WO2014200375A1

WO2014200375A1 - Method and system for monitoring electrical wire aging

Info

Publication number: WO2014200375A1
Application number: PCT/PT2014/000037
Authority: WO
Inventors: Abel BORGES FERREIRA MENDES; Fermando António DOS SANTOS SIMÕES
Original assignee: Active Space Technologies, Actividades Aeroespaciais, Lda.
Priority date: 2013-06-09
Filing date: 2014-06-05
Publication date: 2014-12-18
Also published as: PT106994A

Abstract

A method and a system for anticipating potential electric faults in cables, namely high impedance or short-circuit caused by wire and harness aging processes, are described. A set of thin electrodes properly distributed around cables, comprising metallic resistors assembled in series and parallel, is utilized to predict the location of imminent insulation breakdown, namely in branched harnesses. The system provides real-time, automatic, live wiring monitoring of cables. The architecture of the series and parallel resistors is optimized according to the characteristics of the wiring to be monitored; the number of elements is a function of the wire length and the preferred sensor configuration. The electric sensor can be embedded in or wrapped around the cable to be monitored. This invention is suitable for determining the condition of electric circuits of aircraft, vessels, and automobiles, e.g. predicting the location and type of fault.

Description

DESCRIPTION

"METHOD AND SYSTEM FOR MONITORING ELECTRICAL WIRE AGING"

Field of the invention

Wiring problems in aging means of transport such as aircraft, watercraft, trains, and automobiles have been identified as a cause of major faults and accidents. On the other hand, identification of wiring health conditions during routine check-up and maintenance periods can be tedious, lengthy, and sporadically ineffective. Intermittent wiring faults that may occur during operation could be even more difficult to identify. Continuous, real-time monitoring of aircraft wiring performance, without affecting the electrical system, is essential to improve aviation safety standards. Although principally aiming at means of transportation, the concept can be applied to the diagnosis of power systems in the industrial, commercial, and residential areas as well. The present invention relates to the detection of insulation faults in electrical wiring that may lead to undesirable impedance variation, electrical arcing, and short-circuits. A fault may imply catastrophic events leading to short-circuit, fire, electrocution or merely to degradation of the insulating layer. The invention presents a method and a system suitable to determine the wiring condition of aircraft as well as that of other means of transport. Background of the invention

Electric failure often occurs when undesired electrical arcs happen, because the electric strength of air and dielectric materials, albeit large, is not infinite. Under specific conditions, insulation fault can lead to electric breakdown. In the context of this patent, electric breakdown refers to a rapid reduction in the resistance of an electrical insulator that can lead to a spark jumping around or through the dielectric. Electrical breakdown is frequently associated with failure of solid or liquid insulating materials used inside capacitors or high voltage transformers. Under sufficient electrical stress, electrical breakdown can occur in solids, liquids, gases, or even in a vacuum. In the present invention, we are mostly interested in electric breakdown characteristics of air, including variations with density and humidity, as well as in materials used for wire insulation. Convex surfaces with low radius of curvature are more prone to causing dielectric breakdown, because the electric field strength is larger near sharp points. Partial breakdown can occur in gases, as corona discharge on high voltage conductors, at points with the highest electrical field density. A bluish glow and sizzling sound is sometimes detected around high voltage wires or along high voltage power lines . Corona also generates radio frequency noise, a sort of buzzing on radio receivers known as static. Lightning and transient luminous events are a clear manifestation of electric breakdown in the atmosphere. An empirical equation known as Paschen law is used to compute the breakdown voltage, i.e., the voltage necessary to start a discharge or electric arc between two electrodes in a gas as a function of pressure and gap length. The voltage, ¾, necessary to arc across the gap can be computed from

where N is the number density, r is the gap distance, and γ - known as Townsend secondary ionization coefficient - is the net number of secondary electrons produced per incident positive ion, photon, excited or metastable particle. Constants F and G are a function of gas composition. For air at standard atmospheric pressure, γ;¾1.3χ10^~2, and F and G are 3.8X1CT²⁰ m² and l.lxlO^"18 V m², respectively (e.g., Raju, 2003 - Dielectrics in Electric Fields, CRC Press) . Differentiation of Equation (01) with respect to N r yields the minimum breakdown voltage ln(l +

