Legacy Theses and Dissertations (1940-2009)
Permanent URI for this collectionhttps://hdl.handle.net/10657/6771
This collection gathers digitized University of Houston theses and dissertations dating from 1940.
Browse
Browsing Legacy Theses and Dissertations (1940-2009) by Department "Chemical and Biomolecular Engineering, Department of"
Now showing 1 - 20 of 280
- Results Per Page
- Sort Options
Item A cell for measuring the influence of magnetic field strength on the thermal conductivity of gases(1969) Elliot, Douglas Gene; Honeywell, Wallace I.; Prengle, Herman William, Jr.; Bannerot, Richard B.An apparatus is described for the determination of the influence of magnetic field strength on the thermal conductivity of gases. It was designed for use with an existing superconducting magnet system capable of field strengths from 0 to 100 Kg. Accordingly, a concentric cylinder type cell with axis parallel to the magnetic field was chosen as an optimum geometry. The design utilizes "cold finger type" cooling with a helium exchange gas reservoir and automatic temperature controller to maintain any selected temperature in the 77 to 300[degrees] Kelvin range. [...]Item A clinical and engineering evaluation of hypothermic fluidotherapy(1978) Diamond, Louis Ralph; Henley, Ernest J.; Huang, Chen-Jung; Kirkpatrick, C. V.; Everett, Robert L.A fluidized bed to deliver cold treatments was designed, built, and tested. Comparisons were made between this fluidotherapy unit and cold water. The effects on patient comfort, change in size of healthy feet, and the ability to remove heat were used as criteria. These tests were completed and the unit was taken to the University of Houston Athletic Department. Any traumatic foot or ankle injuries incurred by the members of the football team would have been treated. However the opportunity to make these treatments did not arise as there were no appropriate injuries during the trial period. [...]Item A clinical and engineering study on thermal therapy(1974) Ho, Purvis Ka-Kwong; Henley, Ernest J.; Flumerfelt, Raymond W.; Kay, Franklin J.A new heat modality was developed based upon the principles of fluidization. This design uses a fluidized airsolid system as the heat transfer medium in thermal therapy application. The new device known as the fluidotherapy unit is currently in use at St. Anthony Center, Houston, Texas, for clinical evaluations. Following the development of the fluidotherapy unit, an engineering and clinical assessment of three modalities in thermal therapy was carried out. The devices evaluated were a) the fluidotherapy unit, b) a whirlpool, and c) a paraffin wax/oil bath. The objective of the study was to compare their Relative effectiveness in delivering heat into a human hand. Steady state heat transfer in hand was considered rather than the transient heating. The amount of heat absorbed by a submerged hand in fifteen minutes were measured. The ratios were found to be 3.9 : 1.3 : 1.0; the fluidotherapy unit was observed to be the most effective heating device. In fifteen minutes, 17.2 Btu was delivered into a hand by the fluidotherapy unit, 5.6 Btu by the whirlpool, and 4.8 Btu by the paraffin. In addition to its relative high efficacy as a thermal therapeutic agent the fluidotherapy unit provides good patient motility. Finally a clinical study on the effects of prescription time on patient progress in rehabilitation was conducted. Active joint motions of patients were measured daily, the increase in joint motion was taken to represent the progress profile of a patient. Results obtained in this study were limited, but indicate that longer treatment times are beneficial.Item A computer model for the calculation of epicardial coronary blood flow(1985) Wiesner, Theodore F.; Nerem, R. M.; Flumerfeldt, R.W.; Chang, H.C.; Luss, DanThe model of epicardial coronary flow used here is one which includes the effects of branching as well as the nonlinear elastic properties of the artery walls. Branches are included on a discrete basis where at each branch point a mass balance and two momentum equations are written. These equations together with an additional equation, i.e., a lump impedance for each branch, determine the flow conditions in the region of branching. Between branch points, a tube is modeled as being straight and elastic in nature. The flow in each such tube of the arterial system is assumed to be that of an incompressible, Newtonian fluid whose motion can be adequately described using the unsteady, one-dimensional equations of motion. The effect of wall friction is specified through a friction factor, and the radial inertia of the tube and fluid is neglected. The elastic properties of the arteries exhibit themselves through an area-pressure relationship. These are entered through a specification of