Sensors and Instrumentation and Nondestructive Evaluation
Energy System Applications
DOE Office of Power Technology
NDE for Ceramics in Microturbines
The concept of distributed energy systems using small gas turbines (< 500 kWe) is being aggressively pursued by DOE. To obtain the desired efficiency target of >45%, these microturbines need reliable ceramic components in the combustor section. We are testing monolithic ceramic components for the radial flow rotors (mainly silicon nitride) and oxide ceramic composites for the combustor. An issue with the silicon nitride rotors is oxidation. As with nonoxide composites, an environmental barrier coating (EBC) is being developed to protect the rotors. Initial EBCs were made of tantalum oxide. Of concern with these coatings are spallation and erosive wear. We are developing NDE technology for assurance of low-cost, high-yield production of monolithic rotors, prediction of spallation of the EBCs, and characterization of the oxide composites. The NDE technologies under development include fast, high-spatial-resolution X-ray computed tomography for analyzing radial flow rotors, laser back scatter for determining the condition of the EBC coatings, and thermal imaging and air-coupled ultrasonic methods for determining the condition of the oxide composites.
NDE for Ceramics in Stationary Gas Turbines
As part of the DOE Office of Power Technology’s Program on Advanced Materials, Argonne is a subcontractor to Solar Turbines, Inc., which is developing and evaluating composite ceramics for application to stationary gas turbines. The Argonne effort is directed toward development of NDE methods for the ceramic materials being developed and tested for the Centaur H and Mercury gas turbine engines under development for stationary generation of electric power.
The evaluated ceramics include monolithic materials for turbine blades and nozzles and ceramic matrix composite (CMC) materials for combustor liners and transition ducts. The monolithic materials are Si3N4, and the CMC materials are both nonoxide (SiC/SiC) and oxide (Al2O3/Al2O3). The Argonne effort includes development of NDE methods to determine the thermal properties (thermal diffusivity) of the CMC materials and obtain full-field mapping of thermal properties, to detect delaminations of the CMC materials, and to detect surface cracks and subsurface cracks in the monolithic materials. We will also attempt to correlate the NDE data with predicted component performance.
Most recent work has focused on composite ceramics. The NDE method based on measurement of thermal diffusivity is now being used to accept/reject ceramic components for use in gas turbines. Full-sized liners (0.3 and 0.8 m in diameter and 0.2 m long) have been successfully analyzed with the thermal imaging method. Tests have been conducted before service and after over 14,000 h of engine test time.
In the thermal imaging efforts, we have acquired a very-high-speed thermal imaging camera and developed the capability for one-sided data acquisition for complex-shaped components. In addition, a new air-coupled ultrasonic system has been set up, and data have been correlated with the infrared image data. The air-coupled ultrasonic system operates at 400 kHz and has a fast scan speed (>10 cm/s). We have added the capability to handle large cylinders, 0.3 and 0.8 m in diameter, so that full-scale turbine engine components can now be examined. Results to date show excellent correlation to the thermal image data. Therefore, we can replace the previous X-ray computed tomography scan method for corroboration of data. However, X-ray computed tomography data are still obtained to establish through-wall information.
NDE to Detect Machining Damage in Monolithic Ceramics
As part of the DOE Heavy Vehicle Technology Materials Program, ceramic valves are being developed in a cooperative program with Caterpillar Technical Center. The program involves study of NDE methods to detect and quantify machining-induced damage on valves. The objective of this project is to develop NDE technology and methods that will determine the extent of machining-induced damage in monolithic ceramic materials, primarily Si3N4. In this work, the Caterpillar Technical Center is producing ceramic specimens under varying machining conditions, including wheel speed, grit type and size, grit-to-wheel bond materials, material removal rate, and downward force. The NDE method under development uses low-power lasers and special optical components in both the input and detector train. Data acquisition is totally automated. The detector train further consists of two specially fabricated optical detectors to provide data from two numerical apertures. Polarization of the reflected light is also accounted for in the detector train. The output from the two detectors is either summed or divided, and each resulting output is used to form an “image” by a raster scan motion of the part under study. The resulting “image,” a speckle pattern of gray levels, is then manipulated by digital image processing. Specifically, the image is viewed as a two-dimensional array of values from which statistical data are derived. In this case a standard deviation and a mean value are determined for each image. Dividing the standard deviation by the mean, we determine a coefficient of variation for each image, which represents some physical area on the specimen. Then, by plotting the coefficient of variation as a function of machining conditions, we obtain an extremely well-behaved correlation; Caterpillar is planning extensive further developments.
NDE for Thermal Barrier Coatings
Thermal barrier coatings (TBCs), usually made of yttria-stabilized zirconia (YSZ), are under development to protect the high-alloy, often single-crystal, materials used in the hot sections of gas turbines. These coating systems are being viewed in new turbine designs as so-called prime reliant. That is, the coating system itself is mandatory for turbine system operation at design conditions. The operating conditions in the turbine engine include thermal excursions, foreign objects passing through the gas path and causing damage, and corrosion conditions. Because of these severe conditions, there is a need for NDE technology that can allow turbine system manufacturers and users to assess the condition of as-coated components as well as to ascertain the condition of the coating during operation. Argonne is developing an NDE technique using elastic optical scattering, wherein the optical translucency of the YSZ is used. That is, a low-power laser is incident on the coating and the light, after penetrating the YSZ, reflects off the TBC/bond coat interface. The scattering of the reflected light is characterized by polarizing filters and special detectors.
This work is being conducted in conjunction with the University of Pittsburgh, University of Connecticut, Praxair Surface Technologies, Solar Turbines, GE Power Systems, Pratt & Whitney, and Siemens-Westinghouse. International collaborators include DHR, the German Aerospace Institute in Cologne, Germany. The figure below shows typical turbine blades and resulting laser scatter data suggesting a delamination. Of primary importance is the ability to predict when the TBC is going to spall.