l/y) at Nr_min =-^e In (1 + 1/γ) , where e is the base of the natural logarithm. The voltage necessary to arc across the gap decreases as the density is reduced and then increases gradually, exceeding its original value. In air, at standard atmospheric pressure, the minimum breakdown voltage is about 350 V. For the Loschmidt constant, N_L=2.7xl0²⁵ m^-3, it is obtained a minimum distance of -12 μπι. Figure 1 shows the breakdown voltage for air at standard atmospheric pressure; the intensity of the electric field to reach breakdown is larger than 1 MV m^"1. Insulating solids yield their dielectric strength from covalent bonds and from their much higher density compared to gases.

In automobiles and aircraft, for example, it is important to quantify the effect of humidity on air breakdown voltage. Assuming a perfect gas approximation, the breakdown voltage is a function of not only composition but also pressure and temperature. In addition, the voltage necessary to start an electric arc between two electrodes varies with humidity. In air, the breakdown voltage increases with humidity because water vapour has higher breakdown strength than the typical nitrogen-oxygen mixture. Water also recombines very quickly after dissociation, which increases its breakdown strength. The effect is most perceptible in uniform fields and less important in non-uniform gaps such as around needles. If atmospheric pressure is increased by 0.01 atm due to water vapour partial pressure, then breakdown voltage increases a few per cent. However, surface resistivity decreases with increasing humidity; if the surface is irregular or contains dust particles, humidity favours voltage breakdown. Thus, water vapour thin films, condensation, and water-related corrosion can play a major role in wire aging. In summary, electric breakdown usually follows these characteristics: (i) voltage breakdown increases at very high and very low densities (Paschen law) ; (ii) humidity increases the voltage breakdown of air; (iii) aging by-products, namely acids, deliver charge carriers through dissociation and decrease the dielectric strength of the medium; (iv) the electric strength varies with composition and ionization energy; (v) sharp points are more prone to causing dielectric breakdown; (vi) surface irregularities, water vapour thin films, condensation, and corrosion hamper electric insulation.

Monitoring, mitigation, and protection techniques against voltage breakdown (e.g. short-circuit or fire) have been widely used in electric engines and machinery, power plants, electric power transportation, residential equipment, etc. Many inventions describe apparatus and methods aiming wire protection against short-circuit but only a few address wire health monitoring. In fact, most systems are used for voltage breakdown and short-circuit mitigation rather than error forecasting, i.e., fault detection instead of failure anticipation. In aircraft critical systems, precautionary measures regarding wire health monitoring are important to minimize risk of catastrophic failure. Systems monitoring can be divided according to the method used (either active or passive) and the strategy chosen (either pre-emptive or causal) . Active systems inject a signal in wiring to detect propagation and reflection patterns. Passive systems do not interfere with wiring and simply identify characteristics of propagating signals (e.g. noise) . Although some address active methods, most patents describe passive, causal techniques; the simplest examples include fuses and circuit breakers. The most common active methods include electromagnetic reflectometry (time domain, frequency domain, and spread-spectrum time-domain) , acoustic pulse reflectometry, mutual inductance, and mutual impedance. Passive methods comprise impedance, inductance, photometry, and noise-domain reflectometry . For example, the noise-domain reflectometry technique uses prevailing data signals on wiring and does not require injection of a signal in the monitoring system. The most common technique involves detection of impedance variations, e.g., during open- or short-circuits. Patents disclosed thus far reflect the state of the art described above.

There are many inventions describing technologies and applications for wire protection using fuses and circuit breakers (JP2001196704 , WO2009069209 , JPH05215799,