the propagation speed, c, of small amplitude pressure waves in an artery. With regard to boundary conditions, an upstream pressure is specified. The downstream boundary condition may be either specified as a pressure, if the artery continues as a finite branching system, or in terms of a terminal resistance and capacitance if the structure of the downstream region is modeled on a lumped basis. This model has been applied to the left coronary circulation which is modeled with epicardial coronary vessels having a discrete branching pattern. Each branch terminates in a lumped impedance which models the intramyocardial resistance of those vessels within the heart muscle. The upstream, driving pressure at the left coronary ostium is taken to be the aortic pressure, and the wavespeed in the coronary vessels is taken from in vivo measurements. Calculations have been carried out for both the dog and for man using geometry taken from measurements of coronary casts. Results have been obtained for the normal coronary circulation as well as for cases which include multiple stenoses and aorto-coronary bypasses.Item A correlation of friction factors for pipe flow over two-dimensional, periodic geometric roughness(1971) Gaddis, Emerson Clifford, Jr.; Worley, Frank L., Jr.; Dukler, A. E.; Dalton, Charles; Flumerfelt, Raymond W.; Hwang, Neddy H. C.Friction effects of two-dimensional, periodic, artificial roughness for fully roughened pipes in turbulent flow were investigated. Friction factor data over many different shapes, sizes, and spacings of roughness were compiled and examined for similarities which could be correlated. Such shapes included fins, round rods, rectangular rods, square bars, v- grooves, and sinusoids. It was found that friction factors for two differently- shaped roughness waveforms of the same spacing, amplitude, and pipe diameter are related by the normalized crosscorrelation coefficient of their waveforms. Such roughness falls into two groups: projections, where the bulk flow interacts with the flow near the wall; and grooves, where captive vortices exist in the roughness cavities and do not interact with the bulk flow. Friction factors were related to sine waves of similar size, modified by linear combinations of roughness dimensions and plotted versus a representing parameter R, which was found to be a dimensionless group containing a measure of spacing, amplitude, and pipe diameter. Two correlation plots were developed, for projections and grooves. These show distinct regions which are shown to delineate geometries where various wall flow processes occur. Predictions can thus be made of friction factor and flow type given only the roughness system geometry and shape.Item A CSTR measurement of high pressure catalytic methanation(1975) Hadjigeorghiou, Georghios A.; Richardson, James T.; Howell, John Reid; Luss, DanA continuous stirred tank reactor for high pressure catalytic methanation studies has been designed, built, and tested. Specifications, operating procedure, and limitations of the system are presented. This reactor is to be used to investigate the mechanism of the methanation reaction, as well as, to investigate the support effect on catalytic activity for various catalysts and supports. The kinetics of methanation over nickel catalyst were studied, in order to demonstrate the use of the reactor and to evaluate its performance. The operating region of the reactor, where prefect mixing prevails, was determined by pulse testing. External diffusional effects on the reaction rate were eliminated by running the reactor at high speed (determined experimentally) where such effects are absent. Intraparticle diffusional effects were eliminated by measuring effectiveness factors, which determines the maximum catalyst size. [...]Item A heuristic method for solving the problem of upgrading the system availability through redundant allocation(1982) Yasutake, Karen Michiko; Henley, Ernest J.; Law, Japhet S.; Pollard, Richard; Luss, DanA heuristic method has been developed to solve the problem of computing the optimal number of redundant components needed to attain an improved system availability without violating a cost constraint. This problem is a nonlinear problem, where the objective function is non-linear, the constraint can oe linear or non-linear and the solution must be pure integer. The heuristic method developed for this thesis was tested with example problems given by other methods found in the literature. This heuristic method solved these problems with ease and produced good results. A program was developed, entitled Upgrading of System Availability Package (USAP) which will: a) Generate the cut sets. b) Assess the system availability before upgrading. c) (via heuristics) It will upgrade the system availability without violating