Statistical modelling and optimization of bacterial cellulose formation by Enterobacter roggenkampii IITISM CP-1 in sweet lime extract media through response surface methodology

Background Cardiovascular diseases are the leading cause of death throughout the developed world, attributed to approximately 17.8 million deaths worldwide in 2017 with increasing prevalence due to the aging population. Cardiovascular diseases generally result in heart failure. While the best treatment option for heart failure patients is heart transplantation, there is a severe deficiency in the availability of donor hearts. Rotary Blood Pumps (RBPs) utilised as Ventricular Assist Devices (VADs) provide an alternative treatment option. These devices are small implantable pumps that support the failing heart by providing power to augment circulation. The development of RBPs generally begins with initial designs obtained using traditional pump design methods (such as that developed by Stepanoff). However, studies have shown that this approach produces RBP prototypes far from optimal in design. Traditional theory relies on design constants derived empirically for large industrial pumps and these do not scale down well when applied to the much smaller RBPs. The initial designs are therefore generally quite poor and require an iterative build-and-test approach to obtain suitable pump prototypes – a process that is expensive and time consuming. Therefore, by improving the methodology for obtaining initial designs to better reflect the final product, development time can be greatly reduced. A popular avenue for analysing the effect of design variations and to further develop early prototypes of RBPs is to employ Computational Fluid Dynamics (CFD) simulations. These numerical simulations provide detailed data regarding the flow fields within these devices. However, a range of simulation options is available, leading to a wide range of potential predictions. In an attempt to provide a benchmark case, the FDA presented a challenge in which a pump design and test conditions were defined, allowing for direct comparison amongst different simulation approaches from a number of labs/RBP developers. The purpose of this thesis was to produce a gross design tool to provide a good starting point in RBP prototyping and a CFD simulation approach for verification that can also be used as a design refinement tool. Methods Formulating a design method for pumps requires the generation of empirical data. A number of pump design variables was identified as having an impact on pump performance, and a large number of experimental tests would have been needed to test the influence of each. Instead, a Design of Experiments (DOE) was utilised to streamline the process. The DOE outputs a relatively small number of tests required to fit a statistical model. Each design specified by the DOE was examined experimentally using a custom-built automated pump test platform to generate a number of performance measures. The obtained results were used to formulate a Response Surface Method (RSM) statistical model that showed acceptable fit to the input data. Coupled with desirability functions, the RSM model allowed for design optimisation. This tool essentially replaces Stepanoff’s traditional design methodology. The RSM model provides a robust tool that allows the user flexibility in design optimisation goals. The FDA pump was investigated in this thesis and a wide variety of simulation approaches was examined to determine which was most accurate. A range of factors were considered which included: mesh density, interface position between the rotating and stationary zones, steady vs. transient simulations, discretisation schemes, time step size and choice of turbulence model. The most appropriate option from each investigative study was selected to determine a recommended simulation approach. Final simulations were performed using these recommendations and were compared to the FDA experimental results to confirm the suitability of the suggested settings. Determination of Centrifugal Blood Pump Characteristics using CFD and Experimental Analysis iii The statistical model developed was used to design two different impellers as validation test cases. The first impeller was designed to optimise the maximum efficiency, P – Q curve slope and efficiency consistency. The second impeller was designed to mimic the approach used in traditional design methods for RBPs in setting a target design point as the primary objective and the aforementioned factors (from the first impeller) as secondary objectives. These two case studies underwent statistical performance predictions, CFD simulations, PIV analysis and experimental hydraulic testing to validate the statistical and CFD models. Results From the initial CFD study, a hybrid SBES turbulence model with full transient simulation on a fine grid with small time steps proved to be the most suitable both in terms of pressure rise generated by the FDA pump and resulting velocity fields when compared to published experimental results. From these findings the CFD modelling strategy was established. CFD results for the two validation pumps showed pressure rises matching the experimental data (8% and 1% difference for each impeller) within an acceptable range (<10% from the mean). The simulated velocity fields also closely replicated the PIV data for the majority of the flow domain. The statistical performance predictions well reflected those measured experimentally with the majority of data points falling within its confidence intervals. The hydraulic results also supported the main goal of this thesis, whereby an impeller generated using the statistical model, operated far closer to the target design point than that of a blood pump designed following Stepanoff’s methodology. Overall, both the statistical model and CFD approach provided accurate predictions and the purpose of the thesis was achieved. Final Remarks The statistical and CFD models developed in this thesis yield an effective design tool and verification methodology and show improvement over the current traditional design methods and accuracy in simulated results. Ultimately, the utilisation of these tools will lead to a reduction in the development time for new RBPs and provide a good understanding of the flow dynamics within these pumps, leading to improved pump designs reaching patients sooner. These tools are readily generalizable and could be adopted as design tools now.