JPH02218903, JPS62196798, JP2012174632 , JPH06160221,

CN1529179, JP2006194794 , US2006025942 , JP2005317313 ,

JP2002131361, JP2000339572 , JPH0658989, RU2020499, SU1260880, JP2008181561, JPS63187167, US2005275411 , JPH0992046,

JPH06130114, SU1241165, SU1195301, US4135134, US3621384, and CA2458785) . For example, patent JP2001196704 discloses a printed wiring board for breakage sensor, e.g. to increase reliability of safes. When an electric seal is broken by any external force, the conductor pattern on the surface of a board is disconnected, cutting off the flow of constant current so that the breakage of the circuit is sensed. Patent JPH02218903 reports a wiring method for a resistor type sensor using a Wheatstone bridge. Patent JPH06160221 presents a method to obtain the wiring response patterns from a strain sensor. Patent JPS62196798 discloses a wiring system for fire sensors. Patent WO2009069209 discusses short-circuit wiring fixture, a method for measuring skew, and a method for adjusting skew. Patent JP2006194794 shows a short-circuit detector device for terminal device wiring. Patent US2006025942 presents a method and a system for determining the position of short-circuits in a branched wiring system. The distance from short-circuit to an impedance measurement point is determined based on the measured impedance of the branched wiring system. Patents SU1260880, SU1241165, and SU1195301 discuss a method of locating short-circuit in electrical wiring. Patent CN1529179 discloses a method for fault precision positioning in F-shape wiring transmission- line. The method collects data in order to obtain signals of voltage and current at multiple ends of the tested wires, and is based in Kirchhoff law. Patent JP2000339572 addresses a method to discriminate both disconnection and short-circuit of a regional acoustic wiring by applying an inverse voltage of the same value to the wiring.

Inventions and technologies have been adapted to other applications as well, including microelectronics (CN201112800 and JP2001318127) , printed circuit boards (JPH01177951 and KR100266936) , engines (US7011735 and JPH03115871) , audio and voice (CN201548651 and CN101923133 ) , safety regarding heat and explosive environments (EP1473226, JP2001144598 , and JP2001351919) , touch screens and optical displays (CN102231430, CN200979579, US2008024461 ) , and radio wave technology (GB1273432) . For example, patent CN102231430 discusses a method for aging wires of organic electroluminescence displays, where multiple wires are arranged on a large substrate glass. Patent JP2012156411 discloses a short-circuit prevention structure of multilayer wiring board. Patent JP2001318127 presents a method and equipment for inspecting short-circuits in wiring; this patent provides an inspection method and equipment in which a short-circuit part in the wiring of a semiconductor device can be located accurately in the atmosphere . Patent JP2001351919 discusses a wiring fault analysis method based in the heat transfer properties of a medium. The objective is to enable a void shape analysis applied to the reserved part of a plug-attached aluminium alloy wiring, based on heat transfer and diffusion of atoms in a crystal grain structure. This relies in an analysis method of wiring faults accompanied with a form change in a wiring. Patent KR100266936 shows a board-wiring fault detection device that includes an integrated circuit board with a peripheral module and a microcomputer. Patent JPH03115871 reports a fault detector for electric machinery and electric wiring. Patent EP1473226 describes a method and apparatus for protecting the wiring of the safe side of a protective barrier against transferring fault energy, into a potentially explosive environment. Patent CN101923133 shows a method and system for testing wiring fault in microchips. Utility model CN201548651 discusses a voice wiring harness fault detector. Patent GB1273432 presents an apparatus for locating short-circuit in DC wiring networks. The keying of an audio frequency oscillator is used to indicate the presence of pulses in a wiring network. Utility models CN200979579 and CN201112800 show an aging wiring board detecting circuit for electrical apparatus releases, comprising optical and electrical components, namely a microelectronic component aging test device. Although addressing faulty wire and sometimes wire aging, the whole set of patents discussed above are marginally related to the present invention.

Time and frequency domain reflectometry techniques are used in aviation wiring for both preventive maintenance and intermittent fault location (Smith et al . , 2005 - Analysis of spread spectrum time domain reflectometry for wire fault location, IEEE, Sensors Journal, 5, 1469-1478) . The spread- spectrum time-domain reflectometry technology has the advantage of precisely locating faults in long, complex wiring. Additionally, this technology is being considered for real-time monitoring of aircraft electric circuits during flight since spread spectrum ref lectometry works on a live wiring. This method has been shown to be useful to locate to locating intermittent electrical faults (Furse et al . , 2005 - Feasibility of Spread Spectrum Sensors for Location of Arcs on Live Wires, IEEE, Sensors Journal) . These techniques are useful for monitoring, for example, the three-phase 115/200 V, 400 Hz source frequently used in commercial aircraft. However, integrated circuits have been developed for fault detection and location, implying a posteriori detection of the electrical failure.

