Developing a nondestructive inspection instrument and taking it from the laboratory setting to industrial field conditions presents its own set of challenges. Research initiated in 1986 under a joint US/Canadian coalition resulted in the development of an electromagnetic testing (ET) wire rope tester. In this month's column, the authors report results of extensive field testing of this computer controlled instrument in a number of mines under rugged industrial conditions. G.P. Singh Associate Contributing Editor

 

Figures 1-2
Figures 3-4
Figures 5-6
Figure 7

Introduction
Nondestructive testing of wire ropes on an industrial scale, with electromagnetic testing (ET) instruments of diverse designs, has been well known and widely practiced for decades. The principle of subject magnetic flux detection technique is based on the measurement of three phenomena, namely: (a) of the leakage flux around a local fault (LF) in a damaged rope section, (b) of the change in the rope's magnetic impedance, due to a loss in the magnetic cross sectional area (lma), and (c) of the change in the magnetic flux value of the magnetic circuit in the sensor head.

While certain features of these instruments did change, some even in quite ingenious ways, their basic concept has remained unaltered for about the past 15 years. It is, therefore, of particular interest to be able to report on the industrial maturity of a major new development, namely on the computerized control of a dual function ET wire rope tester.

The basic idea of substituting for analog signal displays digitized ones, with conjoint computer controls, is fairly self evident. So are the many advantages - such as the considerably improved tester flexibility, with more user friendly and less operator dependent handling - that accrue from successfully performing this transition at economically acceptable costs, and with designs suited to the tough industrial conditions that prevail. In practice, though, the development in question proved to be far from straightforward.

---

The basic idea of substituting for analog signal displays digitized ones, with conjoint computer controls, is fairly self evident.

---

Ongoing achievements, covering a period of some 10 years, were previously reported in an extensive series of articles. This paper presents the concluding, and possibly most important results of subject development, namely ones obtained by extensive testing of the computer controlled instrument in a number of mines under rugged industrial conditions.

A major research and development project, initiated some 10 years ago at the Canadian Federal Government's Canada Centre for Mineral and Energy Technology (CANMET) in Ottawa, and supported by other governmental and industrial agencies both in Canada and abroad, has been described in previous publications (Geller et al., 1992, Geller et al., 1995; and Hamelin et al., 1995). The authors do not, therefore, consider it necessary to reiterate the project's original mandate or justification, or list again the results which culminated in the decision to develop a state of the wart, more user friendly and less operator dependent NDT system. Its basic design specifications called for a computer controlled, dual function, ET wire rope tester, with permanent magnets, Hall sensors, and digitized signals, suitable for rope sizes from 13-64 mm (0.5-2.5 in.).

This paper is a direct follow-on to Geller et al., 1995. In this it was stated that the laboratory and theoretical results would be followed by results obtained under exacting field test conditions - a necessary adjunct to any research and development work laying claim to ultimate acceptance by industrial users. Field testing was performed in five important Canadian mines: two in New Brunswick, and one each in Manitoba, Ontario, and QuŽbec. Ropes tested included hoist, balance, and guide ropes; stranded and locked coil (LC) constructions; and sizes ranging from 24 mm to 57 mm (0.94 to 2.25 in.).

While the authors obtained a wide range of results, and drew both practical and theoretical conclusions, in this report they concentrate on points of special interest to hands-on practitioners of rope testing, and to those who need to act upon the latters' reports.

Instrumentation
The theoretical basis of the technique in question, namely of the permanent magnetic field method for detecting wire rope anomalies, has been extensively studied and described in a wealth of publications. In simple terms, the LF and LMA signals of the NDT instruments represent the electronic equivalent of the mechanical anomalies present in the wire rope. The saturating magnetic field of the tester makes the latter visible to the magnetic sensors placed around the rope. The process is somewhat similar to making an NDT examination of a human body with X-rays, where density variations of the patient are made visible by greater or lesser absorption of these rays.

An overview of the practice is provided by the relevant standard (ASTM, 1996). A recent summary of the magnetic method in question was also published (Martyna, 1997), while a detailed description of the analog design sensor head used by the authors can be found in Kitzinger and Wint, 1978. Detailed operating instructions for the computerized tester are given by Geller and Kitzinger, 1996.