Statistical optimization for maximum production of a hyperthermostable, Ca2+-independent and high maltose-forming alpha-amylase by Geobacillus thermoleovorans. G. thermoleovorans was cultivated in 250 ml flasks containing 50 ml of chemically defined glucose-arginine medium (g l(-1): glucose 20; arginine 1.2; riboflavin 150 microg ml(-1); MgSO4. 7H2O 0.2; NaCl 1.0; pH 7.0). The medium was inoculated with 5 h-old bacterial inoculum (1.8x10(8) CFU ml(-1)), and incubated in an incubator shaker at 70 degrees C for 12 h at 200 rev min(-1). The fermentation variables optimized by 'one variable at a time' approach were further optimized by response surface methodology (RSM). The statistical model was obtained using central composite design (CCD) with three variables: glucose, riboflavin and inoculum density. An over all 24 and 70% increase in enzyme production was attained in shake flasks and fermenter because of optimization by RSM, respectively. A good coverage of interactions could also be explained by RSM. The end products of the action of alpha-amylase on starch were maltose (62%), maltotriose (31%) and malto-oligosaccharides (7%). RSM allowed optimization of medium components and cultural parameters for attaining high yields of alpha-amylase, and further, a good coverage of interactions could be explained. The yield of maltose was higher than maltotriose and malto-oligosaccharides in the starch hydrolysate. By applying RSM, critical fermentation variables were optimized rapidly. The starch hydrolysate contained a high proportion of maltose, and therefore, the enzyme can find application in starch saccharification process for the manufacture of high maltose syrups. The use of this enzyme in starch saccharification eliminates the addition of Ca2+.

The three‐dimensional (3D) printed poly lactic acid (PLA) bone plates lack mechanical strength, resulting in premature failure. Coating these plates with polydopamine (PDM) forms covalent bonds with the PLA molecular structure, enhancing their mechanical properties. The mechanical strength of the coated bone plates is influenced by infill density, submersion time, shaker speed, and coating solution concentration. However, conducting experiments for each parameter value to achieve maximum biomechanical tensile strength (BTS) and biomechanical flexural strength (BFS) is time‐consuming and costly. Overall, the combination of response surface methodology (RSM) and machine learning (ML) enables determination of the best printing parameters, leading to reduced material waste, personalized bone plates tailored to individual anatomy, improved implant fit, and functionality. Moreover, this approach has the potential to reduce the need for additional surgeries and overall costs. To optimize coating parameters, this study employs RSM and ML techniques, including genetic algorithm (GA), particle swarm optimization (PSO), random search optimization (RSO), and differential evolution (DE). Experimental validation of the optimized process parameters and their corresponding fitness values is carried out using both RSM and ML approaches. The results demonstrate that GA has the closest relationship between experimental and fitness values, followed by DE, RSM, PSO, and RSO. Highlights Direct immersion coating of polydopamine on 3D printed PLA bone plates. Evaluating mechanical strength for bone plates coated at varying parameters. Statistical modeling and optimization of mechanical strength using RSM. Mechanical strength optimization and convergence properties for ML models. Experimental validation of RSM and ML‐based optimization algorithms.