A few patents from the automotive sector are, to some extent, also relevant to the present invention. Patent US6608486 discusses an automotive test device for detecting short- circuits in automotive wiring. The test device uses two lamps; one lamp is connected in series with the fuse junctions of a particular electrical circuit of the vehicle while the fuse is removed; when a second lamp is illuminated, then everything is normal and the particular electrical circuit of the vehicle is healthy; if the second lamp is not illuminated and the first lamp is illuminated, this means that there is a short-circuit. Patents EP0450807 and EP0450808 disclose a fault detection and isolation for automotive wiring harness. Patent JPS59133796 discusses a fault detector for intensive wiring systems of automobiles. The objective is ensuring an accurate error control by separating a certain terminal control, apart the object control of a central control device part, when the error detecting frequency of the terminal control part exceeds a prescribed level .

A couple of patents from the aeronautics sector are also relevant to the present invention. Patent US2002097056 describes a series arc fault diagnostic for aircraft wiring. The device comprises a capacitive probe designed for clamping to the outer insulation layer of the wire to sense a specific voltage. The device further comprises a floating high- impedance meter, having a ground reference coupled to the source of a common-mode voltage at a second node. The floating high- impedance meter is adapted to measure a voltage difference between the two nodes, being further adapted to indicate the presence of the series fault when the measured voltage exceeds a predetermined level. Patent US2003201780 (also published as US6927579, US2002130668 , and US6777953) discusses a similar system for locating parallel arcing faults in a set of wires. The system includes three devices that can be used in combination or alone. A first device applies a current to a wire while grounding the remaining wires of the set of wires so as to cause the parallel arc. The first device is adapted to locate the parallel arcing fault using one or more leading edges of one or more electromagnetic waveforms being conducted on the wire under test. A second device comprises a controller and two or more receivers, each receiver being electrically coupled to the controller, for receiving one or more leading edges of one or more electromagnetic waveforms being radiated by the parallel arcing fault. A third device senses one or more leading flanks of one or more electromagnetic waveforms as well as the ultrasonic emissions emitted from the parallel arcing fault . Most techniques and inventions currently available are directly related to a posteriori fault detection, i.e., failures in wire/harness disrupt the circuit and their effects are subsequently identified. Fuses are the simplest devices to protect electric circuits. In some cases, circuit breakers are more elaborate because they prevent a major failure from occurring by circuit interruption. Active methods can detect faults, e.g., monitoring radio noise, signal variability, or electric/acoustic echo generated by wiring failure. Nonetheless, advanced, effective monitoring systems should be able to assess wire aging conditions and alert the user before catastrophic faults happen. The most suitable methods and systems for wire health monitoring should: (i) provide pre-fault reliable information, i.e., before faults actually happen, to allow prophylactic action; (ii) minimize interference on signals carried by wires (passive techniques are therefore preferable) ; (iii) offer flexibility for operation during maintenance, routine checkup, and nominal operation; (iv) identify the location including branched harness - where electrical breakdown is imminent; (v) provide easy installation, both in new systems and those already in operation; (vi) offer low cost, low mass, simple solutions. Unlike the systems and methods available, the present invention addresses confidently all these requirements, namely in advance detection of insulation faults that may lead to electrical arcing. Specifically, this patent describes a system and method of diagnostic by which insulation faults in wiring systems can be revealed and located before they develop into arcing faults. Although mostly aiming at detection of wiring faults in aircraft, this method can be applied to other means of transportation and infrastructure as well.

