NON-DESTRUCTIVE AND ON-LINE METHOD AND APPARATUS
FOR DETERMINING THE MECHANICAL PROPERTIES OF
STAINLESS STEEL CABLES
Field of the Invention
This invention relates to the measurement of the mechanical properties of stainless steel (hereinafter, briefly, SS) cables and particularly to determining failures in the mechanical properties of such cables due to stress in use, and to detecting fatigue damage that might cause failure of the cables.
Background of the Invention
US 5,619,135, of April 8, 1997, discloses a steel hardness measurement system and method of using the same for measuring at least one mechanical or magnetic characteristic of a ferromagnetic sample as a function of at least one magnetic characteristic of the sample. In said patent, the steel sample is preferably rolled sheet steel, i.e., steel having a thickness of approximately 2 mm or less. Said patent provides means for testing only a sample of the sheet, but, in order to test a significant part of its surface area, a very large number of sensors is required. A complete test of sheets is not practical according to US 5,619,135, as it mainly relates to the testing of stationary steel sheet samples; and the tests of steel structures other than steel sheets is not considered therein.
PCT application WO 98/20335 describes and claims a method for measuring the mechanical properties of ferromagnetic, elongated structures, particularly cables, which comprises the following steps:
- creating at least a magnetic field that varies either as a function of time or the function of position or both, along at least a given length of ~ cable;
- for each section of the cable to be tested, sensing a plurality of Barkhausen Signals due to the time or position related variation of said magnetic field;
- measuring the Barkhausen Signal rate as a function of the value of the applied magnetic field for each said section; and
- determining the desired mechanical properties, particularly yield strength and tensile strength, of each such section from said sensed Barkhausen Signals relative to it.
The aforesaid PCT application also describes and claims an apparatus for measuring the mechanical properties of elongated, ferromagnetic structures, particularly cables, which comprises means for generating a magnetic field that varies along at least a given length of the elongated structure; a plurality of Barkhausen Signal sensors distributed along said structure length; and means for comparing the signals relative to each section of the structure to be tested.
The austenitic stainless steels, specifically the AISI 3XX type family of steels, are composed of mostly a non-magnetic phase called austenite with minor quantities of magnetic phases like δ ferrite or α and α1 martensite. The austenite in these steels is not a completely stable phase and transforms under certain conditions - low temperature or mechanical (plastic) .deformation - to the more stable phase, martensite, which is a ferromagnetic phase, whereas the austenitic phase is paramagnetic. This transformation has been documented in the literature, see e.g. Hecker, M.G. Stout, K.P. Staudhammer, and J.L. Smith, Metallurgical Transactions, 13A,
619-635 (1982), in which this transformation is studied in 304 SS by several methods, and it is shown that the percentage of the martensite phase increases as a result of strain, up to 85% under certain conditions. Its appearance is shown to be associated with a decrease in the work hardening rate with possible premature local instability.
Stainless steel wire ropes are manufactured using either cold or hot drawing techniques. In either case, the product contains varying, but small, amounts of a magnetic phase. A typical Barkhausen Signal (hereinafter, briefly, "BS") from a typical AISI 302 SS wire is presented in Fig. 1. Subjecting the wire to external loads will cause a phase transformation which will lead to the formation of martensite, i.e., more magnetic phase will be present in the microstructure of the cable. It is a well established fact that deformation induced martensite significantly enhances strength generated by cold work. The amount of martensite in the material can be monitored using existing laboratory equipment (e.g., the Ferritoscope) with low accuracy. This measurement is performed off-line, on a small sample, and is destructive, because a sample has to be taken from the cable or the wire.
It is a purpose of this invention to provide a method and apparatus for checking steel, in particular stainless steel, cables and wires that constitutes an improvement on the methods and apparatus of the prior art.
It is another purpose of this invention to provide a method for checking steel, in particular stainless steel, cables and wires that is not destructive.
It is a further purpose of this invention to provide means for testing the complete volume of a cable, and when necessary, for testing the whole length of the cable.
It is a still further purpose of this invention to provide a method and apparatus for checking steel, in particular stainless steel, cables and wires on line.