the cost constraint, or upgrade the system availability until a desired availability is attained even if it violates the cost constraint. d) Assess the final system availability. The advantage of this heuristic method is that the cost constraint is not restricted to a linear form. The parallel, identical components within a stage may be in hot or cold standby and they may also be a voting system.Item A hybrid computer study of the dynamics of a tubular chemical reactor(1970) Lee, Hong-Mou; Motard, Rodolphe L.; Bannerot, Richard B.; Childs, S. Bart; Flumerfelt, Raymond W.; Honeywell, Wallace I.The stable hybrid computer solution of a time-dependent tubular chemical reactor represented by a system of parabolic or elliptic- parabolic partial differential equations is studied. In the classical approach to the serial hybrid solution of the one-dimensional diffusion equation using the continuous-space-discrete-time (CSDT) technique, there exists an undesirably large amount of positive analog loop feedback. This makes the classical hybrid method highly unstable in the study of higher frequency transient behavior. The serial decomposition method used in this study replaces the linear second order differential operator by two stable first order operators integrating in opposite directions and yields one-pass solutions instead of the usual iterative solutions. Thus, considerable computation economy can be expected. [...]Item A low temperature investigation of the effect of magnetic fields on the resistance characteristics of a thermistor(1969) Stein, Louis Edward; Honeywell, Wallace I.; Witte, Larry C.A low temperature, experimental study was conducted to determine the influence of magnetic fields on the resistance characteristics of a cryogenic Keystone Carbon thermistor. [...]Item A model of photochemical smog including gas-to-particle conversion for humid and dry climates(1978) Nicknish, Christine A.; Worley, Frank L., Jr.; Rabalais, J. Wayne; Payatakes, Alkiviades C.A model of photochemical smog is presented that includes the interaction of the gas phase molecules with aerosol particles. The model demonstrates how high relative humidity maintains the lower oxidant levels observed in humid climate cities such as Houston as compared to dry climate cities such as Los Angeles. A quasi-Langrangian approach is taken by modeling a homogeneous air parcel which follows an air trajectory. Four of the twenty species modeled are solved by the quasi steady state approximation (QSSA). The set of sixteen ordinary nonlinear stiff differential equations is solved by Michelsen's (1976) semi-implicit Runge-Kutta method. This numerical method directly handles the explicit time dependence of the emission rates, photodissociative reaction rates, and the empirically determined aerosol growth patterns. At high relative humidity there is an increase in the growth of aerosols resulting in a large available surface area for collision. The model assumes that certain molecular species that either directly or indirectly contribute to oxidant levels are partially removed through absorption by aerosols. Empirically determined aerosol growth patterns were presented for humid and dry climates. The humid versus dry climate aerosol growth patterns were compared under a variety of theoretical absorption conditions. Case studies included studying the effect of different combinations of molecules and radicals considered to be absorbed at various accommodation coefficients, the effect of initial concentrations of the species, and the effect of different emission rates. When only radicals were considered absorbed the model showed insignificant differences between the two types of aerosol growth patterns. When molecules as well as radicals were considered absorbed, oxidant levels were considerably lower for humid climate aerosol growth patterns under all conditions studied.Item A new algorithm for multi-stage, multi-phase distillation and extraction(1982) Chen, Sze-Cheng; Henley, Ernest J.; Pollard, Richard; Rhodes, Benjamin T., Jr.The stage-to-stage flash algorithm (MSEQ) was extended for use in the three-phase, liquid-liquid-vapor, distillation problem. The simulation model can be used not only for the simulation of three-phase distillation columns, but also for columns with any combination of two- and three-phase stages. The subprogram BSOLVE (which uses Marquardt's method) is used to solve the non-linear material and energy balances in adiabatic flash problems. The computer program developed, entitled Three-Phase MSEQ Algorithm (TPMA) : a) Uses FSOLVE to solve the adiabatic flash problem on all stages between the condenser and reboiler. b) Performs a percent vapor flash on the reboiler. c) Calculates the bubble point of the condenser. d) Iterates the whole procedures until successive