The field test data presented in this paper were obtained with the commercially available model of a dual function Canadian instrument, equipped with Hall sensors and permanent magnets. It is known as the Magnograph in its analog design, and as the Magnograph II in its digitized design version. The results obtained with the analog version are representative of signals as obtained with most other currently available dual function analog wire rope testers. These results are, therefore, well suited as benchmark illustrations with which to compare the digitized displays. The full range of the digitized model's flexibility, and of its many playback options, can not be covered in the limited space available here. This has, however, been done in (Geller et al., 1995) and in (Geller and Kitzinger, 1996). Moreover, many important features, made possible only by digitizing the signals, are presented in the illustrations of the present article.

While two distinct sensor head designs were used for the field tests in this project (the analog and the digitized models) playbacks of both were handled by the same computer circuitry. This was done so as to facilitate the juxtaposition of the respective results, i.e., to get around the difficulties inherent in handling, and publishing, the analog design's long paper strip-charts. This hybrid operation was possible, because the necessary work had already been completed on a prototype instrument, in anticipation of requests from present day users of the analog tester to have their instruments retrofitted for computerized use.

Field Test Results

General Remarks
It may be helpful to make some general comments about the authors' field testing practice and experience, before presenting a selection of the results:

• Two runs were always performed along all rope sections that were to be tested. This practice is advocated by others as well, e.g., by (Poffenroth, 1996) of the US Department of Labor's Mine Safety and Health Administration (MSHA). Some practitioners may disagree because of the additional costs, but the authors contend that any extra costs that may accrue are more than offset by the greatly increased reliability and accuracy of the resulting process.
• Testing was always initiated as close as possible, say within 1.5 m (5 ft), to the rope ends (skips, cages, counterweights, cappels, etc.), safety considerations alone being the limiting factor. We advocate this, contrary to the practice of others who often choose to initiate their tests no closer than some 6 m (20 ft) from the rope end.
  
  The authors' tests showed that this is not only unnecessary, but may even be detrimental, in cases where the best rope sections are short ones, which are immediately adjacent to the rope end(s):
• Whenever two sensor head setups were necessary along the same rope, a very frequent occurrence, the authors always attached a marker at the appropriate location of the overlapping rope section. This procedure, while widely recommended, is not used in practice to the extent which is desirable. Moreover, testing at the second rope end was, as it always should be, recommenced so that magnetic polarizations matched each other at both ends of the rope.
• While testing could proceed at any steady rope speed of up to some 3.5 m/s (690 ft/min), unduly rapid speed changes had to be avoided. These produce spurious signal anomalies. However, there is no
  optimum test speed as far as instrument design is concerned. For overall safety
  reasons some mines prefer not to exceed a speed of about 1.3 m/s (250 ft/min). Others operate in the 3.1 m/s (610 ft/min) range, or faster. Equally good results are obtained at crawling speeds, needed when locating specific anomalies of interest.
• Spurious signal anomalies also arose when testing passed from a nonmagnetized to a premagnetized section of the rope, or when it was influenced by the rope's history of remnant magnetism. These anomalies can be readily recognized, and thus discounted, provided sufficient importance is routinely attached to their presence, e.g., by using two passes along each rope section - as recommended under the first point, above.
• The test results were consistent and repeatable to an extraordinary degree. Specific characteristics of rope constructions, such as locked coil (LC) or plastic filled valley (PFV) designs, as well as rope anomalies, could often be easily recognized from their test signal displays alone.

A selection of field test results follows, to illustrate the state-of-the-art possibilities now available through the use of subject computerized and digitized sensor head design. Salient design features include:

• The use of highest quality, samarium cobalt alloy, permanent magnets.
• Incorporation of Hall sensors.
• Electronic signals that record, at every 4 mm of rope travel, features such as the rope's LF and LMA conditions as well as its direction, distance, and speed of travel.
• Elimination of the need for chart paper and ink supplies.
  

Evaluation of the Rope's Loss of Metallic Cross Sectional Area (LMA)
As an illustration of the computerized system's flexibility, the authors' standard procedure for establishing LMA results is described:

• For both the long and the short rope ends, play back the complete length of each test pass onto one screen display. Print out these displays onto letter sized (216 ´ 279 mm [8.5 ´ 11 in.]) sheets. From these, establish the location of the best and worst rope sections, and the severity of corrosion all along the tested rope sections. The LMA values at any location can easily be obtained by moving a level cursor to the point of interest.
  