Summary of the invention

Wiring problems in aging structures, namely spacecraft and automobiles, have been identified as a cause of major faults and sometimes accidents. Precocious aging is sometimes associated with thermal stress, humidity-related corrosion, and mechanical chafing. To solve this problem, the present invention combines several concepts . The method proposed in this patent minimizes interference with live wiring, and is based in impedance metering, e.g. comprising a resistive sensor that combines a set of resistors made of a thin filament assembled in series and parallel . The method and associated disclosed system rely on elementary circuit theory, namely Kirchhoff law and Thevenin theorem, as well as resistivity and conductivity properties of conductors. The system can be embedded in or wrapped around wires/harness to improve fault detection. In addition, the system can be used in new and operational cabling, and can detect both open- and short-circuits (e.g. infinite impedance or grounding faults). The system is suitable for monitoring electrical circuits with multiple branches. Nonetheless, the most important advantage of the present system, compared to concepts available, is its relevance for pre-fault detection, unveiling prophylactic capabilities for circuit malfunction assessment. The present system is a low power, low mass, highly miniaturized sensor aiming at not only detection but also prediction of electric circuit problems. The system includes a Sensing Unit (SU) with one or more Series (S) and Parallel (P) resistor meshes, one Multiplexing Unit (MU) whenever two or more branches are included, and a Data Acquisition and Signal Processing Unit (DASPU) . The simplest architecture includes a set of electrodes properly distributed around the monitoring wire, and an SU with sampling and processing capabilities, comprising an analogue- to-digital converter and a straightforward signal processing algorithm. In this patent, the words 'fault', 'malfunction', and 'breakdown' are used to refer to sudden, unwanted variations of impedance in electric circuits, namely degradation of wire insulation, electric arcing, and open- and short-circuits. An individual resistive wiring component, used to build wire grids that cover the sensing area, is designated electrode; the wire grid ensemble is referred to electric sensor or passive sensing network.

The electrodes are made of thin wiring of known resistance and low temperature coefficient of resistivity. Utilization of specific materials, e.g. alloys, whose resistivity is not sensitive to temperature gradients, is important to guarantee that the electric sensors resistance variation with temperature is minimal. Since some faults may produce small variations of the electric sensor equivalent resistance, it is very important to select materials with low temperature coefficient of resistance, because large temperature gradients would otherwise lead to degradation of the system accuracy. This issue is most significant when the system is used for monitoring faults in cables with multiple branches, since higher accuracies would be required. Resistivity, length, and cross-section of the electrodes can be effectively gauged and evaluated to achieve the wanted values of resistance.

Weathering processes, chafing, humidity, temperature cycles, mechanical stress, abrasion from lubricants and hazardous chemicals contribute to the degradation of the quality of wiring and associated insulation layers. When insulating layers crack and bare wire is exposed to the environment or grounded to the vehicle frame - or aircraft fuselage - the risk of current leakage and short-circuit increases significantly. The solution proposed in the present invention is wrapping cables with thin electrodes, which offer specific sensing functions, for inferring wiring health conditions. When the sensing unit is damaged, e.g. fragile electrodes are broken or grounded due to mechanical chafing, pertinent information is sent to the user. Unless extensive damage occurs suddenly, affecting not only the electric sensor but also cable integrity, it is possible forecasting serious faults before they actually damage the cable, i.e., the system can be used for assessing wire aging. The layout of electric sensors must take into account the shape and topology of the cable to be monitored. On the one hand, a set of resistors in series and parallel is necessary to ensure sensing uniform distribution along the cable and to determine where faults do occur. A single resistor could be used to envelope the cable, but its failure would not identify faults location, which is detrimental in long wire bundles. Specific arrangements of electrodes combining resistors both in series and parallel can locate faults with much higher accuracy. On the other hand, a more thoughtful distribution of electric sensors might be able to identify multiple failures of similar or different origin. The efficiency of the fault detection system is a function of the number of electrodes and of the number of resistor nodes in the network, and a trade-off is necessary to adjust sensor sensitivity versus location accuracy. In general, a large number of electrodes help to improve assessment of location and type of failure, but fewer elements usually produce more discernible variations in the equivalent circuit when faults happe .

The DASPU element comprises standard electronic components for signal acquisition, conditioning, and processing, namely one analogue-to-digital converter with at least 12 bit resolution, a central processing unit and memory to store simple algorithms necessary for data analysis. The method and electric sensor architecture are, in fact, the key innovative technologies, since commercially available components can be used for the data analysis and signal processing required by the DASPU. Electric sensor multiplexing is sometimes necessary to distinguish multiple branches in complex harnesses; several MU elements are therefore required to differentiate branches in electric circuits. A simple, effective strategy for identifying branches in an electrical circuit is the insertion of indexation resistors in specific locations to offset the output voltage and, consequently, discriminate the response between branches.

