Yun Li

Numerical Predictions of Mild-Steel Corrosion in a H2S Aqueous Environment Without Protective Product Layers

The corrosion of mild carbon steel in a H2S aqueous environment without corrosion product layers is predicted using the computational fluid dynamics method. The Abe–Kondoh–Nagano low-Reynolds-numbe...

A Study on the Corrosion Behavior of Carbon Steel Exposed to a H 2 S-Containing NH 4 Cl Medium

NH4Cl corrosion failure often occurs in the overhead systems of hydrotreaters, and this failure is always accompanied by the appearance of H2S. A combination of electrochemical and surface spectroscopic (SEM/EDS, AFM, XRD) techniques was used to investigate the effect of different factors, including the surface roughness, temperature, dissolved oxygen, pH and H2S concentration, on the corrosion behavior of carbon steel in an NH4Cl environment with the presence of H2S. The effect of H2S concentrations (at the ppm level) on the corrosion behavior of carbon steel was systematically revealed. The experimental results clearly indicated that the corrosion rate reached a minimum value at 10 ppm H2S. The steel surface was covered by a uniform corrosion product film in a 10 ppm H2S environment, and the corrosion product film was tight and protective. The ammonia from NH4Cl helped maintaining the protectiveness of the corrosion films in this environment. Dissolved oxygen mainly accelerated the cathodic reaction. The cathodic limiting current density increased with increasing temperature, and the anodic branch polarization curves were similar at different temperatures. The anodic current density decreased as the pH decreased, and the cathodic current density increased as the pH decreased. The absolute surface roughness (Ra) of carbon steel increased from 132.856 nm at 72 h to 153.973 nm at 144 h, and the rougher surface resulted in a higher corrosion rate. The critical innovation in this research was that multiple influential factors were revealed in the NH4Cl environment with the presence of H2S.

FEM Analysis and Preload Evaluation of Bolted Joints Subject to Low Temperature

Bolted joints are broadly used in various industrial products. Especially in chemical engineering, the reliability of bolted joints can have a significant influence on the safety of chemical systems. In some fields, such as liquefied natural gas (LNG) industry, equipment usually works in low temperature (−162°C). Different materials with various thermal expansion rates are usually used in bolted joints; Therefore the preload and stress distribution of bolted joints in low temperature can change apparently due to the different axial and radial thermal deformation. If the preload design in normal temperature is inappropriate, the bolted joints may either encounter relaxation or over-tightening in working temperature. In this study, a theoretical analysis is proposed in order to evaluate the appropriate preload selection of bolted joints work in low temperature. FE analysis is made to examine the accuracy of the theoretical formula. The result shows that the error is less than 10% in most cases. The stress distribution on the bolt thread region is studied and the result shows that the maximum tensile stress on the bolt thread region is much higher than in non-thread region. Availability and safety should be both considered in preload selection of bolted joints working in low temperature. Finally the effect of radial thermal deformation difference is discussed when the thermal expansion rate difference is high. The effect of radial thermal deformation difference should not be neglected in some cases.Copyright

A Universal Semi-Analytic Model for Axial Mixing in a Straight Pipe

The industrial applications of common pipelines to simultaneously deliver multiple types of gases or liquids widely occur in petroleum and natural gas industries. Therein, the pipelines compose of significant horizontal and vertical sections can range from several meters to several km in length. The longitudinal dispersion of matter is an important aspect of such flows in safety and reliability. This paper sets out an analytic simulation model on the basis of the profiles of velocity and turbulent viscosity via steady numerical solutions by computational fluid dynamics (CFD) method. The mixing length and the wasting lengths of the original and latter fluids were applied as the key parameter to describe the extent of axial mixing in the pipe. In both turbulent and laminar pipe flows, the relative errors in the present model were reduced to 5% or less, which is 5%–15% lower than those in previous models. Moreover, the accuracy of the present model was nearly not less than that of unsteady numerical solutions in the same meshing density, whereas the computational load of the present model was far less than that of unsteady numerical solutions. Further, through investigating effects of buoyancy on axial mixing, it is found that the wasting length of the lighter fluid was higher than that of the heavier fluid.Copyright

Comparison of various sound models for predicting transmission loss of compressor mufflers based on the modal series expansion method

With the rising public awareness and concern for noise control, properly designed mufflers for compressors are now becoming more and more important. Reactive mufflers are commonly used in refrigerating and air conditioning compressors to reduce the gas pulsation generated by the suction and discharge systems. Due to the size limitation and expense of these mufflers, it would be beneficial to predict the transmission loss (TL) characteristics at the design stage. To accomplish this, many methods have been studied, including plane wave method, finite element method (FEM), boundary element method (BEM), and modal series expansion method (MSEM). Three MSEM sound models, that is, point model, surface model, and line model, have been compared on the basis of some criteria, such as accuracy and computation time, in this article. A newly revised point model verified by 3D FEM has been proposed to overcome singularity of a previous point model. The computational accuracy and efficiency of the surface model, the circular line model, and the revised point model has been compared. It is found that the proposed point model not only has higher computational efficiency than that of the circular line and the surface models but also has high accuracy.