnih uslova, i to temperature,<br />molskog odnosa reaktanata, količine katalizatora i brzine<br />me&scaron;anja, na tok procesa epoksidovanja je bilo osnov za<br />definisanje graničnih vrednosti procesnih uslova unutar<br />kojih je tražen maksimum prinosa epoksida. Oblast dugih<br />vremena reagovanja, koja nije od interesa za industriju, je<br />izbegnuta adekvatnim izborom temperature.<br />Pri optimalnim vrednostima procesnih uslova<br />epoksidovanja lanenog ulja, određenim primenom<br />metodologije odzivne povr&scaron;ine, postignuto je dobro<br />slaganje očekivane i eksperimentalno određene vrednosti<br />maksimalnog relativnog prinosa epoksida, sa odstupanjem<br />od 3,28%.<br />Za potrebe određivanja optimalnih uslova izvođenja<br />procesa epoksidovanja biljnih ulja kori&scaron;ćenjem kinetičkih<br />modela, razvijena su tri pseudohomogena modela<br />ispitivanog trofaznog multireakcionog sistema. Pored<br />kinetike osnovnih reakcija formiranja persirćetne kiseline i&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; reakcije epoksidovanja dvostrukih veza triglicerida biljnog ulja,&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; kao i kinetike sporedne reakcije otvaranja epoksidne<br />grupe sa sirćetnom kiselinom, u predloženim modelima je<br />uzeta u obzir i raspodela sirćetne i persirćetne kiseline<br />između uljne i vodene faze sistema. Opisan je i uticaj<br />masno-kiselinskog sastava ulja, odnosno broja dvostrukih<br />veza u masno-kiselinskim lancima triglicerida, na kinetiku<br />reakcija. Za koeficijent raspodele sirćetne kiseline između<br />tečnih faza reakcionog sistema predložena je empirijska<br />korelacija koja je dala dobro slaganje izračunatih<br />vrednosti sa eksperimentalnim podacima. Kinetički<br />parametri modela su određeni fitovanjem<br />eksperimentalnih podataka o promenama količina<br />dvostruke veze i epoksidne grupe sa vremenom izvođenja<br />procesa epoksidovanja. Na osnovu statističkih pokazatelja<br />uspe&scaron;nosti fitovanja eksperimentalnih podataka, potvrđena<br />je prepostavka da je pseudohomogeni model publikovan u<br />literaturi unapređen uzimanjem u obzir pomenutih<br />fenomena raspodele komponenata reakcione sme&scaron;e i<br />masno-kiselinskog sastava sirovine pri modelovanju<br />reakcionog sistema epoksidovanja biljnih ulja<br />persirćetnom kiselinom.<br />Kori&scaron;ćenjem predloženih pseudohomogenih modela<br />reakcionog sistema za određivanje optimalnih uslova<br />izvođenja procesa epoksidovanja lanenog ulja in situ<br />formiranom persirćetnom kiselinom u prisustvu<br />jonoizmenjivačke smole, dobijeno je odstupanje od 5,51%<br />očekivane od eksperimentalno određene vrednosti<br />relativnog prinosa epoksida.<br />Bolje slaganje predviđene sa eksperimentalno određenom<br />vredno&scaron;ću relativnog prinosa epoksida u kontrolnom<br />eksperimentu je dobijeno primenom metodologije odzivne<br />povr&scaron;ine u poređenju sa kori&scaron;ćenjem kinetičkih modela pri<br />određivanju optimalnih vrednosti procesnih uslova. To je i<br />očekivano, s obzirom da regresiona jednačina kori&scaron;ćena u<br />okviru metodologije odzivne povr&scaron;ine bolje fituje relativni<br />prinos epoksida. Standardna devijacija relativnog prinos<br />epoksida za regresionu jednačinu je 8,9 puta niža od one<br />izračunate za kinetički model koji najbolje predviđa<br />optimalne procesne uslove epoksidovanja lanenog ulja<br />persirćetnom kiselinom.</p> / <p>Vegetable oils can be transformed into added value<br />products by various chemical modifications, such as<br />epoxidation. The epoxidized vegetable oils have a<br />wide range of applications in the chemical and<br />polymer industry. The quality, and consequently the<br />application, of epoxidized vegetable oil is influenced<br />by the epoxy group content. Since the epoxy groups<br />are formed by the oxidation of double bonds in<br />triglycerides, the main constituent of vegetable oils,<br />highly unsaturated vegetable oils, such as linseed<br />oil, are desirable raw materials.<br />The manufacturing of epoxidized vegetable oils<br />requires the optimization of the process conditions<br />in order to achieve complete conversion of double<br />bonds and high selectivity of the process in respect<br />to the epoxy groups. Therefore, the aim of this<br />doctoral thesis is to determine the optimal process<br />conditions for the epoxidation of linseed oil with<br />peracetic acid, formed in situ from acetic acid and<br />30% hydrogen peroxide in the presence of an ion<br />exchange resin as the catalyst. The optimal process<br />conditions were determined by response surface<br />methodology, as well as by using developed pseudohomogeneous<br />kinetic models that describe the<br />investigated reaction system. For both optimization<br />methods, the relative epoxy yield was selected as an<br />objective function to be maximized.<br />The effects of process conditions, such as<br />temperature, molar ratio of reactants, catalyst<br />amount and steering speed, on the kinetics of the<br />epoxidation were studied in order to define<br />constraints for the optimization. To avoid long<br />reaction times, which are not of interest in<br />manufacturing, an adequate temperature range was<br />selected. Under the optimized process conditions for the<br />epoxidation of linseed oil, which were determined<br />by response surface methodology, good agreement<br />between the calculated and experimentally<br />determined relative epoxy yields was achieved<br />within 3.28%.