This paper is an interdisciplinary study. The core idea is to observe the combustion process by means of nondestructive testing. A new combustion chamber detection method is explored through low cost and non-polluting nuclear imaging. In this research, the fluid imaging of a combustion chamber was obtained by Positron Emission Tomography (PET). In experiment, the chamber is constructed in PET scanner. The relevant engineering details and operations are described. The kerosene was labeled by ¹⁸F tracer. The PET detected the entire combustion flow field of the chamber. In results and analysis, this paper showed the PET images and used Computational Fluid Dynamics (CFD) to analyze the causes of feature formation in PET images. By Correlation calculation between PET images and various parameter distributions of burner fluid, the correlation results showed that the image features had a close relationship with Turbulent Viscosity (TV) distribution. Through this experiment, PET imaging is proved to be feasible and effective to trace diffusion flame fluid.

The bottom structures in shaking bioreactors have great influence on the flow field characteristics. In this work, four kinds of bottom structures, such as conical bottom, ellipsoid bottom, flat bottom, and sphere bottom are presented, and the mixing, dissolved oxygen and shear strain rate of bioreactors with different bottom structures are simulated by means of CFD. The simulated free surface shapes were verified by the observed images in video. The results showed that turbulent kinetic energy were 0.0014 m²·s⁻², 0.0016 m²·s⁻², 0.0012 m²·s⁻² and 0.0011 m²·s⁻², and the values of k_La are 2.6 h⁻¹, 2.9 h⁻¹, 2.1 h⁻¹ and 2.0 h⁻¹ for the conical bottom, ellipsoid bottom, flat bottom and sphere bottoms, respectively. This indicates that the mixing and dissolved oxygen characteristics of the bioreactors with the ellipsoid and conical bottoms were superior to the flat and sphere bottoms. The shear strain rates in the bioreactors with different bottoms were mainly in the range of 8–10 s⁻¹. The highest average shear strain rate was found in the bioreactor with the flat bottom, and it was still in the low shear strain rate range and suitable for the growth of shear-sensitive cells.

A numerical method based on cut-cell method was extended for turbulent flows in this study. Computational studies with modeling near wall boundary layer have emerged as the one of practical method in the industries. Most of merits of the method is due to a serious reduction of computational costs. A cut-cell method can obtain better spatial resolution because: (1) an obliquity of walls are captured and (2) conservations of mass and momentum are easier of attainment for each cross section. Conventional cut-cell methods focused on flux control across on the computational cells. In a proposed method, we changed the calculation procedure of stress tensor directly by using the distance from wall and velocity information. This extension made easy use of wall laws for laminar and turbulent boundary layers. The function can be selected from some proposed wall function model. In this study, the following two model were used: non-slip wall model and Allmaras wall model. As a result, the law of the wall model can be effectively incorporated on the Cartesian grid with high engineering applicability. In addition, higher efficient algorithm for orthogonal structured grid can be applied without serious reconstruction of programming source code. To demonstrate the method, (1) two-dimensional laminar flow in an inclined channel, (2) flow around a cylinder, and (3) turbulent flows over a wall-mounted hump were calculated. As a result of applications, the proposed method could easily take into account the flows with distorted boundary walls.

Finite Gaussian mixture model (GMM) has recently proven to be a powerful unsupervised treatment for monitoring nonlinear processes with multiple operating conditions. The performance of GMM-based monitoring method largely depends on the number of mixture densities. However, the popular penalty method, such as Bayesian information criterion (BIC) and Akaike’s information criterion (AIC), usually tend to yield noisy model size estimates. Moreover, the parameter estimates in GMM are susceptible to outliers. To overcome these deficiencies, this paper proposes a new process monitoring technique based on a robust infinite Gaussian mixture model (Ro-IGMM). Specifically, a separate weight at each point is assigned to the precisions as a measure of smoothness, representing the similarities to other data points. The Chinese restaurant process is then placed on a prior to turn into infinite groupings. The informations, such as a distribution over the number of clusters, the cluster assignments, and the parameters associated with each cluster, can be given by the posterior which is obtained by a collapse Markov chain Monte Carlo (MCMC) inference. Simulation results on the benchmark Tennessee Eastman process show that Ro-IGMM-based process monitoring method is more insensitive to outliers during process modeling, compared to traditional methods working with BIC model selection.

Fault diagnosis is a critical task in the daily operation of chemical processes. In this paper, a hybrid fault diagnosis method is proposed that combines a process-knowledge-based qualitative reasoning technique with fault detection based on a data-driven process-monitoring technique, without using any faulty datasets. Extended attributes, which are additional process feature variables generated from normal-operating-condition knowledge, are utilized to integrate the two techniques. The process qualitative reasoning model is simplified for combining these techniques and easing the modeling. This fault diagnosis method provides multiple reasoning routes for several potential fault root candidates. Each candidate and variable in its reasoning routes are weighted according to the results of the data-driven fault-detection method. Therefore, a priority list is presented to chemical engineers for further field examinations. The effectiveness of this method is validated using the Tennessee Eastman process, and novel diagnosis results are subsequently achieved.

We developed and evaluated an efficient method for the regeneration of activated carbon using dimethyl ether (DME), which is an organic compound that exists in gaseous state at room temperature and atmospheric pressures and liquefies above approximately 0.6 MPa at room temperatures. Liquefied DME dissolves both water and organic matter. When spent activated carbon adsorbs organic matter and water is immersed in liquefied DME, the organic matter and water are desorbed from the activated carbon and dissolved in the liquefied DME. Therefore, the spent activated carbon can be regenerated, and the activated carbon, organic matter, and water can be separated when the liquefied DME is depressurized. We developed and evaluated batch and flow methods for regenerating spent activated carbon polluted with acetic acid using DME. Using the batch method, the removal efficiency of the regenerated activated carbon for 1.0 wt% acetic acid solution was approximately 68%, whereas that of virgin activated carbon was 79%. The removal efficiency remained almost unchanged at 68% when activated carbon was repeatedly regenerated five times. We also conducted flow experiments to regenerate the spent activated carbon polluted with acetic acid. The removal efficiency of the regenerated activated carbon for 1.0 wt% acetic acid solution using the flow method improved with an increase in the contact time between the activated carbon and liquefied DME without being affected by the volume of the liquefied DME from 20 to 40 cm³. The removal efficiency of the regenerated activated carbon was approximately 80%, an improvement of 18% compared to that obtained using the batch method, which almost corresponds to that of the virgin activated carbon obtained using the batch method. Additionally, the removal efficiency remained almost unchanged at approximately 80% when the regeneration process was repeated five times.