answers are within tolerance. e) Prints out the final answers. One very important advantage of the algorithm is that one can set initial values of the compositions of all three phases, vapor phase fraction, liquid phase separation ratio, and temperature for each stage arbitrarily and still obtain results. In this thesis NRTL was used for determining the liquid phase activity coefficients. As the vapor phase is often close to ideal, it was assumed that the equilibrium vaporization ratio or K value is a function of the system temperature, pressure, and composition of liquid phase only. The number of stages, feeds and sidestreams is not restricted. An extended application to liquid-liquid counter current extraction using the stage-to-stage iterative procedure was also developed. For this case, the Modified Wilson equation was used to calculate the equilibrium data. The methods developed were tested with examples found in the literature. They solved these problems with ease and produced good results.Item A new model of the constricted unit cell type for granular porous media and collocation solution of the creeping Newtonian flow problem(1977) Neira, Mariano Alberto; Payatakes, Alkiviades C.; Bailey, James E.; Tiller, Frank M.; Wheeler, Lewis T.The constricted unit cell model for granular porous media developed by Payatakes, Tien and Turian (1973) is extended here to take in account the random orientation and interconnectivity of the flow channels. In the proposed model each unit cell corresponds to a pore (cavern) and has two coaxial constricted inlet and outlet ports ("throats"). The unit cells have random dimensions and orientations, the distributions of which can be determined from simple experimental measurements. The flow through a unit cell is assumed to be identical to that through a segment of the corresponding periodically constricted tube. A collocation solution of creeping newtonian flow through periodically constricted tubes is obtained. The profile of the wall of the type of tube considered is piecewise smooth, composed of symmetric parabolic segments and accords with the unit cell geometry of the porous media model. A transformation of the domain of interest into a rectangular one is obtained, which allows satisfaction of all boundary conditions. The collocation solution gives the stream function in terms of the new independent variables and can easily be converted to the original cylindrical coordinates. Axial velocity and radial velocity are obtained in analytical form, and the pressure drop is calculated both from integration of the energy dissipation function and of the Navier-Stokes equations. The results are compared with the finite-difference solution by Payatakes et al. (1973b) and are found in good agreement. Differences between the two solutions are attributed mainly to discretization error in the finite-difference solution. Based on the collocation solution, the new model is used to study the flow of fluids through granular randomly packed beds.. Experimental and theoretical predictions of permeability for two samples are compared and found in remarkable agreement. Intended applications of the model coupled with the present flow solution are in the modeling of processes taking place in the macroporous space of beds of monosized or nearly monosized grains (deep bed filtration, etc.). A modified version of this model will be applied to the modeling of secondary and tertiary oil recovery from oilbearing rocks.Item A numerical study of the displacement of fluids confined by parall plates(1974) Smith, Gene Stanford; Flumerfelt, Raymond W.; Tiller, Frank M.; Dalton, Charles; Henley, Ernest J.An investigation of the vertical displacement of one Newtonian fluid by another is reported. The fluids are immiscible, incompressible and contained by vertical parallel walls. The displacement process was numerically simulated with a computer algorithm (UHMAC) which solves the two-dimensional transient form of the Navier-Stokes equations by use of a modified form of the marker and cell method (MAC method). A comparison of the UHMAC results for the solution of an inlet flow problem and a confined wake problem indicate good agreement with the steady-state solutions obtained by other investigators. UHMAC results for a transient, two-phase problem (fractured diaphragm) were also quite good. The displacement process was simulated for several pairs of fluids with different density ratios and viscosity ratios. The effects of the density ratio and viscosity ratio on displacement efficiency were determined and compared with an approximate analytical solution given previously. The UHMAC results show that displacement efficiency increases as the ratio of the lower fluid density to the upper fluid density is increased. They also show that displacement