It is a still further purpose of this invention to provide such a method and apparatus that are simpler, quicker and more convenient than those of the prior art.
It is a still further purpose of this invention to provide such a method that does not require the quantitative evaluation of the Barkhausen signals at each test.
Other purposes and advantages of the invention will appear as the description proceeds.
Summary of the Invention
It has been found, and this is the basis of the present invention, that the change in the amount of the magnetic phase (martensite) in a stainless steel wire, cable, or wire rope, can be monitored through the Barkhausen Signals (BS) and correlated to the change in its mechanical properties. Hereinafter, when reference is made to a cable, it should be understood that what is said applies to wires and wire ropes as well. The inventors have determined that subjecting, for example, an AISI 302 wire to low cycle fatigue, while increasing dramatically the amount of magnetic (martensitic) phase in the wire, causes a concurrent change in the BS from the cable. A prominent
feature of this change is the appearance of the intensity (clearly seen as a peak, such as in Fig. 2) on the low field side of the original BS peak shown in Fig. 1.
The change in microstructure is clearly evident when comparing the BS peak of the original wire (Fig. 1) to that of a wire which has been fatigue cycled, i.e., submitted to severe back and forth bending cycles (Fig. 2). A new peak on the low magnetic field side can be clearly observed in this figure. This peak is attributed to the increased fraction of martensitic phase. This martensitic phase appears in clusters which are the result of local ordering of dislocations which, at certain density and arrangement, transfer a region of austenite into a martensitic phase. These clusters are therefore, initially, free of dislocations, and consequently the magnetic domains inside them can move freely, resulting in low coercive force and low field BS, which is manifested by the additional low field peak in Fig. 2 (in the figure, to the left of the peak of the original wire). As more strain causes the formation of more martensite, this peak grows in intensity relative to (and instead of) the original peak. The additional strain, furthermore, causes hardening of the newly formed martensitic phase, thus resulting in shifting of some of the BS intensity to the right (to higher fields) towards the original peak position.
There are, therefore, two competing processes: one contributing to higher intensity on the low field peak, and the other ultimately causing its disappearance and merging with the original peak. The inventors have experimentally proved that the balance of these two processes results in a growing contribution to the low field peak, while causing its broadening, until the premature breaking of the wire. This, obviously, happens before the disappearance of the low field peak, because a premature failure will
WO 99/56122 .β. PCT/IL99/00152 occur in a region where the existence of the martensitic phase resulted in low ductility, while there is still a large percentage of the martensite with low dislocation density.
Since stainless steel (SS), as its name suggests, is not prone to develop corrosion, visual checking does not give much information about its mechanical state. Being mostly non-magnetic, the flux leakage methods are not useful (and not accurate) for detecting LMA (Loss of Metallic Area) or broken wires, as is done with ferromagnetic cables. But SS cables are susceptible to most failure mechanisms described in said PCT application WO 98/20335, as much so as any other metallic cable, and its strength cannot be measured on line with any available method.
This invention provides a method and apparatus for determining the mechanical properties of SS cables or wires, especially of the AISI 3XX type, that is applicable on line, both in production and in use. Hereinafter, for brevity's sake, only the term "cable" will be used, but what is said should be understood as applying to wires as well.
The method according to the invention comprises comparing the amount of martensitic phase induced by strain to the amount of the original magnetic phase in the cable being tested. The original magnetic phase is determined by taking out a baseline reading of the BS for a master cable. For each family of stainless steel cables to be tested, the master cable is the same cable to be tested, but as it was when new, either prior to or right after installing the cable in service, or is an equivalent cable. Said baseline reading can be called "BS calibration". By "equivalent cable" is meant herein a cable which is sufficiently similar to the cable being tested as to give essentially the same baseline reading that this latter would have given
when new. A cable obtained from the same production hne from which the cable to be tested has been obtained, is in general a satisfactory equivalent cable. The amount of martensitic phase induced by strain is determined, preferably at regular intervals and at any rate whenever the cable is to be tested, by taking out a successive reading or readings of the BS of said cable.
