This paper reports a comparison of two methods to perform residual stress measurements. The specimens tested by each method were two blades from a shrouded centrifugal compressor impeller. The first method is the conventional hole drilling strain gage method which was used to predict residual stresses across the blade surface. The residual stresses are released by drilling a hole in the blade. The second method is called the nonlinear harmonic (NLH) scanning method and is based on the principal that the magnetic domains of ferrous materials vary in a non-linear way relative to internal stress. The effects of residual stress may be either helpful or harmful, depending on the magnitude of the residual with respect to the operating stresses. If not adequately relieved by heat treatment, residual tensile stress that develops in the welding process of shrouded impellers, will add to the stress developed by rotation which moves the point to the right on the Goodman diagram and reduces allowable alternating stress. The results showed comparable residual stress measurements of the NLH method compared to the conventional hole drilling method.

In the current emerging development of technologies for heavy crude oil pumping, it is essential to be able to predict and verify the reliable operation of new pumping technologies. Engineering studies are required for verification of the reliable operation of crude oil pumps and pumping systems. The specific requirements to be met by the pump and pumping system for ensuring reliable operation vary depending on their characteristics. However, it has been commonly based on the author’s experience to perform several analyses specifically for heavy crude oil applications including positive displacement pumps. Such analyses include skid mechanical vibration study, coupled acoustic-mechanical piping analysis, and lateral and torsional rotordynamic study. This paper presents a case study where the previously mentioned studies are conducted for a gas turbine driven screw pump train. A lateral rotordynamic analysis has been conducted by the turbomachinery original equipment manufacturer (OEM) and its details are not included in this paper. The results from the skid mechanical vibration study confirm that vibration issues are not predicted during normal or transient operation. A coupled acoustic-mechanical analysis identifies areas of concern that can then be addressed to avoid excessive vibration levels on the piping. Finally, results from a torsional rotordynamic study verify that vibration levels are expected to be within allowable limits.

The current study uses a transient computational fluid dynamic (CFD) analysis to predict the rotordynamic force coefficients for an unshrouded radial inflow expander and labyrinth wheel seal in a high pressure natural gas turboexpander. Direct and cross-coupled terms are predicted for both the stiffness and damping for the expander wheel and labyrinth seal, including the aerodynamic cross-coupling forces or cross-coupled stiffness coefficients that impact rotordynamic stability. A static eccentric method was applied where a transient simulation where the rotor portion is displaced with a known static eccentricity and the resulting forces allow for the calculation of direct stiffness and cross-coupled stiffness. In addition, a whirling method using transient simulations at various whirling frequencies is applied to solve for the full set of rotordynamic coefficients, including direct and cross-coupled stiffness and damping terms through curve fitting a KCM model for both the expander wheel and labyrinth seal. Wheel force coefficients are presented for several operating conditions, including low, nominal, and high flow regimes. In addition, a parametric study is performed, including hypothetical cases to evaluate sensitivity of wheel tip clearance, rotational speed, gas density on the expander cross-coupled stiffness. The predicted wheel cross-coupled stiffness shows a very strong dependence on the wheel tip clearance compared to the other parameters studied.

The current work shows the implementation of a high-speed turbomachinery test rig to measure vibration and internal shaft temperature differentials at a journal bearing for a rotor system designed to induce the Morton effect rotordynamic phenomenon. The vibration and shaft temperature measurements are compared to predictions using an analytical code described by Tong and Palazzolo [6,7]. An existing high-speed test rig was adapted, including a new rotor with six equally-spaced RTDs embedded at the journal bearing centerline. The rotor configuration included an overhung rotor design with a 58 mm (2.3 inch) diameter, 5-pad tilting-pad journal bearing. The in-rotor temperature measurements were conditioned using a custom on-board amplifier and extracted with a high-speed commercial slip ring. Test conditions included various levels of unbalance, bearing oil inlet temperature, and operating speed. Test measurements show that the temperature differential across the shaft is dependent upon operating speed, as well as vibration amplitude. Operating conditions included rotational speeds up to 19.5 krpm and vibration levels approaching the magnitude of the bearing clearance. Testing near the rotor first lateral natural frequency (or critical speed) with a high level of initial unbalance resulted in the highest temperature differentials across the shaft of approximately 11 °C (20 °F) plus. Vibration measurements show hysteresis in the synchronous vibration response in the Bode and polar plots. This measured vibration hysteresis is consistent with the rotor hot spot or temperature differential changing the unbalance level of the rotor. Overall, both the test measurements and predictions show notable temperature differentials and hysteresis behavior in the vibration response that are believed to be associated with the Morton Effect. These conditions are considered precursors to the spiral vibration or fully developed synchronous instability typically associated with the Morton Effect.

In order to reduce the amount of carbon dioxide (CO2) greenhouse gases released into the atmosphere, significant progress has been made in developing technology to sequester CO2 from power plants and other major producers of greenhouse gas emissions. The compression of the captured carbon dioxide stream requires a sizeable amount of power, which impacts plant availability, capital expenditures and operational cost. Preliminary analysis has estimated that the CO2 compression process reduces the plant efficiency by 8% to 12% for a typical power plant. The goal of the present research is to reduce this penalty through development of novel compression and pumping processes. The research supports the U.S. Department of Energy (DOE) National Energy Technology Laboratory (NETL) objectives of reducing the energy requirements for carbon capture and sequestration in electrical power production. The primary objective of this study is to boost the pressure of CO2 to pipeline pressures with the minimal amount of energy required. Previous thermodynamic analysis identified optimum processes for pressure rise in both liquid and gaseous states. At elevated pressures, CO2 assumes a liquid state at moderate temperatures. This liquefaction can be achieved through commercially available refrigeration schemes. However, liquid CO2 turbopumps of the size and pressure needed for a typical power plant were not available. This paper describes the design, construction, and qualification testing of a 150 bar cryogenic turbopump. Unique characteristics of liquid CO2 will be discussed.