The most efficient way of embedding the electric sensors in new wiring would be during the design, extruding, and electro- facing procedures, which means the electric sensors would be added during the manufacturing, jacketing procedures. This approach improves sensor reliability and reduces costs as well. Cables are jacketed with the electric sensors and another thin insulating layer. To ensure the electric sensor elements do not short-circuit each other, specific wire coating is used. The electrodes are coated with a very thin layer of insulating material, e.g. varnish. One or more polymeric film insulating layers, often of two different compositions to provide a robust and uniform insulation layer, can be applied to the electrodes. For straightforward manufacturing, most wiring can use insulation that acts as a flux when burnt during soldering. Both solid- core and stranded wiring are possible but the electrodes are usually so thin that a solid-core solution is more effective. Preferably, the electric sensors should be installed around the monitoring wire/harness and detect defects before they affect the inner layers. Wire health monitoring may be also necessary for cables already in operation. In this case, an external sensing blanket comprising many electrodes can be crafted, attached to the cable, a posteriori , as a wrapping film. Blankets with embedded sensors are also suitable for monitoring large surfaces, e.g. to detect fractures, cracking . Brief Description of the Drawings

The foregoing aspects and expected advantages of the present invention are briefly described to accompanying the detailed description. A table with important properties of typical metals and alloys used in wire manufacturing is included. To improve legibility of the proposed method and system explained in the description, the following drawings are included, wherein:

Figure 1 elucidates the breakdown voltage of air as a function of gap distance as given by the Paschen law.

Figure 2 shows the equivalent electric circuit of the ribbon layer used for enveloping wires and harnesses .

Figure 3 demonstrates the thin resistive electric sensor (electrodes) used for enveloping wires, cables, and harnesses .

Figure 4 exemplifies a cross section of the thin resistive electric sensor used for enveloping cables and harnesses shown in Figure 3. Figure 5 indicates the equivalent electric circuit of the electrodes distribution shown in Figure 3.

Figure 6 presents the normalized voltage variation with respect to nominal operation as a function of specific wiring faults in one resistor; Sh and Op define short- and open- circuit, respectively.

Figure 7 illustrates the normalized voltage variation with respect to nominal operation conditions as a function of series to parallel resistor ratio and specific wiring faults. Figure 8 presents the normalized minimum resistance between faults as a function of parallel to series resistor ratio for the given topology.

Figure 9 shows a typical architecture for detecting faults in branched harnesses. The boxes identify specific electronics reading units (DASPU) .

Figure 10 presents architecture for detecting faults in branched harnesses, along with the equivalent circuit considering branch indexing resistors.

Figure 11 illustrates a thin resistive electric sensor in a flat band configuration used for wrapping cables and harnesses .

Figure 12 exemplifies a cross section of the resistive electric sensor flat configuration wrapping a group of cables .

Table 1 summarizes a few properties of typical metals and alloys used in wiring. Detailed description of the preferred embodiment

Wiring issues in aging structures, namely aircraft, trains, automobiles, submarines and other watercraft have been identified as a cause of major faults and accidents. Identification of wiring health conditions during vehicle or aircraft routine check-up and maintenance periods can be tedious, lengthy, and sporadically ineffective. For example, intermittent wiring faults that may occur during flight could be even more difficult to resolve. Continuous, live monitoring of cables, without affecting the electric system and interfering with communications and control, is very important regarding aircraft safety. Significant research has been carried out to develop live wire monitoring technology for the aviation, nautical, and automotive industries. The systems available in the market only detect faults after they actually occur, may interfere with the electrical systems of the vehicle/aircraft, and require dedicated assistance from the user. These systems are therefore useful during routine check-up and maintenance, but can hardly determine the position of the fault in branched harness. In this invention it is proposed a passive, interference- free method and system to assess the quality of wiring over time. Unlike other solutions, the present method presents several advantages for wire aging evaluation: (i) real-time, live wire monitoring solutions are offered (suitable for maintenance periods, routine check-ups, and during operation); (ii) automatic, standalone, with minimal assistance from an operator; (iii) sensors are light and can be attached to, sometimes embedded in, wiring; (iv) wire aging forecasting solutions are offered (detection before faults occur, e.g. due to short-circuit); (v) easy installation, both in new systems and those already in operation; (vi) easy identification of the location of imminent electrical breakdown, including in branched harnesses. The formation of aging or stress-induced cracks in the insulation and repetitive condensation, wetting, and temperature cycles, namely in aircraft, increase the risk of electrical breakdown. The effect is particularly dangerous in electric power distribution and electromechanical control circuits of aircraft. Enabling wire aging monitoring, preferably to enhance forecasting capabilities, is then of paramount importance .