The comparison of the amount of martensitic phase induced by strain to the amount of the original magnetic phase is effected by determining a first value of a parameter which represents the amount of martensitic phase induced by strain to a second value of the same parameter which represents the amount of the original magnetic phase, and determining the ratio of said two values. Said parameter will be called hereinafter "phase parameter" (hereinafter sometimes abbreviated as Pp) and their ratio will be called "the martensite index". The martensite index may be: a) the ratio R between the two area integrals relating correspondingly to the two peaks obtained the two relevant BS readings (viz. the reading of the cable at the moment in which it is tested and the readings of the same cable when new or of an equivalent cable); b) the ratio defined by R=Z(N)/Z(I), while the function Z is defined by Z = ∑(uι - Vi)2 u and v being the BS intensities determined in the two relevant BS readings for the same field value; c) the ratio R=U/V, wherein U and V are parameters expressing BS intensities, as hereinafter specified d) the ratio of the values of the FWHM (Full Width at Half Maximum) of the BS peak of the two relevant readings;
e) the ratio of another phase parameter that is a statistical index pertaining to the shape of the peaks (such as its standard deviation, kurtosis, etc.) of the two relevant readings; f) the ratio of a combination of the above parameters.
Each individual reading is preferably carried out by the method and by the apparatus described in WO 98/20335, and hereinbefore summarized.
Preferably, according to the invention, the chosen martensite index is related to the number of fatigue cycles. This can be done by submitting a master cable, as hereinbefore defined, to a number of fatigue cycles up to its failure, and determining the martensite index at successive numbers of fatigue cycles. A diagram - hereinafter, "the fatigue diagram" - can thus be constructed having the number of fatigue cycles as a first coordinate (e.g. the abscissa in Fig. 3) and the martensite index as the other coordinate. When another cable is tested, its martensite index is determined and the fatigue diagram is entered using it as the first coordinate, whereby the other coordinate will indicate the number of fatigue cycles that the cable has been submitted to, and will show how far it is from failure.
Brief Description of the Drawings
In the drawings:
Fig. 1 is a BS diagram of a section of a new AISI 302 cable;
Fig. 2 is a BS diagram of the same section of cable after a few cycles of fatigue (i.e., bending);
Fig. 3 is a diagram illustrating the variations of the martensite index
(hereinafter defined) caused by fatigue cycles;
Figs. 4a, 4b and 4c schematically illustrate an embodiment of apparatus for carrying out the invention;
Fig. 5 schematically illustrates another embodiment of said apparatus; and Fig. 6 is a block diagram illustrating the method and the apparatus according to an embodiment of the invention.
Detailed Description of Preferred Embodiments
The inventors have found that stressing a SS wire or cable (i.e. an AISI 302 wire) to low cycle fatigue by bending not only dramatically increases the amount of magnetic (martensitic) phase in the wire, but causes concurrent changes in the BS from the cable which can be quantitatively related to the cable's strength or to the expected remaining life of the cable.
A prominent feature of this change is the appearance of intensity (a peak in some cases) at the low field side of the original BS peak measured as a function of magnetic field. The comparison of Figs. 1 and 2 shows this clearly. Fig. 1 depicts the measured number of BS as a function of applied magnetic field as measured for a new cable made of AISI 302 SS. Fig. 2 depicts the results for the same cable after it has been subjected to about 90 cycles of fatigue, in which the measurement of the BS was taken at the point subjected to the fatigue; while at any other point of the cable, the result was similar to that shown in Fig. 1.
The Barkhausen Signals of the tested cable can be measured in any suitable way, and particularly in any one of the ways described in said WO 98/20335. For the sake of illustration, one such apparatus and method will now be described
The magnetic field, which generates the BS, may vary either as function of time, as a function of position, or as a function of both. A device for measuring the field strength, such as a Hall probe can be used. The field can
be produced by at least an electromagnet fed with an alternating current, at least a sensor being provided in a fixed position with respect to the electromagnet, the cable being stationary or moving with respect to the apparatus unit constituted by the electromagnet and the sensor, and the time (or cable speed of movement), with respect to which the BS is measured, being recorded.