efficiency increases as the ratio of lower fluid viscosity to upper fluid viscosity is increased. These results agree qualitatively with the results of the analytical solution. The simulation results also show that part of the upper phase fluid may be bypassed and left against the wall when the upper phase is more dense than the lower phase. The results, although limited to parallel plate geometry and Newtonian fluids, show the nature and importance of the variables involved in practical displacement problems such as oil and gas well completion operations.Item A preliminary study of gravity filtration(1977) Ah-Kee, Felix; Tiller, Frank M.; O'Neill, Michael W.; Bailey, James E.A methodology based on falling head gravity filtration was developed to allow cake filtration to be rapidly determined in the laboratory. Tests were run on talc and kaolin, and the effects of various filteraids on the filtration resistances of the two materials were studied. Plots of average filtration resistances versus mass fraction of filteraids in the solids mixture were found to fit a general equation for hyperbolas. I . Based on Shirato's Additive Law for ideal mixtures, an equation was developed to calculate the average filtration resistance of a mixture of two solids. The discrepancies between the experimental values of average filtration resistances and the corresponding ones calculated by means of the Additive Law increased with increasing lack of match between the particle size distributions of the two components. Calcium carbonate was unexpectedly found to act as a filteraid when mixed with talc, but not when mixed with kaolin. Calcium carbonate had low positive zeta potentials at pH values close to 7.0 and low negative zeta potentials at pH values above 8.1. Both kaolin and talc had nearly identical negative zeta potentials over a wide range of pH values (4.0 to 10.0). A study of scanning electron microscope pictures of talc and calcium carbonate mixtures revealed highly flocculated, irregular particles of calcium carbonate lodged between unevenly distributed, plate-like particles of talc. Although no information could be drawn about the relative sizes of the pores, it was apparent from the photomicrographs that some interaction had occurred which had changed the shape of the pores of the talc particles.Item A preliminary testing of a photochemical smog model and an organic analysis of airborne particulate matter(1981) Choinski, Mary B. Carlson; Worley, Frank L., Jr.; Pollard, Richard; Weinheimer, Alfred J.There has been considerable interest in modeling the photochemical processes that occur in the atmosphere. Most models either simulate gas phase reactions or aerosol particle growth. Recent model developments have included the interaction of the volatile phase with the aerosol phase. In this study, a-preliminary test of a photochemical smog model is performed. This model includes the loss,of molecules and radicals to aerosols along with the gas phase reactions. Observed concentration profiles of oxidants and precursor species were compared to simulated profiles. In addition to this comparison, the organic composition of airborne particles was investigated to obtain more information about the nature of aerosol formation by photochemical mechanisms. A gas chromatograph/mass spectrometer/data system instrument was used for the analysis. In the comparison of the observed and simulated profiles, there is evidence that the chemical kinetic mechanisms used in the model are acceptable. It is also apparent that the emission function in the model needs to be revised to better simulate early morning hour concentrations. There were over 100 compounds identified in the aerosol phase. Many of the compounds were oxygenated indicating that they are formed by chemical reactions either in the gaseous or aerosol phase. These results lend additional weight to the understanding of the photochemical processes occurring in the atmosphere.Item A quantitative analysis of the effects of Fluidotherapy heat treatment on the human hand(1980) Lober, Steve B.; Henley, Ernest J.; Nerem, Robert M.; Flumerfelt, Raymond W.; Luss, DanThe quantitative effects of heat treatment (by Fluidotherapy) on the human hand were measured in order to establish guidelines for choosing optimal treatment conditions. Tissue oxygen consumption rate changes, the criterions used to evaluate the effectiveness of the treatment, were obtained clinically via oxygen mass balancing in the blood supplying the treated tissues. Possible rate limitation caused by diffusional resistance was then investigated using a mathematical model of the tissue system. Arterial and venous blood samples obtained before and after treatment provided the 'inlet' and 'outlet' oxygen concentrations for the balance. Blood flow rates were