Under specific conditions, insulation fault can lead to electric breakdown. In the present invention, we are mostly interested in electric breakdown characteristics of air and polymeric dielectrics, including their variations with composition, density, and humidity. Equation (01), known as Paschen law, is used to compute the breakdown voltage between two electrodes in air. Insulating solids derive their dielectric strength from covalent bonds and their much higher density compared to gases. When insulation of dielectric materials is significantly degraded, e.g. due to cracking in the insulator, the breakdown voltage may decrease significantly. Figure 1 shows the air breakdown voltage versus gap distance as given by Paschen law. The curve is plotted from Equation (01) considering the following parameterization: F=3.8xl0^~20 m², G=l.lxl0^~18 V m², and γ=1.3χ10^"2. Although humidity increases the breakdown voltage of air, formation of thin films on the surface of the cable decreases the insulator dielectric strength because surface resistivity decreases with increasing water condensation. Irregular or unclean surfaces containing dust particles or humidity favour voltage breakdown. Thus, water vapour thin films, condensation, and water-related corrosion hamper electric insulation and play a major role in wire aging.

Electric elements, namely wiring, offer resistance to passage of current. The electrical resistance of a conductor is the opposition to the passage of an electric current through the medium. Conductors are made of high- conductivity materials, namely metals such as copper or aluminium. Resistors, on the other hand, can be made of a wide variety of materials. For many materials and conditions, the voltage and current are directly proportional to each other. Although the resistance can vary with temperature and strain, it is frequently assumed that the Ohm law applies. The resistance of a given metal or alloy depends of its shape as well as intrinsic properties (conduction band) and environmental conditions (temperature and humidity) . An object of uniform cross section has a resistance proportional to its resistivity and length, and inversely proportional to its cross-sectional area. The resistance, R, of wiring materials can be computed from

R = PJ . (03) where p is the resistivity of the medium, and L and A are the length and uniform cross -section of the conductor. For example, a manganin wire of 0.1 mm diameter has a resistance of about 60 Ω per meter.

As discussed above, the electrical resistivity of most materials, namely metals, varies with temperature. When the temperature, T, does not change much, a linear approximation can be used p(T) = p₀{l + a(T - T₀)} , (04) where is the temperature coefficient of resistivity; T₀ and p₀ are a gauge temperature, e.g. room temperature, and the corresponding reference resistivity. When the temperature varies considerably, the linear approximation is insufficient and a more elaborate solution should be used. Nevertheless, this approximation is useful to estimate resistance variations as a function of temperature. From Table 1 and considering ΔΤ-100 K, the variation of resistivity for copper and nickel is 70% and 60%, but the variation for some alloys such as constantan and manganin is only 0.08% and 0.02% for the same temperature difference. Since resistance variations due to faulty electric circuits are usually small, selection of materials with low temperature coefficient of resistivity is quite convenient. In addition to intrinsic variations of resistivity, adjustment of wiring shape, e.g. length and cross section, is fundamental (Equation (03)).

