·Home ·Table of Contents ·Workshop - Reliability

Application Parameters - A Definition and Their Role in NDE Reliability Matthew J. Golis, Advanced Quality Concepts, PO Box 141388, Columbus, Ohio 43214 (614) 268-0518 Email : mattgolis@columbus.rr.com Contact

Background

The reliability model, adopted at the First Reliability Workshop held in Berlin in June 1997, identified three functions that are believed to capture the dependence of NDT reliability as it relates to system capability, specific application features and human factors. The conceptual model suggests that the reliability of an inspection technique will never be better than the idealized or intrinsic capabilities (IC) (based on scientific principles) of each NDT technique. The model further suggests that reliability will tend to be degraded when using the technique for specific industrial applications because deviations from the idealized assumptions used in determining the technique's intrinsic capability must be considered. Idealized assumptions that may be affected are called the application parameters (AP). In general, application parameters relate to the bulk materials through which the probing media must pass, the naturally occurring discontinuities found in such materials, and the geometrically based constraints related to NDT sensor positioning and access. Human factors (HF) are treated as a separate degrading element of the model, applicable primarily in those situations requiring manual manipulation practices.

Thus, NDT technique reliability, R, as it relates to a singular type of discontinuity can be represented using the conceptual model shown below where f, g and h are still-to-be-determined functions dependent upon IC, AP and HF, respectively.

This conceptual model merely suggests that some function, f, dependent upon the technique's intrinsic capabilities, is representative of the technique's reliability under idealized conditions. When application parameters (deviations to idealized assumptions) are considered, the idealized reliability will be reduced in accordance with the nature of another function, g. When human factors associated with manual inspections are considered, the effective reliability is reduced even further, following function h.

A conceptual drawing of this relationship is shown to the right. On the left is an idealized level of reliability, established by the Intrinsic Capability of the inspection technique being used. On the right is the predicted level of reliability derived from the estimated negative impacts of specific application parameters and the degrading effects of human factors.

The purpose of this paper is to define what constitutes an application parameter. The nature of the functional relationship g(AP) is driven by the general types of formats considered applicable to expressing reliability performance. The characterization of each of these three functions (and identification of ways in which they can be linked in a quantitative manner) was discussed in depth during the second Reliability Workshop and still remains to be done. This paper presents the definition for an Application Parameter and how they might be evaluated and used to realistically predict the most probable reliability of a specific NDT technique used for a specific application.

A Basis for Intrinsic Capability

The concept of an Application Parameter cannot be developed without a clear definition of what constitutes the basis for an inspection technique's intrinsic capability. An inspection technique is said to have an Intrinsic Capability based on the technique's scientific principles. In order to simplify the definition of an application parameter, let us assume that a singular discontinuity must be detected. Thus clusters of discontinuities and issues of sizing are not of interest at this time. We will be striving to determine the sensitivity of the inspection technique under specific operating conditions. The degrading of inspection sensitivity and reduction in operating range will be used as measures of the technique's application-dependent reliability degradation.

An inspection using the ultrasonic method of NDT, for example, would have an intrinsic capability based, in part, on the electronic instrument's performance characteristics, e.g., noise level, amplitude dynamic range, time-base dynamic range and frequency response. However, the intrinsic capability for a specific technique must include additional variables such as the transducer's performance characteristics (conversion constants and frequency response) and any constraints programmed into the instrument to modify the appearance or utility of received signals (e.g., reject, filtering, and gate threshold settings).

In assessing Intrinsic Capability, we assume the test material is homogeneous and isotropic. Surfaces are smooth and flat. Acoustic wave velocity and impedance are known and assumed constant. The useful range of the ultrasonic instrument is the region above the electrical noise level and below the upper limit or saturation point of the instrument. The acoustic wave characteristics in this ideal material would be predictable based on material wave velocity, transducer size, pulser output and transducer conversion coefficients (d₃₃ & g₃₃). The simulated discontinuity would be oriented at an optimum angle (perpendicular), comprised of an air-backed flat, smooth surface of diameter equal to the diffraction limited extent of the acoustic wave (possibly l /10 or so). Static readings could be taken from any direction. Given these variables, it is possible to predict the ultrasonic wave's geometrical beam characteristics (directivity function) and its signal strength, relative to the initial energy input. The signal sensitivity could be estimated and its operational dynamic range would be at its maximum.