obtained from both literature values (plethysmographic measurement) and clinical velocity change measurements, using a pulsed Doppler meter. The results indicated that Fluidotherapy treatment at 46[degrees]C for 20 min-r utes slightly more than doubles (110% increase) the tissue metabolic rate. An unsteady state model of the capillary/tissue system was then used to predict tissue oxygen tension profiles during treatment. Solution by fourth order collocation technique indicated that the oxygen tension never falls below the critical level; i.e., the reaction rate remains relatively independent of oxygen concentration. The conclusion was therefore that diffusional resistance did not affect the observed metabolic rate increase. The absence of diffusional limitations dictated a treatment strategy of always using the highest treatment temperature possible. Estimation of individual tissue layer metabolic rates clearly showed that the treatment duration should be chosen according to the depth of the tissue layer being treated. Finally, the sizeable rate increase observed (more than double the normal rate) strongly indicated that the frequency of treatment should be increased from the standard one treatment per day.Item A reactor for chemical vapor deposition of silicon(1982) Guerner, Herve F.; Pollard, Richard; Henley, Ernest J.; Wolfe, John C.A rotating disk reactor has been designed and built to study chemical vapor deposition of silicon. The process chosen is reduction of silicon tetrachloride in excess hydrogen at high temperatures (900 - 1200[degrees]C) and atmospheric pressure. The silicon single crystal substrate on which deposition was obtained is rotated at a speed in the range 300 - 1200 rpm. The reactor is designed to minimize the effects of natural convection, radial diffusion, and reactant gas recirculation. The rotating susceptor, to which the silicon wafer is attached, is heated by induction and the substrate temperature is measured by optical pyrometry. Significant variations in nucleation and growth rates of silicon on a silicon substrate have been observed. Microstructures associated with epitaxial and polycrystalline growth were obtained at different radial positions on the same wafer. These experimental results can be attributed largely to temperature gradients across the wafer and to the presence of trace impurities, such as oxygen and carbon, on the wafer surface. A new reactor design is proposed which includes the modifications necessary to obtain uniform epitaxial growth of silicon with the rotating disk system.Item A revised approach to the theory of cake filtration(1979) Ville, Francois; Tiller, Frank M.; Crump, Joseph R.; Wheeler, Lewis T.Conventional constant pressure filtration theory has been revised to account for variation in average cake porosity and average filtration resistance as they change with pressure drop across compressible cakes. The revised theory points to the impossibility of obtaining medium and cake resistances with a single linear plot of instantaneous reciprocal rate dt/dv, or average reciproqual rate t/v, versus v. At the beginning of filtration, those plots are shown to have marked curvature which can easily be missed experimentally. They converge at t=0 to the medium resistance. For long filtrations, tangents to these curves tend to intersect at the same intercept, termed the "false medium resistance" which is approximately (1-n) Rm for moderately compressible materials. The intercepts approach zero for highly compressible cakes, giving the erroneous impression of a null Rm . In addition, this revised theory has been applied to constant rate and variable pressure - variable rate filtrations.Item A segregated model of the growth of Baker's yeast(1983) Hjortso, Martin A., 1954-; Luss, Dan; Amundson, Neal R.; Pollard, Richard; Donaghey, Charles E.; McInnis, Bayliss C.Increased activity in biotechnology has motivated new interest in better mathematical models of microbial growth. This work derives and solves segregated growth models for populations of the budding yeast Saccharomyces cerevisiae, or Baker's yeast, in different genetic and environmental contents. These models are an improvement over the classical nonsegregated models in two respects: first they predict the distribution of cells over the different parts of the cell cycle instead of just the bulk properties of the population, and, second, these models relate the bulk properties to kinetic events on the single-cell level, thus forming a connection between single cell kinetics and observable system behavior. The unequal division model of the Baker's yeast cell division cycle is used to formulate a population balance model for the growth of this organism in well-stirred batch and continuous flow reactors. It is assumed that all cell divisions and