Edward James Nieters

A Multizone Technique for Billet Inspection

An ultrasonic inspection system has been developed in response to FAA recommendations for improved inspection of titanium billet [1]. This prototype system — called Multizone — has been transitioned to the factory floor and has inspected 1,000,000+ pounds of Ti billet in 1993–94. It is a real-time, PC based platform that employs custom built analog electronics using up to 8 parallel (non-multiplexed) channels, each with a remote pulser/receiver matched to the ultrasonic transducer. Scanned helically, the billet is divided into concentric zones with a focused transducer used to acquire peak detected C-Scan image data for each zone. The depth of each zone is established by the depth of focus of that transducer. C-Scan image data from all channels are displayed simultaneously on a 1024×1280 CRT and scroll as the inspection advances along the billet length. The data are written to optical storage upon completion of the inspection. The analog electronics are fully synchronous and could provide a baseline system for the acquisition of full waveforms. Custom post scan analysis software has been developed to detect flaws using signal to noise based algorithms. This software provides more reproducible results than conventional systems and greatly reduces operator fatigue and the chance for error. This paper will discuss the system architecture and operation. A companion paper in this volume discusses inspection results. [2]

Sampling-ring saturation pulse for two-dimensional magnetic resonance selective excitation

A method for suppressing sampling-ring artifacts produced by spiral-scan-based 2D selective excitation pulses, such as a those exciting a `pencil-shaped` region, employs a 2D annular saturation pulse followed by a gradient `crusher` lobe which dephases the transverse magnetization in the annular region. The annular saturation pulse is itself based on a spiral k-space trajectory having a limited number of cycles and a small outer radius, and is designed to saturate magnetization of tissue of the subject corresponding to the artifact rings of an excitation region while not affecting a central region. The annular saturation pulse may also be reshaped to limit the peak RF power to levels currently used for clinical MR imaging, while preserving bandwidth and the 2D excitation profile.

Error correction for Moiré based creep measurement system

Due to the high temperatures and stresses present in the high-pressure section of a gas turbine, the airfoils experience creep or radial stretching. Nowadays manufacturers are putting in place condition-based maintenance programs in which the condition of individual components is assessed to determine their remaining lives. To accurately track this creep effect and predict the impact on part life, the ability to accurately assess creep has become an important engineering challenge. One approach for measuring creep is using moiré imaging. Using pad-print technology, a grating pattern can be directly printed on a turbine bucket, and it compares against a reference pattern built in the creep measurement system to create moiré interference pattern. The authors assembled a creep measurement prototype for this application. By measuring the frequency change of the moiré fringes, it is then possible to determine the local creep distribution. However, since the sensitivity requirement for the creep measurement is very stringent (0.1 micron), the measurement result can be easily offset due to optical system aberrations, tilts and magnification. In this paper, a mechanical specimen subjected to a tensile test to induce plastic deformation up to 4% in the gage was used to evaluate the system. The results show some offset compared to the readings from a strain gage and an extensometer. By using a new grating pattern with two subset patterns, it was possible to correct these offset errors.

Continuous turbine blade creep measurement based on Moiré

Moiré imaging has been used to measure creep in the airfoil section of gas turbine blades. The ability to accurately assess creep and other failure modes has become an important engineering challenge, because gas turbine manufacturers are putting in place condition-based maintenance programs. In such maintenance programs, the condition of individual components is assessed to determine their remaining lives. Using pad-print technology, a grating pattern was printed directly on a turbine blade for localized creep detection using the spacing change of moiré pattern fringes. A creep measurement prototype was assembled for this application which contained a lens, reference grating, camera and lighting module. This prototype comprised a bench-top camera system that can read moiré patterns from the turbine blade sensor at shutdown to determine creep level in individual parts by analyzing the moiré fringes. Sensitivity analyses and noise factor studies were performed to evaluate the system. Analysis software was also developed. A correlation study with strain gages was performed and the measurement results from the moiré system align well with the strain gage readings. A mechanical specimen subjected to a one cycle tensile test at high temperature to induce plastic deformation in the gage was used to evaluate the system and the result of this test exhibited good correlation to extensometer readings.

Design and Fabrication of Forged Ti-6Al-4V Blocks with Synthetic Ti-N Inclusions for Estimation of Detectability by Ultrasonic Signal-To-Noise

In the work described subsequently, synthetic “hard alpha” inclusions have been fabricated within Ti-6A1-4V (Ti64) forgings. Several compositions of synthetic hard alpha were made by arc melting Ti sponge and TiN powder. Small solid cylinders of the titanium-nitrogen alloys were made by electro-discharge machining (e. d. m.) the arc-melted ingots to diameters of 0.031, 0.047, 0.062, and 0.078 inches, respectively, with heights equal to the respective diameter. Sets of eight or sixteen each identical cylinders were hot isostatic press (HIP) bonded within forged Ti64 blocks to yield uncracked inclusions with sharp interfaces with the Ti64 matrix. Detectability of the uncracked hard alpha was estimated as a function of flaw size, orientation, and nitrogen content from ultrasonic signal to noise ratios determined from C-Scan images of the blocks. Relationships of detectability to physical properties of hard alpha, and methodologies of signal to noise determinations are discussed.