A Basis for Application Parameters

The concept of application parameters arose from the thought that success with inspections is often less than ideal due to symptomatic, but often acceptable irregularities in the test parts themselves, limitations in equipment being used, uncertainty regarding the exact nature of naturally occurring discontinuities, and the strategies used for transducer placement, orientation and scanning. Thus, Application Parameters can be divided into these four general categories. The first category involves bulk material characteristics whose presence can alter the pristine nature of the probing energy. The second involves the limitations of the instrument/transducer combination used to create and detect the ultrasonic pulses. The third involves naturally occurring discontinuity characteristics that differ from those idealized in defining the technique's Intrinsic Capability. The fourth involves the geometrical constraints due to test item shape, ability to introduce probing media from optimum directions and the practical issue of scanning speeds and inter-pass intervals. For example, let us examine Ultrasonic NDT.

Application Parameters for an Ultrasonic Inspection Technique

Material Characteristics
In the case of Ultrasonic NDT, attenuation and anisotropy are the major material characteristics that degrade the passage of wave fronts. Attenuation has two materials-related components[Note that beam spread due to diffraction is considered an idealized beam characteristic and thus is part of the Intrinsic Capability estimate. Although it relies on the acoustic wave velocity in the test material, it is not considered a material-dependent attenuation Application Parameter.]. One is scatter due to inherent reflectors, e.g., grain boundaries and inclusions. The other is non-elastic absorption in some materials, e.g., plastics and rubber. Each of these distributive features reduces the sound energy available for detecting discontinuities.

Scatter from material texture, grain boundaries, small inclusions and voids also degrade the wave's coherence, and in addition, introduces backscatter noise into the detected signal. In the case of materials that exhibit backscatter, the effective noise level of the ultrasonic instrument increases and therefore the available operating range of the instrument/transducer combination is reduced. Engineered materials such as composites exhibit differential propagation characteristics (anisotropy) in accordance with the materials used and orientations of reinforcing fibers or aggregates. Such material effects are often summarily included in measures of material propagation characteristics such as bulk wave group velocity and signal level attenuation.

Another wave-coherence degrading effect is caused by rough surfaces upon entry and upon internal reflection from exterior surfaces. Each interaction of an ultrasonic wave with a rough or irregular surface contributes to the deviation of the wave front from that of the coherent (in-step) waves initially created by the pulse excitation of the transducer.

Inspection Equipment Characteristics
The capability of the inspection equipment is the basic reference for overall system reliability. A piece of equipment, coupled with a transducer-type sensor, establishes whether or not an inspection can be conducted. It must have a sufficient useful dynamic range with a suitable level of sensitivity for each specific application. The ideal instrument will be capable of showing the presence of the smallest physically detectable indication from a test target. Presumably, this will be at or about the diffraction limit of a crack tip. When a reference reflector, e.g., flat-bottomed or side-drilled hole, is used to set the sensitivity of an ultrasonic instrument, this merely serves to establish the operating level of interest.

The useful operating range of the instrument in a specific application is the region between the reference level (or some specified percentage thereof) and where the instrument becomes saturated, e.g., at the top of the viewing screen or where the gain can no longer increase signal response. The lower extreme of the overall operating range in a specific application is limited by the instrument's electrical noise or the backscatter noise from the test material. The upper extreme will be limited by the effect that material and interface attenuation needs to be overcome by the gain capabilities of the system.

Discontinuity Characteristics
Reflector impedance differences, orientation, flatness, smoothness and existing stresses are the major characteristics that degrade signal response from naturally occurring discontinuities. Factors such as irregular shapes and compressive stresses all tend to degrade the likelihood of detecting naturally occurring discontinuities. In some applications, e.g., laminations in plate or fatigue cracks in parts loaded in pure tension, the reflectors are statistically flat, relatively smooth, in a predictable orientation and exhibit air-backed conditions. When the transducer can be positioned properly, this class of discontinuity can be detected with a relatively high level of reliability as long as the sensitivity called for can be produced by the ultrasonic equipment.

In other applications where orientation, morphology and stresses are unknown, the detection probabilities can decrease precipitously. In such cases, the response affects of different discontinuity characteristics may be approximated using the statistical probabilities of their respective occurrences. The figure to the right schematically shows the effects of the degrading influences of material noise, material attenuation, discontinuity orientation and shape and their attendant reductions in system operating range. The issue of exactly how to define the ultimate sensitivity of a system remains an open question and is represented by the question mark, (?).