cell births occur at discrete points in the cell cycle. The population balance equations may then be reduced to sets of coupled linear first order differential equations that have been solved analytically. Steady-state frequency functions for cell mass and DNA content are derived under different assumptions concerning cell mass and DNA synthesis kinetics. Unequal division of budding yeast gives rise to a characteristic bimodal distribution of cell masses. The accumulation of bud scars on the cell surface, which eventually leads to cell death as a maximum number of scars accumulate, is introduced in two models: in the first it is assumed that the scars do not affect the single-cell mass growth rate, while in the second it is assumed that the growth rate decreases with the number of scars on the cell surface. It is found that scar accumulation can significantly alter the shape of the mass frequency function, even to the extent of removing the bimodality. The different models are combined with substrate (glucose) and product (ethanol) balance equations to determine the steady-state operating characteristics of a continuous-flow bioreactor. Again it is found that accumulation of bud scars can significantly alter the system behavior. A stepwise method for obtaining transient solutions to the population balance is described, and solutions are found for two special cases. The response of a yeast culture subjected to an abrupt environmental change, resulting in a step change in single cell growth rate, is obtained. It is shown how the response can yield information on single-cell mass synthesis kinetics and on the manner in which individual cells control certain parameters of the cell cycle. The return to steady state of a culture that is partly synchronized has been found and described in terms of the fraction of cells of different ages, as defined by the number of bud scars on the cell. When the cell division is unequal the response is oscillatory, while this is not the case for symmetric division. The population balance model is used to investigate plasmid stability in steady state growth with selection pressure. Plasmids are assumed to be partitioned at random at cell division, allowing for the possibility of cells being born without plasmids. Two models of plasmid replication regulation are used: the first assumes that, independent of the number of plasmids in the newborn cell, each dividing cell has the same fixed number of plasmids. The second model assumes that the number of plasmids synthesized in each cell cycle is a constant. The results of this analysis show that the two models can give qualitatively very different copy number distribution (bimodal versus unimodal). It is also found that the average plasmid content is significantly more growth-rate dependent for the first model than for the second. This result holds if the probability of a plasmid moving into the bud is presumed constant or is assumed to increase with the bud size. The loss of unstable plasmid from a population following a shift to nonselective medium is investigated. Differences between the predictions of the population balance model and simpler nonsegregated models increase as cell division becomes more unequal. This occurs at lower overall growth rates of the yeast population. A central parameter in the model of populations with unstable plasmids is the probability that a cell will be born without any plasmids. This parameter is determined experimentally by isolating subpopulations of newborn cells and cells just about to divide using a flow cytometer with a cell sorter and testing these cells for plasmid content by plating them out on selective and nonselective media.Item A selective catalyst for the production of LPG from synthesis gas(1977) Santner, Chris R.; Richardson, James T.; Motard, Rodolphe L.; Howell, John ReidA study of the production of liquified petroleum gas (LPG) from the hydrogenation of CO over nickel catalysts is reported. Coprecipitated nickel alumina catalysts reduced between 300 [degrees] C and 400 [degrees] C are shown to have a high selectivity to LPG. Reactor temperatures of 275 [degrees] C to 310 [degrees] C give satisfactory conversion levels without sacrificing LPG yields. High contact times and a H2/CO ratio of two appears to be a good compromise between conversion, LPG production, and catalyst fouling. Magnetic studies to determine crystallite size distributions are reported. Lifetime studies for a nickel alumina catalyst (Catalyst A) show rapid deactivation initially down to 40% conversion where further deactivation is slow. Reduction temperature has no effect on this. Regeneration is possible with no loss of activity or selectivity. Temperature programming to maintain conversion is also possible with little sacrifice of selectivity. [...]