The apparatus, referred to herein by way of illustration, schematically showing Figs. 4a, 4b and 4c, comprises a number of permanent magnets 10. They are at least two, they fully encircle the cable (not shown) and are magnetized parallel to their axis. They are arranged along the path of the tested cable (not shown) with opposing magnetic poles facing each other, as designated by S and N in the drawing. A few Barkhausen sensors 11 (at least four) are spaced equally between the magnets along the axis of the apparatus, viz. the tested cable path, at positions of known, predetermined values of the applied magnetic field. The magnets are split so that it is possible to open the apparatus, consisting of two halves 13, in order to encircle the tested cable. A Relative Cable-Sensor Speed (RCSS) measuring unit 14 is placed inside the apparatus housing to measure the speed of the cable relative to the sensor. Both the magnets and sensors are mounted on adjustable mounts so as to fit various cable diameters.
Another apparatus, slightly different from the previous one, is schematically shown in Fig. 5. It contains at least two DC electromagnets 20, Barkhausen sensors 21, and RCSS 22, mounted on a casing 23. This apparatus cannot be opened SQ, as to be mounted around a cable, and is therefore intended to be used on the production line of wire ropes (not shown), or at any other installation where the tested wire rope can be inserted through the apparatus.
The apparatus used in carrying out this invention, therefore, comprises:
- means for generating a magnetic field that varies along at least a given length of the elongated structure;
- a plurality of Barkhausen signal sensors distributed along said structure length;
- electronic system for detecting, amplifying and transferring the BSs to a computer;
- means for comparing the signals relative to each section of the structure to be tested;
- computer means for elaborating the said signals to determine parameters defining the martensite index from said sensed Barkhausen signals: and, optionally,
. memory means for storing said parameters measured on the new or equivalent cable and on the cable to be tested, referred to the conditions under which they have been measured; means for calculating the martensite index; and means for calculating the number of fatigue cycles that a tested cable has undergone.
As shown in the diagram of Fig. 6, the master cable or the tested cable are sensed, as indicated at 30 and 31 respectively, in an apparatus 32 to determine the Barkhausen signals. Numeral 33 generally indicates computer means, to which the BS are transmitted. The computer means 33 calculates from the transmitted BS the phase parameters (chosen among the various Pp that may be used, as hereinbefore set forth) of the master cable or ihe tested cable, as the case may be.
When the master cable is sensed, the variations of the chosen Pp are due only to fatigue cycles to which the master cable has been deliberately
submitted, and the fatigue cycle counts are inputted into the computer, as indicated at 35. From the Pp and the corresponding fatigue cycle counts, the fatigue diagram of Fig. 3 is constructed by the computer, as indicated at 36.
When a tested cable is sensed, the chosen Pp is similarly derived from the BS. By comparison between it the corresponding Pp of the master cable before it was submitted to fatigue cycles, the martensite index of the tested cable is computed, as indicated at 37.
Said martensite index is then compared to the said fatigue diagram, as indicated at 38, by entering it as the ordinate and determining the number of fatigue cycles that corresponds to it. This does not mean that the cable's martensitic phase has necessarily been generated by fatigue cycles. It may have been generated by other causes, but by determining the number of fatigue cycles that corresponds to it, and therefore the residual number of fatigue cycles that the cable could still be submitted to before causing failure, the state of the cable can be evaluated, and, in the light of the conditions to which it has been submitted in its previous service, its residual expected life under the same conditions can be estimated. The fact that the output of the computer 33 is indicated at 39 as "number of fatigue cycles" should be understood in the light of what has been said, viz. as meaning an index of the state of the cable and of its expected residual life, whether the cable has been and/or will be stressed by fatigue cycles or by any other cause.
The martensite index, which is correlated to the change in the mechanical strength of the cable, is calculated by any one of the methods described below. Such a change may be due to successive fatigue stresses, and this is the case to which particular reference is made herein; however, it could be
due to another cause, and the method of this invention will be applicable in the same way, though the diagram relating the martensite index to the cause of the change in mechanical properties will not be a fatigue diagram.