  In case of doubt, such as perturbations due to remnant magnetism or excessively rapid rope speed changes, individual rope sections can be examined in great detail by narrowing the vertical or the horizontal display ranges. This procedure is also used to verify whether doubtful LF peaks are due to wire breaks or to a gain in metallic cross sectional area (MA).
• Using the available software options, the zero percent LMA level is placed at the best rope section. The pertinent LMA values can be obtained at any rope location by using the movable level cursor. If the best rope section should be known to be in its as manufactured condition, the rope's catalogue weight is assigned to it. If the test can be completed with one sensor head set up, the LMA evaluation process can now be completed.
• If two sensor head set ups have been used, the measured rope weight (in kg/m or lb/ft) at the marker location (i.e., in the overlapping long short rope sections) must also be recorded in that rope length along which (say the short length) the overall zero percent LMA level has previously been located. Using the software, the other rope section display (i.e., the long one in this example) must now be moved so that the same kg/m (or lb/ft) is recorded at its marker location as has been measured in the short rope end. The LMA evaluation process can now be completed.
• As noted in the foregoing, the computerized instrument's software displays both the rope's point to point weight in kg/m (or lb/ft), and the corresponding percent LMA values. If the displayed weight in the zero percent LMA section equals the as manufactured value, the percent LMA data equals the percent total change of metallic area (TCMA) data. If not, then the percentage of difference between the two is to be added to the LMA data, so as to arrive at the TCMA information (i.e., at the rope's cross sectional metallic area change as a percentage of its original condition).

We illustrate the foregoing comments in Figures 1a and 1b. These record the second pass test results, as obtained with the authors' computerized sensor head system, on a 54 mm (2.12 in.), 6 ´ 27 (this expression indicates the number of wires in a strand wrapped around a core of other wires), skip hoist rope. In Figure 1a, the short rope end, we placed the rope's as manufactured 11.6 kg/m (7.8 lb/ft) weight at the display's best section. At the time of this test, the subject rope had only five months of use. Also, the best section is adjacent to the hoist drum, with the rope fully extended. Consequently, the original rope weight of 11.6 kg/m (7.8 lb/ft), and the zero percent LMA TCMA level, could be assigned to this best section on the signal display. Thus the corollary rope weight, and percent LMA data, at the marker position in the short rope end were seen to be 11.14 kg/m (7.47 lb/ft) and 4.16 percent, respectively. The corresponding marker values in the long rope end signal display (Figure 1b) are 11.14 kg m (7.47 lb/ft) and 4.17 percent.

Note that although some 890 m (2920 ft) are displayed in Figure 1b on a letter sized print out, the distance at any given test location can still be easily ascertained, because the program places a conspicuous length mark every 5 m (25 ft) of rope travel.

 

Clear Distinction Between Loss, or Gain, of Metallic Cross Sectional Areas
As mentioned in the first point above, it is easy to distinguish between LF peaks that correspond to a loss of metallic area (LMA), and the ones caused by the addition of metal (e.g., by welds and markers). As an example: along the 1280 m (4200 ft) second pass print-out of a 40 mm (1.57 in.), 34 ´ 19, plastic filled valley (PFV) skip balance rope, conspicuous LF peaks appeared every 305 m (1000 ft). The authors were unaware of the true origin of these. Playback of several of these sections, one of which is shown in Figure 2, revealed the presence of some 1m long (3.3 ft) ferrous inclusions in the rope's fiber core. Note the synchronized rise and then dip of both the LF and LMA signals.

Although the illustration is from the test's second pass, the software restores proper polarity of the LF signal whichever way the rope is moving through the sensor head. The program also ensures that the direction in which the rope moves through the head at the start of the test is always recorded as the forward direction, while opposite movement is always recorded as the reverse one.

Equally clearly the LMA and LF signals move, again in synchronization, in the opposite direction in case of a metallic area loss. As an example note Figure 3. It is a sectional playback from a 1259 m (4130 ft) first pass along a 32 mm (1.25 in.), 1 ´ 149, LC skip hoist rope.

