Nondestructive testing

From Free net encyclopedia

Revision as of 08:34, 15 April 2006; view current revision
←Older revision | Newer revision→

Template:TOCright

Nondestructive testing (also called NDT, nondestructive evaluation, NDE, and nondestructive inspection, NDI) is testing that does not destroy the test object. While destructive testing usually provides a more reliable assessment of the state of the test object, destruction of the test object usually makes this type of test more costly to the test object's owner than nondestructive testing. Destructive testing is also inappropriate in many circumstances, such as forensic investigation. That there is a tradeoff between the cost of the test and its reliability favors a strategy in which most test objects are inspected nondestructively; destructive testing is performed on a sampling of test objects that is drawn randomly for the purpose of characterizing the testing reliability of the nondestructive test.

The content of the American Society for Nondestructive Testing's journal Materials Evaluation and of the textbook Nondestructive Evaluation: A Tool for Design, Manufacturing and Service suggests a smaller scope for the discipline called "nondestructive testing" than is implied by its label. For example, the discipline appears to exclude nondestructive biomedical tests, such as urinalysis.

Contents

[edit]

Applications

NDT is used in a variety of settings that covers a wide range of industrial activity.

[edit]

Methods and techniques

NDT is divided into various methods of nondestructive testing, each based on a particular scientific principle. These methods may be further subdivided into various techniques. The various methods and techniques, due to their particular natures, may lend themselves especially well to certain applications and be of little or no value at all in other applications. Therefore choosing the right method and technique is an important part of the performance of NDT.

New methods that use advanced sensing technology or that have become possible with today's computing power are always being developed.

[edit]

Elements

Regardless of application or method, all nondestructive testing shares the same basic elements:

Source 
Source provides a medium for testing.
Modification 
The probing material must get modified due to variation in the source.
Detection 
A detector which will determine the changes on the probing medium.
Indication 
Some means of indicating/recording signals received from indicator.
Interpretation 
A method of interpreting indications.
[edit]

Reliability

Its testing reliability is defined by a test's perfect gold standard test, by a binary variable, to be termed OK/not_OK, whose values are elements in the set {OK, not_OK}. By definition, the nondestructive test errs when its measurement of OK/not_OK differs from the measurement of the same quantity by the perfect gold standard test. The Cartesian product of the sets of values respectively measured by the perfect gold standard test and the nondestructive test defines the four, possible events that belong to a measurement: a "false negative" occurs when the test object is not_OK but it is identified as OK by NDT; a "false positive" occurs when the test object is OK but it is identified as not_OK by NDT; similarly, a "true positive" occurs when the test object is not_OK and it is identified as not_OK by NDT and a "true negative" occurs when the test object is OK and NDT identifies it as OK.

[edit]

Defect detection tests

NDT's defect detection tests attempt to detect defects (e.g., cracks, intergranular corrosion, pits and inclusions) before they can cause structural failure, leaks or other, unfavorable outcomes. The existence of a "probability of detection," corresponding to the probability of a true positive given a true positive OR false negative (inclusive disjunction implied by OR) is more often asserted for these tests than delivered by their design. Similarly, the existence of a "probability of false call," corresponding to the probability of a false positive given a false positive OR true negative is more often asserted than delivered. This topic is amplified below.

[edit]

Violations of probability theory

Oldberg and Christensen (1995) and Oldberg (2005) report that today's defect detection tests violate probability theory empirically. Often, a probability of false call is not defined. Sometimes, the "probability" of detection is not a probability. In these ways, the testing reliability is ill-defined.

The false impression that the testing reliability is well defined can be established by the use of terms that imply the preservation of probability theory under conditions in which probability theory is empirically violated. An article that is posted by the U.S. Nuclear Regulatory Commission, 2000 at its Web site provides a case in point. In the article, the agency applies the terms "probability of detection," "signal" and "noise" to a Defect Detection Test that is used in the safety inspections of nuclear power reactors. The terms "signal" and "noise" imply that probability theory is unconditionally preserved, but Oldberg and Christensen (1995) demonstrate that it is only conditionally preserved in this test. The term "probability of detection" implies that probability theory is preserved, given the event of a true positive OR false negative but Oldberg and Christensen (1995) demonstrate that, for this test, this is not true.

Some of the consequences of this kind of terminological abuse are: a) it appears that the testing reliability can be determined when it can't b) it appears that the expected utility of testing can be determined when it cannot and c) it appears that decisions can be made about testing on a basis that is considered rational by decision analysts when one's ability to make such decisions is actually crippled. In a defect-sensitive technology such as nuclear power, a further consequence might be significant, unnecessary loss of life and property damage.

Potential and actual users of Defect Detection Tests should also be aware of the fact that the expected utility of such a test depends upon a) the nature of the statistical populations to be tested and b) the utilities of the user. As the populations are undefined and the associated utilities vary by user, claims that testing with a particular technology has a greater, expected utility than non-testing or that testing with a particular technology as a greater, expected utility than testing with another technology should be met with skepticism.

[edit]

See also

[edit]

References

  • Bray, D.E. and R.K. Stanley, 1997, Nondestructive Evaluation: A Tool for Design, Manufacturing and Service; CRC Press.
  • Materials Evaluation, monthly series; American Society for Nondestructive Testing, Columbus, Ohio.
  • Oldberg, T. and R. Christensen, 1995, "Erratic Measure" in NDE for the Energy Industry 1995; The American Society of Mechanical Engineers, New York, NY. Republished by ndt.net at http://www.ndt.net/article/v04n05/oldberg/oldberg.htm .
  • Oldberg, T., "An Ethical Problem in the Statistics of Defect Detection Test Reliability," 2005, Speech to the Golden Gate Chapter of the American Society for Nondestructive Testing. Published by ndt.net at http://www.ndt.net/article/v10n05/oldberg/oldberg.htm .
  • U.S. Nuclear Regulatory Commission, 2000, "Issues Stemming from NRC Staff Review of Recent Difficulties Experienced in Maintaining Steam Generator Tube Integrity", http://www.nrc.gov/reading-rm/doc-collections/gen-comm/reg-issues/2000/ri00022.html#_1_6 .
[edit]

External links

fr:Contrôle non destructif nl:Niet-destructief onderzoek uk:Неруйнівний контроль ru:Неразрушающий контроль


Template:Industry-stub

Retrieved from "http://www.netipedia.com/index.php/Nondestructive_testing"