Electricity: measuring and testing – Impedance – admittance or other quantities representative of... – Lumped type parameters
Reexamination Certificate
1998-07-29
2001-01-09
Metjahic, Safet (Department: 2858)
Electricity: measuring and testing
Impedance, admittance or other quantities representative of...
Lumped type parameters
Reexamination Certificate
active
06172511
ABSTRACT:
FIELD OF THE INVENTION
The present invention is directed to a method and apparatus for determining the size, growth and extent of a crack in a material, and is especially suited for the automated, continuous monitoring of cracks in materials and fatigue life monitoring.
BACKGROUND OF THE INVENTION
Materials used in structures are susceptible to failures due to cracking of the material which may result from fatigue. The failure of a component always results in inconvenience, and may result in considerable financial loss, for example where a plant needs to be closed down for replacement of the component, during which time production is lost. In some cases, the failure of a component can lead to injury or loss of life, especially where the failure occurs in a transport system such as an aircraft or train or plant operating at high temperatures or pressures. It is therefore important to determine how components will behave under certain conditions to estimate accurately when they will fail so that the component may be replaced before it fails, or to inspect components regularly to identify any weakness. Test measurements on samples may be made to estimate the behaviour of components in use.
There are two environments where monitoring is required, for components in service, and for laboratory testing. Conventionally very different techniques have been used in these different environments.
Engineering components are designed to have a specified “life”, when subjected to fatigue loading in service. This is known as the fatigue life and is a conservative estimate, based on design calculations. For most of this life, microcracks may be initiating. However, it is only during the latter stages of the components'life that a detectable crack begins to propagate.
For highly stressed components, fatigue life monitoring is of particular benefit, since if a component experiences higher loads than predicted by design calculation, there is the potential for in-service failure, with attendant hazards and economic penalties. Sensors are thus advantageously used to monitor fatigue life, or “residual life”, i.e. that proportion of the fatigue life remaining prior to the onset of failure.
In addition, monitoring has advantages for extending the life of components, to economic advantage. Due to the conservative nature of design calculations, most components may be made to operate far longer than anticipated. This is highly important in the current competitive climate, since it reduces maintenance and replacements costs.
However, as the component ages, it is increasingly important to have accurate knowledge of the condition of the equipment. Current techniques rely heavily on increasingly regular shutdowns and visual inspection. However, on-line monitoring would offer significant advantages, since it would provide improved knowledge of plant condition to avoid the need to stop operation regularly.
Current devices for monitoring fatigue in components in service include the “fatigue fuse”, made by Tensiodyne in the US and a fatigue monitor manufactured by STAS of France. Both these devices indicate the fraction of fatigue life consumed by the fracture of ligaments using the simple life fraction calculation based on “Miner's rule”. This states that for a given load, the fraction of life consumed is N/N
f
, where N is the number of cycles experienced and N
f
, the number of cycles to failure. However, both these devices are limited in terms of (i) only providing a periodic indication, rather than the capability to continuously monitor, (ii) inherent inaccuracy (STAS claim ±20% as being in ideal circumstances) and cannot accommodate variations in loading conditions. Moreover, once cracking has initiated in the component, neither of these devices can provide information on crack growth, which is critical to accurate prediction of failure in the component.
Where cracking is occurring in a component, the presence and extent of cracks can be determined by ultrasonic testing or magnetic particle inspection. These methods are conventionally used to test components in service. These methods do not allow continuous monitoring to determine continuous changes in the crack, but can only determine a crack size at a particular instant. A travelling microscope is conventionally used in a laboratory to observe cracking of a test sample. All these methods are manual methods which do not lend themselves to automation, and therefore are extremely labour intensive. Furthermore they are unsuitable for detection of cracks under non-ambient or extreme conditions, for example at high or low temperatures.
Automated techniques for determining the extent of cracks in a workpiece are primarily electrical detection methods, such as an electrical potential drop technique. This is generally used on a test sample in the laboratory. In this case, a constant current is passed along the sample. Any cracks in the material cause an increase in the impedance compared to an expected value. There is a voltage drop in the current passing through the sample, and this depends on the impedance of the sample. Therefore, by measuring the voltage drop across the sample, and comparing this to an expected voltage drop, the extent of cracks in the sample can be determined. The comparison is made either with experimental data, or from published data which may be based on experimental data on standard samples, for example by growing a crack to a certain length in a sample, splitting the sample, and measuring along the fracture surface. This requires that the electrodes are correctly placed on the sample, and requires that the sample is electrically conducting.
As the change in impedance due to cracks in the sample are small, it is necessary to have a large current passing through the sample so that the resulting voltage drop can be measured accurately, and so that variations in the voltage drop due to cracks can be distinguished from noise. Typically a current of 50 A is used to determine the size of cracks in a laboratory sample, although currents as high as 300 A may sometimes be used. Even with such high currents, a crack extension of 2×10
−6
m in length will typically only give a voltage drop of 1 &mgr;V.
Usually d.c. is used for electrical potential drop measurements of crack size, although this can lead to problems with thermoelectric electromotive forces generated at the junction between the electrodes and the sample. Further, metallurgical changes such as annealing and aging often cause changes in the resistivity of the sample, and crack tip resistivity and crack closure can affect the measurements.
The use of a.c. is less common due to the “skin effect” which means that the use of a.c. only allows detection of cracks at the surface of the sample. With a.c. at a frequency of 1 kHz, the skin depth within which a crack can be identified is only 0.1 mm. With higher frequencies, the skin depth decreases, although the sensitivity improves. With a.c., a current as low as 2 A can give an accurate determination of crack size. A.C. electrical voltage drop methods require correction for stress induced resistance contributions, and, where a high frequency current is applied, for inductively induced electro-motive forces.
The use of an electrical voltage drop method for detecting crack size is temperature dependent, as the impedance of a sample varies with temperature. To stabilise the temperature and electronics to give accurate results, a warm up period of up to 24 hours is required.
One of the most common methods of determining crack length of crack growth in a test sample is a compliance method. This can be carried out over a wide range of temperatures. In a compliance test, a load is applied to the sample, and the displacement induced by that load is determined. Under elastic conditions, the displacement in the sample will be dependent on the geometry and the size of cracks in the sample. The load applied to the sample may be measured by a strain gauge or load cell, and the displacement by a clip gauge at ambient temperatures
Nicholls John Rayment
Tidswell Roger David
Baker & Daniels
Cranfield University
Kerveros James C.
Metjahic Safet
LandOfFree
Measuring device does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Measuring device, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Measuring device will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2526039