 

Clear Distinction Between Genuine and Spurious Rope Anomalies
Figure 4 is from the playback of both the first and second passes, with the authors' analog instrument, along the short end of a 29 mm (1.15 in.), 18 ´ 7, cage balance rope. It illustrates how one can readily, and quite clearly, distinguish between signal perturbations caused: (1) by genuine anomalies, (2) by remnant magnetism, and (3) by wire breaks (loss of MA) or markers (or other causes of MA increase). The 2.2 percent first pass and 2.4 percent second pass LMA perturbations are seen to be genuine, since: (a) these occur in both the first and second passes at the same rope locations, and (b) these clearly coincide with a major change in the severity of the corrosion pattern along the LF signal. However, the 2.3 percent LMA is caused by remnant magnetism, in this case by testing passing from the non premagnetized rope section onto the one that had been magnetized during the first pass. This is seen to be so because: (a) the 2.3 percent LMA occurs only during the first pass, and (b) it occurs shortly ahead of the marker. The first test pass has, of course, ended shortly after this point.

LF peaks at wire breaks and markers were again easily distinguished from each other by expanded playbacks of the relevant rope sections.

Improved signal displays with the digitized sensor head design Figures 5a and 5b illustrate the major differences between the immediacy with which the LMA signal of the digitized sensor head design (Figure 5a) follows the MA changes in a wire rope, as compared to the gradual, because averaged over the length of the sensor head, LMA signal shifts obtained with the analog design (Figure 5b). This anomaly (one of a large number) occurred in a 29 mm (1.15 in.), 18 ´ 7, cage balance rope, a companion rope to the one referred to in Figure 4. Test 5b was obtained with the analog design sensor head in January, 1996; test 5a was obtained with the digitized design four months later. While Figures 5a and 5b cover only the second test pass along the rope, these differences are just as clear in the first pass as well.

 

Documentation of Unduly Rapid Speed Changes
As mentioned in the fourth point, of the "General Remarks" section, any steady test speed from stationary to maximum permissible produces equally good results. However, unduly sudden speed changes, such as often occur when stopping to affix markers, cause spurious LMA and LF signal swings, not unlike the vibrations discussed and illustrated by Costello and Phillips in the dynamic analysis section of their contract report on wire rope stresses (Costello and Phillips, 1983).

This spurious anomaly is a well-known occurrence, often referred to by others. It is, therefore, of particular interest that the many playback options of the computerized tester system include one that readily documents the test's entire speed history.

 

Calibration of the LF Signal Displays
Another matter of considerable practical consequence is the computerized system's ability to tie in the amplitude of the LF signal peaks with specific ropes and specific wire breaks. On the basis of theoretical considerations and laboratory testing, it could always be shown that the LF signal's shape and amplitude are functions of several basic parameters, including the rope's weight and size, the broken wires' size and location within the rope, and the gap size of the break. With the computerized system the history of the signal voltages during the entire test process can now be recorded and displayed. Therefore, the influence of these parameters can be quantified. The authors conducted a number of relevant laboratory experiments, with results such as the ones shown in Figure 6.

The same type of tests can be performed for the LF signal patterns in the field, thus establishing quantified benchmarks for very specific rope sizes and constructions, for specific anomalies and corrosion severity, and for quite specific operational conditions. These LF patterns are known to be remarkably consistent and repeatable.

 

In Situ Versus Reel-to-Reel Testing
The question of the comparability of in situ tests and reel-to-reel tests is of considerable practical interest. The authors have always maintained that these are quite comparable. Their conclusions have been based both on numerical calculations (Geller et al., 1990), and on the CANMET research and development project; it contained many test results obtained on the same ropes, tested first in situ and subsequently reel-to-reel.

Even so, it is of renewed interest to refer to this question, now that results can be presented in much greater detail. As an example, please note Figures 7a and 7b. These were obtained with the authors' computerized system on a 29 mm (1.15 in.), 18 ´ 7 balance rope in situ (Figure 7a) and a month later reel-to-reel (Figure 7b). The respective displays are seen to be almost identical.

 

Summary
This paper contains only a brief selection of the field test results obtained to date with the new computer controlled wire rope test instrument. Even so, the many novel features and advantages of this digitized system can be appreciated. The authors will be glad to provide further details on request. The equipment in question is being used on a daily basis, and specific modifications can be made when users ask for these. As an example, software changes have already been performed to accommodate requests for multilevel security access codes, for ease of display evaluation, and for ease of completing standard lists of information.