Response surface methodology was used to optimize lincomycin production by Streptomyces lincolnensis NRRL ISP-5355 in submerged fermentation. Screening of fermentation medium components to find their relative effect on lincomycin production was done using Plackett-Burman design. Malt extract, dextrin, soluble starch and (NH4)2SO4 were the most significant nutrient influenced on lincomycin production. Central composite design was applied to determine optimal concentrations of these factors and the effect of their mutual interactions. The interaction between soluble starch and (NH4)2SO4 was found to enhance the production, whereas malt extract and dextrin exhibited an influence independent from the other two factors. Using this statistical optimization method, maximum lincomycin concentration of 1345 μg/ml was obtained which represented a 40.5 % increase in titer than that acquired from the non-optimized medium. This statistically optimized medium was employed for lincomycin production through immobilization of Streptomyces lincolnensis by adsorption on synthetic cotton fibers. Immobilization technique improved the concentration to 1350 μg/ml higher than that produced from free cells cultures and could be maintained for longer than 17 days in a repeated batch system.

Response surface methodology was used to optimise toluene degradation by immobilised Brevibacillus parabrevis ATHH40. The pH, temperature and toluene concentration were the most significant factors in toluene degradation. Central composite design was applied to determine optimal concentrations of these factors and the effect of their mutual interactions. The interaction between pH and temperature with toluene concentration was found to enhance toluene degradation, whereas pH, temperature and toluene concentration exhibited an independent influence. Using statistical optimisation, a maximum toluene degradation was achieved when pH, temperature and toluene concentration were adjusted to 6.93, 34.76°C and 774.76 mg L−1. This optimised medium was employed for toluene degradation by Brevibacillus parabrevis immobilised by multi-walled carbon nanotubes that improved toluene degradation to 93.87%, much higher than achieved by cultures of free cells.

In this study, the statistical modeling and optimization of the regeneration of spent bleaching earth (SBE) for re-use in the bleaching of crude palm oil (CPO) oil was examined. Having a good model will assist with the successful optimal regeneration of SBE and hence minimize the environmental pollution associated with its current disposal method which is based on dumping as landfills. The SBE samples were de-oiled with the Soxhlet extraction method, using n-hexane for 1 h at 60 °C; treated at temperatures ranging from 300–500 °C; at carbonization time between 30 and 45 min; and with hydrochloric acid concentrations between 1 and 2 M, at a constant stirring time of 30 min, respectively. The operating conditions for the experiment were according to the Central Composite Design (CCD) experimental design using the Design Expert software version 13. The modeling and optimization of the SBE regeneration process was carried out with the Response Surface Methodology (RSM) and Artificial Neural Network (ANN) techniques. Five regression models were developed from the RSM approach and the best one selected based on model selection parameters recommended in the literature. Similarly, ten ANN models with the number of neurons in the hidden layer that varied from 2 to 16 were considered and the best one selected using the mean square error (MSE) and correlation coefficients (R) for the training, validation and testing performances. Results showed that the ANN technique led to a model with a better predictive ability than the RSM one. The optimum experimental bleachability of 71.5% for the regenerated de-oiled SBE was obtained at carbonization temperature of 500 °C, hydrochloric acid concentration of 2M and carbonization time of 45min. Using the Genetic Algorithm (GA), the ANN model resulted in an optimum bleachability of 70.87% with corresponding optimum factors at 468.19 °C, 2 M and 45 min, while the RSM approach gave an optimum bleachability of 73.52% at the corresponding factors of 498.99 °C, 1.57 M and 41.14 min for the carbonization temperature, acid concentration and carbonization time, respectively. The optimum experimental bleachability of the regenerated SBE achieved was 12.5% higher than that of virgin bleaching earth (VBE).

Additive manufacturing of thermoplastic elastomers is challenging using fused deposition modeling due to their high melt viscosity, low column strength, and poor fusion among layers. Solvent‐cast 3D printing (SC‐3DP) is an efficient alternative to successfully 3D print such materials. However, selection of suitable 3D printing parameters is crucial to realize parts with optimum physicomechanical properties. In this work, statistical modeling and SC‐3DP parameter optimization for styrene–ethylene–butylene–styrene (SEBS) block copolymer were performed. The effect of 3D printing process parameters on shrinkage, relative density, and tensile strength was analyzed using response surface methodology. Experiments were planned as per central composite design and analysis of variance was performed to evaluate the significant parameters. SEBS content in the polymer solution significantly affected the shrinkage of SC‐3DP samples. Moreover, relative density and tensile strength were significantly affected by print speed and layer height. A significant interaction between print speed and layer height was also noticed for tensile strength and relative density of printed samples. Multi‐objective optimization using genetic algorithm was also performed to minimize shrinkage and maximize relative density and tensile strength. Finally, a case study was conducted comparing the physicomechanical properties of SC‐3DP samples printed at optimized process parameters and compression molded samples. Highlights Statistical models were developed using response surface methodology. Genetic algorithm based multi‐objective optimization was performed. Optimum solvent‐cast 3D printing (SC‐3DP) process parameters were determined.