<br />Three models describing the three-phase multireaction<br />system of vegetable oil epoxidation with<br />peracetic acid were developed and further used for<br />the optimization. The models are pseudohomogeneous<br />with respect to the catalyst. Besides<br />the kinetics of the main reactions of peracetic acid<br />and epoxy group formation, the models take into<br />account the side reaction of the epoxy group opening<br />with acetic acid. The partitioning of the acetic acid<br />and peracetic acid between the oil and aqueous<br />phases is considered. In two proposed models, the<br />effect of fatty acid composition on the kinetics of the<br />process is also described by considering the number<br />of double bonds in the fatty acid chains. The<br />developed empirical correlation for the partition<br />coefficient for acetic acid between the liquid phases<br />shows good agreement between the calculated and<br />experimental data. The kinetic parameters of the<br />proposed pseudo-homogeneous models were<br />determined by fitting the experimentally determined<br />changes of the double bond and epoxy group<br />amounts with reaction time of the epoxidation.<br />Statistical values of the models` parameters<br />determination confirmed the hypothesis that the<br />pseudo-homogeneous model proposed in the<br />literature can be improved by considering the<br />partitioning phenomena and the effect of the oil fatty<br />acid composition on the kinetics of the vegetable<br />oils epoxidation with peracetic acid.<br />Under the optimized process conditions for the<br />epoxidation of linseed oil with peracetic acid formed<br />in situ in the presence of the ion exchange resin,<br />which were determined by using proposed pseudohomogeneous<br />models, the experimentally<br />determined relative epoxy yield was 5.51% lower<br />than the calculated.<br />Better agreement between the calculated and<br />experimentally determined values for the relative<br />epoxy yield, achieved under the optimal process<br />conditions, is obtained when the response surface<br />methodology (RSM) was applied as opposed to<br />when the kinetic models were used for the<br />determination of the optimal process conditions.<br />This is in accordance with better fitting of the<br />relative epoxy yield by RSM regression equation<br />than by kinetics models. Standard deviation of the<br />relative epoxy yield for RSM regression equation is 8.9 times lower than the standard deviation for the<br />most successful kinetic model used for prediction of<br />the optimal process conditions for the epoxidation of<br />the linseed oil by peracetic acid.</p>

Hexapod trusses are an important element in many mechanical design systems. The natural frequency and stiffness behavior under geometric and mass variations of such structures is largely undocumented. Furthermore, the ability to quickly model hexapod designs and explore a large design-space in finite element software packages is, in general, time consuming and inefficient. The purpose of this project was to develop software tools that made design-space exploration (modeling and simulation processes) for hexapod structures drastically more efficient. Secondly, the project included an experimental analysis portion to demonstrate the various modal study techniques and to validate finite element analysis predictions. Lastly, the project investigated a specific hexapod design problem as a means of exhibiting the modeling/optimization software tools and to develop an understanding of the natural frequency behavior of hexapods. To this end, the research could be used for the design of telescope secondary support structures and other hexapod optimization engineering problems.

There is growing interest in commercial aircraft formation flight as a means of reducing both airspace congestion and the carbon footprint of air transportation. Wake vortex surfing has been researched extensively and proven to have significant fuel-saving benefits, however, commercial air transportation has yet to take advantage of these formation benefits due to understandable safety concerns. The realization of these formations requires serious consideration of formation contingencies and safety during closer-in maneuvering of large commercial aircraft. Formation contingency scenarios are much more complex than those of individual aircraft and have not yet been studied in depth. This thesis investigates the utility of optimization modeling in providing insight into generation of aircraft escape paths for formation contingency planning. Three high-altitude commercial aircraft formation scenarios are presented; formation join, formation emergency exit, and formation escape. The model-generated paths are compared with pilot-generated escape plans using the author’s pilot expertise. The model results compare well with pilot intuition and are useful in presenting solutions not previously considered, in evaluating separation requirements for improvement of escape path planning and in confirming the viability of the pilot-generated plans. The novel optimization model formulation presented in this thesis is the first model shown to be capable of generating escape paths comparable to pilot- generated escape plans and is also the first to incorporate avoidance of persistent and drifting wake turbulence within the formation.