Geometrical Constraints Characteristics
In addition to the material, equipment and discontinuity characteristics, constraints introduced through inspection protocols and physical encumbrances can also degrade the reliability of NDT. For example, in ultrasonic shear-wave inspection, only a limited set of transducer aspect angles is used. Scanning of test parts may be in accordance with a less than 100% sampling scheme. Test part curvature in pipes and nozzles degrades coupling effects and beam consistency. Even some corrective measures, such as contoured shoes, limit the ability to rotate transducers in order to enable off-normal detection of discontinuities. The more convoluted the test part surface, the more the effectiveness of the inspection is compromised. These conditions must also be considered part of the collection of application parameters which affect the reliability of an NDT activity.

A Suggested Listing of Ultrasonic Method Application Parameters

Listed are the application parameters that can affect the performance on an ultrasonic inspection. Note that changes in parameters (D variable) are mostly found with regard to the material being inspected and the nature of the discontinuities being detected. Only the instrument's "filtering" settings, selected in accordance with a specific inspection protocol, are considered Application Parameters.

ULTRASONIC METHOD VARIABLES
Contributing Element	Intrinsic Capability	Application Parameters
Material	Homogeneous	D Noise (Scatter)
	Isotropic	D Attenuation (Absorption & Scatter)
	Smooth Surfaces	D Surface Roughness
	Flat Surfaces	D Surface Curvature
	Acoustic Impedance	D Acoustic Impedance
	Acoustic Velocity
Instrument1	Electrical Noise Level
	Dynamic Range
	Pulse Output
	Unfiltered Signals	D Filter Settings2
Transducer	Effective Size
	Wavelength in test part
	Conversion Coefficients
	Acoustic Impedance
Discontinuity3	Best Case Scenario DZ, Maximum Orientation, Right Angle Size, Diffraction Limited Flat and Smooth	Change in DZ (material & stress) D Orientation D Size D Flatness or Smoothness
Transducer Positioning	Unconstrained	Limits of constraint
	Coverage	Scan Intervals & Rates
	Directions	Aspect Angle(s)
	Static Readings	Surface Access (inherent & external)

Notes:
1. Output signals measured at Near Field - Far Field Transition point, (Yo+).
2. Filter Settings include Reject control, Gate Threshold, Smoothing, etc.
3. Discontinuities singular in nature.

A Suggested Listing of Eddy Current Application Parameters

Listed are the application parameters that can affect the performance on an eddy current inspection. Note that changes in parameters (D variable) are mostly found with regard to the material being inspected, consistency of coupling and the nature of the discontinuities being detected. Both "filtering" settings and heating caused by exciting-current induction are considered Application Parameters.

EDDY CURRENT METHOD VARIABLES Encircling Coil (Absolute)
Contributing Element	Intrinsic Capability	Application Parameters
Material	Homogeneous
	Isotropic
	Smooth - Straight - Round	Dh : coupling, fill factor, noise
	Conductivity Constant, s	D s : inclusions, temperature, heat treat noise
	Permeability Constant, mr = 1	D mr, : inclusions, heat treat, work harden noise
Instrument	Electrical Noise Level	D Environmental (elec/mag) noise intrusion
	Dynamic Range
	Exciting Current	D Induction heating drift
	Unfiltered Signals	D Filter Settings, thresholds, phase selection
Coil	Fill Factor, h = 1
	Length
	Number of Turns
Discontinuity	Best Case Scenario s = 0, Perpendicular, On Surface Size ~ TBD	D Orientation D Surface to subsurface D Size
Coil Positioning	Aligned with axis
	Static Readings	Throughput Rates

Examples using other NDT methods and scenarios could be developed, but the general approach would be the same as that used for Ultrasonic and Eddy Current NDT. Having reviewed the potential approach, how should the term Application Parameter be defined?

What is an Application Parameter?

An Application Parameter is any physical condition or state identified as pertinent to the physical aspects of an NDT technique, which when changed, alters the performance of the technique from that expected under idealized conditions.

Application Parameters are largely related to material and discontinuity features different from those assumed in an idealized inspection. They can be modified by operational settings related to instrument settings designed to ease or modify signal interpretation or selectivity. They can be related to operational procedures that require use of specific, but limited, sensor positions caused by physical impediments and economically attractive scanning practices.

© AIPnD , created by NDT.net

|Home|    |Top|