Statistical modeling and optimization of itaconic acid reactive extraction using response surface methodology (RSM) and artificial neural network (ANN)

Pseudomonas sp. SUK producing an extracellular collagenolytic protease was isolated from soil samples from meat and poultry industrial area based in Kolhapur, India. Response surface methodology was employed for the optimization of different nutritional parameters influencing production of collagenolytic protease by newly isolated Pseudomonas sp. SUK in submerged fermentation. Initial screening of production parameters was performed using Plackett-Burman design and the variables with statistically significant effects on collagenolytic protease production were identified as gelatin, peptone, and K2HPO4. Further, optimization by response surface methodology (RSM) using Central Composite Design showed optimum production of collagenolytic protease with 12.05 g L−1 of gelatin, 12.26 g L−1 of peptone and 1.29 g L−1 of K2HPO4. Collagenolytic protease production obtained experimentally has very close agreement with the model prediction value and the model was proven to be adequate. The statistical optimization by response surface methodology upsurges collagenolytic protease yield by 2.9 fold, hence the experimental design is effective towards process optimization. Moreover, ammonium sulphate precipitated, partially purified enzyme has shown to cleave collagen from bovine achilles tendon, which was observed by phase contrast microscopy, and SDS-PAGE. Hence, extracellular collagenolytic protease of Pseudomonas sp. SUK could have considerable potential for industrial as well as medical applications.

Statistical modeling and mix design optimization of fly ash based engineered geopolymer composite using response surface methodology

A protease producing Bacillus sp. GA CAS10 was isolated from ascidian Phallusia arabica , Tuticorin, Southeast coast of India. Response surface methodology was employed for the optimization of different nutritional and physical factors for the production of protease. Plackett-Burman method was applied to identify important factors (anchovy waste, KH 2 PO 4 , NaCl and temperature) influencing protease production. Further optimization was done by response surface methodology using central composite design. Under the optimized conditions by central composite design, the protease experimental yield (842.102 U/ml) closely matched the predicted yield by the statistical model (830.307 U/ml) with R 2 = 99.94%. The time course of protease production was increased using the RSM optimized medium (856.29 U/ml) (anchovy waste 20.50 g/l, KH 2 PO 4 3.06 g/l, NaCl 42.91 g/l, temperature 43.36°C, 42 h and pH 9) compared with the un-optimized basal medium (267.33 U/ml). The improvement of protease production by microbial conversion of anchovy waste suggested its potential utilization to generate high value added products using cheap carbon and nitrogen substrates. Keywords: Protease, anchovy waste, statistical optimization, ascidian associated bacteria. African Journal of Biotechnology Vol 13(27) 2741-2749

ABSTRACTIn this work, experimental and statistical modeling and optimization of process parameters for maximizing the o/w emulsion stability was carried out using the multiobjective artificial neural network-genetic algorithm (ANN-GA) coupled with response surface methodology. The independent model constrains were oil concentration (10–50% v/v), surfactant concentration (2–10% v/v), stirring intensity (2000–6000 rpm), stirring time (5–20 min), and pH (2–12). The responses were turbidity (τ) and emulsion stability index (ESI24). This fact that there is a reasonably good agreement between the experimental data and the predicted values was shown by the modeling results. The optimized conditions predicted by hybrid ANN-GA model to maximize ESI24 ( = 94.71) with 4.8% error were: oil concentration 50% v/v, surfactant concentration 5.571% v/v, stirring speed 6000 rpm, stirring time 5.97 min, and pH 12. The accuracy of the model is confirmed by the comparison between the predicted and experimental data. The proposed hybrid ANN-GA model was found to be useful for the modeling and optimization of process parameters for emulsion stability analysis and the other emulsification process.