Moreover, once the basic problems of digitization had been successfully overcome, a host of other very interesting possibilities also presented themselves. These included:

• construction of a miniaturized simplified version of subject computerized system, with battery operation and hand held controls for rope sizes of up to 25 mm (1 in.)
• signal displays onto wall size screens
• increase of the presently available top speed of some 3.5 m/s (690 ft/min) to 12 m/s (2362 ft/min), or more, and
• design and construction of a permanently installed, computerized, wire rope monitoring system, with displays and options similar to those of the presently available portable system.

All of the foregoing improvements are well on the way to being implemented.

 

Acknowledgments
The authors wish to thank M. Hamelin, scientist at the Noranda Technology Centre, for his extensive contributions to the research and development work during the early stages of this project.

We also wish to recognize G. Sobkowski's valuable contribution to the project's computer programming, undertaken on the basis of the Canadian Federal Government's supply arrangements with industry.

 

References
Costello, G.A., and J.W. Phillips, "Stress Analysis of Wire Hoist Rope", Univ. of Illinois, Dept. of Theoretical and Applied Mechanics, 104 S. Wright Street, Urbana, IL, Contract J0100011, pp. 1-103, 1983.

Geller, L.B., and F. Kitzinger User's Guide for the Computer Controlled Magnograph II, Heath and Sherwood (1964) Ltd., Kirkland Lake, Ontario, Canada, P2N 3J2, pp. 1-75, 1996.

Geller, L.B., D. Poffenroth, J.E. Udd, and D. Hutchinson, "Evaluation of Electromagnetic Rope Testers: Joint Canadian-US Work," Materials Evaluation, 1992, Vol. 50, No. 1, pp. 56-63.

Geller, L.B., D. Poffenroth, J.E. Udd, and D. Hutchinson, MRL report #90-071, "Canada/NB MDA Project On Mine Shaft Rope Testing; Testing of 1 1/2 in., 6 ´ 30, H.R.," CANMET, Energy, Mines and Resources Canada, 555 Booth St., Ottawa, K1A 0G1, pp. 1-128, 1990.

Geller, Lorant B., K. Leung, and F. Kitzinger "Computerized Operational Control of an Electro-magnetic Wire Rope Tester," Materials Evaluation, 1995, Vol. 53, No. 9, pp. 1002-1006.

Hamelin, M., F. Kitzinger, G. Rousseau, and L.B. Geller "Techniques to Better Exploit the Possibilities of Wire Rope Testing with Permanent Magnet Equipped Electromagnetic Instruments," MINING Technology, 1995, Vol. 77, No. 888, pp. 249-256.

Kitzinger, F. and G.A. Wint, "Magnetic Testing Device for Detecting Loss of Metallic Area and Internal and External Defects in Elongated Objects", 1978, US Patent 4,096,437.

Martyna, R. "The effect of environment and other problems on the magnetic testing
of steel wire ropes," 1997, Proceedings, O.I.P.E.E.C. Round Table Conference on "The Application of Endurance Prediction for Wire Ropes," Univ. of Reading, UK, R.C. Chaplin, ed., pp. 77-88.

Poffenroth, Dennis N., "Nondestructive Testing of Elevator Suspension and Governor Ropes," Elevator World, 1996, pp. 73-75.

"Standard Practice for Electromagnetic Examination of Ferromagnetic Steel Wire Rope," ASTM Designation E 1571-96, 1996, American Society for Testing and Materials, West Conshohocken, PA 19428.

 

* Mining and Mineral Sciences Laboratories, CANMET, Natural Resources Canada, 555 Booth St., Ottawa, Ontario, Canada, K1A 0G1; (613) 992-6792 (lab.); fax (613) 992-2597; e-mail lgeller@nrcan.gc.ca or eleung@nrcan.gc.ca.

+ Noranda Technology Centre, 240 Hymus Blvd., Pointe Claire, Québec, Canada, H9R 1G5; (514) 630-9552; fax (514) 630-9379; e-mail kitzinger@ntc.noranda.com.

Copyright © 1998  by the American Society for Nondestructive Testing, Inc. All rights reserved.

[ Materials Evaluation ]