A different martensite index will correspond to different strengths of the cable. The result, for an increasing number of fatigue cycles, is shown in Fig. 3, wherein the martensite index has been calculated by method 2, hereinafter defined. In this particular example, each point of the graph is an average of three different measurements - a number of different measurements can be used in any embodiment of the invention - and the measurements were taken at predetermined intervals of the same number of fatigue cycles. If the fatigue cycles cannot be numbered, the measurements will be taken at predetermined time intervals. The ordinates of a graph such that of Fig. 2 or Fig. 3 indicates a total amount of deformation induced by the amount of the magnetic phase present in the cable and hence permits to estimate its condition.
An index R, which is what has been called "martensite index", for the condition of a tested cable can be determined by one of the following procedures:
1. By calculating a ratio R between the two area integrals relating correspondingly to the two peaks obtained in one measurement of BS signals relating to the tested cable (such as of Fig. 2). The two peaks are: the higher, central peak in Fig. 2, that is similar to that found in testing the master cable (the peak of Fig. 1), and the new peak - the left-hand, lower peak - of Fig. 2, which did not appear in the measurements of the master cable. The border points, between which the integration for each peak is performed, are: for the lower (left-hand in Fig. 2) peak, from the beginning
of the diagram (lowest magnetic field strength) to the minimum between the two peaks (point I in Fig. 2); for the higher (central in Fig. 2) peak, from the minimum between the two peaks (point I in Fig. 2) to the point in which the BS intensity has decreased to background noise value.
2. By calculating a ratio R=Z(N)/Z(I), while the function Z is defined by
Z = ∑(uι - vi)2 u being the result of the measurements on the tested cable and v being a the results of the corresponding measurements on the master cable. Ui and Vi are therefore two measurements performed for the same "i" field value. The said calculation can be carried out for all the relevant values N, or for only the first M values. I is the point dividing the two peaks, such as is indicated in Fig. 2.
3. By calculating a ratio R=U/V wherein U is the BS intensity measured at a predetermined field value, which is presumed to be the point at which the maximum of the first peak (at a lower field value) occurs in the BS diagram of the tested cable and V is the BS intensity measured at a field value where the maximum of the second peak (at a higher field) occurs in said diagram.
4. By measuring the FWHM (Full Width at Half Maximum) of the peak, which is the width of the peak at half its height.
5. By calculating a ratio R of the values of any other statistical index pertaining to the shape of the peak (such as its standard deviation, kurtosis, etc.), for- the tested cable and the master cable, respectively.
Any one of said R indicators, or a combination of them, can be selected for use, according to the existing conditions, or as is convenient.
As stated hereinbefore, the BS readings are obtained periodically and each obtained in one of the resulting readings is analyzed by the following procedure. For example, if the first of the above procedures is used, the area of the left-hand part of the plot, which includes the peak, is measured. The ratio R for the measurement is determined and compared to a reference curve such as that of Fig. 3, previously prepared for the same cable, or for a master cable, indicating the R ratio for the master cable for different number of fatigue cycles. The expected lifetime, before failure of the cable under consideration, can be calculated if a constant fatigue rate is assumed, when the ratio for the same rate of fatigue cycles is compared between the master cable and the tested cable. More particularly, if a ratio Ri is calculated for the tested cable, and the master cable is known to have the same ratio Ri after, for example, 140 fatigue cycles, it can be assumed that the tested cable has been passed about 140 fatigue cycles during its period of use.
Of course, in order to improve accuracy, the calculation of the expected lifetime can be made by considering the ratio determined by one of the above procedures, or by averaging the ratios determined by two or more of the above procedures.
While examples of the invention have been described by way of illustration, it will be apparent that the invention can be carried out with many modifications, variations and adaptations, without departing from its spirit or exceeding the scope of the claims. In particular, any way of measuring the BS of a cable described in the aforesaid PCT application WO 98/20335, or any other suitable way of measuring them, can of course be employed to carry out the invention.