The NDE work also focused on development of an acousto-ultrasonic method for determination of damage in hot gas filters as a function of axial position along their length. Both monolithic and oxide composite materials were studied. The work has been conducted with the University of Tampere (Finland), Schumacher Corp. (Germany), Southern Research Institute, Southern Company Services, and Oak Ridge National Laboratory. The work demonstrated that data from this NDE method can be directly correlated to the remaining strength of the hot gas filter material. The Argonne acousto-ultrasound method was automated and used to characterize a full-length (1.5 m) filter in under 5 minutes.
Advanced Sensors for Advanced Natural-Gas Reciprocating Engines
This program, completed in FY 2003, was supported by the DOE Office of Power Technology under the Advanced Reciprocating Engine Systems (ARES) project. The objective was to develop advanced sensors and control systems for real-time combustion monitoring of advanced natural-gas reciprocating engines. The proposed development includes sensors to measure NOx emissions and natural-gas composition and a feedback control system. Such sensors are needed to optimize engine combustion and reduce NOx emissions. Optimization can generally be achieved by adjusting the natural-gas/air ratio to control NOx emissions.
The targeted real-time NOx emission sensor must give a fast response and demonstrate high sensitivity to NOx emissions in the range of 1-100 ppm. The sensor must be able to function under an ARES engine exhaust environment that is typically at high temperature and contains more than 10% moisture. Conventional solid-state electrochemical sensors generally have slow response time and a narrow working temperature range. Our approach was to develop a practical NOx sensor based upon ion-mobility spectrometry (IMS). The status of the IMS sensor is outlined below:
A nonradioactive spark-discharge ionization source that replaces the conventional 63Ni b source was successfully developed and demonstrated.
- A laboratory prototype was built and tested. Negative NO2 ions were the primary ions detected in the two carrier gas streams tested, dry nitrogen and simulated exhaust gas.
- Negative-ion current intensity was correlated with both NO2 and NOx concentrations up to 200 ppm with high sensitivity and linear dependence.
- The present IMS sensor can produce a measurement in less than ten seconds, depending on the speed of the data acquisition system.
The second sensor being developed concerns the natural-gas composition. The approach taken is based on the acoustic properties of the gas mixtures, namely, the speed of sound and sound attenuation. A typical fuel gas composition of a natural-gas reciprocating engine consists of 82-96% methane, 2-7% ethane, 0.4-1.1% propane, and up to 0.6% higher hydrocarbons. Semiconductor sensors generally cannot fulfill the requirements because of slow response time, limited detection range, and cross sensitivity. The advantages of an acoustic sensor are fast response, robustness, and low cost; the drawbacks are poor sensitivity and applicability primarily to binary gas mixtures. We, therefore, focused on measurements of additional acoustic parameters such as acoustic attenuation for natural-gas identification.
Our attenuation measurements cover an f/p range from 0.05 to 0.5 MHz/atm. In this range, an
acoustic relaxation peak was observed for methane. The figure below shows the dimensionless attenuation
as a function of f/p in a 1:1 mixture of nitrogen/methane. A relaxation peak at f/p ≈ 0.06
MHz/atm is detected. The solid curve shows the theoretically predicted decay of the relaxation process.
To better fit the data, we derived an empirical expression, which is plotted as the dashed line in
where μmax is the peak maximum, and A is a constant. Figure 3 shows the acoustic attenuation
for pure methane, pure ethane, and a methane/ethane (0.9:0.1) mixture. The pure methane showed a relaxation
peak at f/p ≈ 0.05 MHz/atm, shifted to a slightly lower f/p as compared to
the (1:1) methane/nitrogen mixture. The peak was shifted to higher values of f/p when ethane
was introduced. The relaxation peak for pure ethane was not resolved very well. The data also show
that the attenuation in an ethane/methane mixture falls between the attenuation in methane and ethane,
but it may not provide a direct quantitative measure of the gas composition.
The status of the acoustic sensor is summarized as follows:
The sensor can detect changes in binary-gas composition by measuring changes in the speed of sound.
- An acoustic relaxation peak was detected for methane gas and its mixtures.
- By combining the sound speed and attenuation measurements, one can determine the composition of a natural gas that contains multiple components.
However, further development is needed to determine the sensitivity and reproducibility. At present, we believe the sensor can determine the percent of ethane or nitrogen in methane, but the percent of propane in methane may be difficult to determine.
NDE for Continuous Fiber Ceramic Matrix Composites
Work on continuous fiber ceramic matrix composites (CFCCs) was transferred from the DOE Office of Industrial Technologies to the Office of Power Technologies in FY 2002. This work is directed toward development of NDE methods that yield information on ceramic process development/reliability, damage level from either oxidation or impact, and effectiveness of repairs. Development efforts were focused on thermal imaging, air-coupled ultrasonics, and water-coupled ultrasonics and were highly interactive with various industrial suppliers: Honeywell Advanced Composites (now GE Power Systems/Composites), Textron, McDermott Industries, B. F. Goodrich, and Composite Optics. In the case of Honeywell Advanced Composites, several specimens were prepared for the study of thermal imaging and air-coupled ultrasonics relative to processing variables in the polymer impregnation process. Successful correlations with mechanical properties were established. For example, sensitivity to damage levels in SiC/SiC 2-D laminated CFCC material, induced by instrumented pendulum impactors, was studied by thermal imaging and air-coupled ultrasonic analysis; correlations were excellent.
Last Modified: Mon, April 19, 2010 3:13 PM