The article captures a brief state-of-the-art encompassing statistical modelling and optimization of bioenergy production from oil seeds. Exploring the research space of bioenergy production from biomass oil, various numerical approaches have been employed to model and optimize the process parameters for maximum response yields. Various performance studies have also been carried out to evaluate the predictive capabilities of emerging Artificial intelligence (AI) algorithms on modelling of bioenergy production from seed oil using the conventional Response Surface Methodology (RSM) as a reference base. For the records, the precision of measurement, management of uncertainties, more accurate data analysis and prediction are techniques which these methodologies have the capacity to do. The commonly used techniques for optimization and modelling studies of bioenergy from biomass oil and the analysis of their usage according to their performance metrics are detailed in the body of this work. Aside the relative limitation of RSM models in highly non-linear processes as compared to the robustness of AI models, RSM models still continue to hover on large scale applications when it comes to modelling and optimization studies on bioenergy production from oil-rich plant seeds. In the era of big data analysis in relation to the test and measure of analytical performances and prediction ability, AI models have begun to gain attention in the last two decades owing to their better estimation capabilities and ability to handle more data points in real-time prediction of bioenergy production. This study has been able to show successful applications of both RSM and AI models in bioenergy field, with a pointer to further adoption of the later in more related studies as a result of its relative advantages.

In this study, the response surface methodology (RSM) and artificial neural network (ANN) were employed to study the adsorption process of 2,4-dichlorophenoxyacetic acid (2,4-D) by using modified hydrogel, i.e., activated carbon poly(dimethylaminoethyl methacrylate) (AC/PDMAEMA hydrogel). The effect of pH, the initial concentration of 2,4-D and the activated carbon content on the removal of 2,4-D and adsorption capacity were investigated through the face-centered composite design (FCCD), optimal design and two-level factorial design. The response surface plot suggested that higher removal of 2,4-D and adsorption capacity could be achieved at the higher initial concentration of 2,4-D and lower pH and activated carbon content. The modeling and optimization for the adsorption process of 2,4-D were also carried out by different design methods of RSM and different training methods of ANN. It was found that among the three design methods of RSM, the optimal design has the highest accuracy for the prediction of 2,4-D removal and adsorption capacity (R2 = 0.9958 and R2 = 0.9998, respectively). The numerical optimization of the optimal design found that the maximum removal of 2,4-D and adsorption capacity of 65.01% and 65.29 mg/g, respectively, were obtained at a pH of 3, initial concentration of 2,4-D of 94.52 mg/L and 2.5 wt% of activated carbon. Apart from the optimization of process parameters, the neural network architecture was also optimized by trial and error with different numbers of hidden neurons in the layers to obtain the best performance of the response. The optimization of the neural network was performed with different training methods. It was found that among the three training methods of the ANN model, the Bayesian Regularization method had the highest R2 and lowest mean square error (MSE) with the optimum network architecture of 3:9:2. The optimum condition obtained from RSM was also simulated with the optimized neural network architecture to validate the responses and adequacy of the RSM model.

Guerbet alcohol (GA) is β-branched primary alcohol having excellent physiochemical properties like lower pour point (PP) and higher kinematic viscosity (KV) in comparison to linear alcohol. Although different aspects of the synthesis of GA, such as methods of synthesis, catalytic systems, and reaction conditions, have been studied, but statistical modeling and optimization of the synthesis of GA have not been carried out. In the present work, the optimization of the synthesis of GA using a mixture of lauryl and myristyl alcohol was carried out with the aid of response surface methodology (RSM) considering the conversion of the reaction, PP and KV at 40˚C & 100˚C as dependent variables. The optimal reaction conditions were temperature, pressure, and time of 220˚C, 300 mbar, and 10 hours respectively. The optimum conversion was 99.141%, including dimer yield of 81.755%, PP of -3˚C, KV at 40˚C & 100˚C of 34.12 cSt & 7.22 cSt, respectively. The results obtained by the RSM were then authenticated, applying artificial neural networks (ANN) generated with the help of MATLAB. The ability of the generated model to predict the response variables was validated by less than 5% error for almost all the models, confirming the statistical significance. Also, the tribological potential for linear Ginol-12,14 (FA) and synthesized branched GA as lubricant additive was evaluated by determining its physiochemical, thermal and tribological properties.