The electric sensor network comprises multiple electrodes arranged in specific configurations to allow wiring health effective monitoring. Although a one-electrode configuration can be used to identify open- and short-circuits, multi- electrode configurations are more effective because they give information of fault location. Wire aging can be associated with two electrical failure scenarios. First, the insulating layer may crack, thus rendering bare wire and electric breakdown more likely; galvanic contact with the structure of the vehicle/aircraft would eventually trigger electric arcing and short-circuit. In the second scenario, wiring is severed and an open-circuit is obtained. Distinguishing between the two types of failure is usually invaluable. To determine the location and type of fault, a more elaborate electric circuit is proposed. Figure 2 shows a typical electric sensor with multiple resistors (electrodes) used for enveloping wires and harness. To be able to evaluate the sensor condition it is used an impedance meter, comprising a controlled current source, forming a voltage divider with the sensor; more sophisticated impedance analysers can also be chosen. Standard signal processing and conditioning units can be used for data acquisition, e.g. employing Analogue-to-Digital Converters (ADC) , as well as for calculation and visualization purposes. The architecture of series and parallel resistors is optimized according to the characteristics of the wiring to be monitored; the number of elements is a function of wire length and of the preferred sensor configuration. Figure 3 shows typical architectures of the electrodes of the electric sensor, comprising multiple windings. The number of turns and pitch is calculated from the electrode length and resistance needs; the number of S and P nodes, or meshes, is determined from the length of the cable, resolution of the DASPU analyser, and of the intended accuracy of the system. The embodiment of Figure 3 presents typical solutions for electrode winding. Linear electrodes are usually lighter but do not cover cables uniformly. Winding electrodes provide better coverage of the surface of the cable. The most effective design, albeit heavier, is reversed zigzagging of the S and P resistors. Figure 4 shows a cross section of the thin resistive electric sensor used for enveloping cables and harnesses shown in Figure 3. This cross section shows the electrodes covering the cable under test as well as the outer layer required for electrical insulation. Figure 5 presents the equivalent circuit of the electrode distribution shown in Figure 3.

The equivalent resistance of the circuit illustrated in Figure 2 can be derived for multiple-electrode architectures using the Kirchhoff voltage law. To a first approximation, since the electric sensors use a DC voltage, the capacitance and inductance of the electrodes can be neglected. The equivalent resistance of a network with N meshes (S-P elements) , R_N, can be computed using a simple recursive method, yielding

Open- and short-circuit conditions in resistors can be approximated by infinite and zero impedance, respectively. The most convenient approach is considering constant S and P values. The resistors S and P work separately as indexation parameters to determine the location failure. For example, if Si=0, V i=l, 2,..., N, it is not possible to locate the failure, because a fault in any P resistor yields the same equivalent resistance. A similar situation happens when the P resistors are discarded. It is the combination of the S and P resistors that allows for fault characterization.

Modelling electrodes response to specific wiring faults is important to explain the measurement procedure and to characterize the response of the electric sensor. Equation (05) can be used to compute the equivalent resistance of the circuit as well as variations due to specific faults, namely open- and short-circuits. For illustration purposes, in this preferred embodiment is considered an electric sensor with five S-P mesh elements. The open- and short-circuit condition means assigning infinite and zero resistance to specific resistances, respectively. In general, it is considered that only one resistance fails at a time. Figure 6 shows the typical resistance difference, normalized to nominal operation, as a function of specific wiring faults in several resistor elements. Stars and squares represent open-circuit and ground- circuit faulty conditions, respectively. Figure 6 shows several important results. First, the resistance difference is always positive or negative for specific failures. Second, failures close to the DASPU analyser, i.e., mesh index close to N, produce larger impedance differences, maximizing sensitivity. Although these values get lower when additional mesh elements are included, the sensor can be designed to include an automatic scaling adjustment so that measurements use the full precision of the ADC. To demonstrate how electrode resistance values can be optimized, variations of the S/P and P/S ratios are also computed. Figure 7 shows the normalized voltage variation with respect to nominal operation conditions as a function of the S/P resistor ratio and specific wiring faults. Stars (top panels) and squares (bottom panels) represent open-circuit and ground-circuit defective conditions, respectively. The left and right panels identify a single fault in either S or P resistors. The subscript number identifies which electrode is at fault. Figure 7 is invaluable to assess the performance of the method proposed for assessing wire aging, specifically the S/P resistor ratio that maximizes system sensitivity. Selection of resistor values may be driven by several requirements. On the one hand, when the number of mesh elements increases, the sensitivity of the sensor must be higher so that discrimination between nodes would be possible. The sensitivity required for the detection of open- and short-circuits is different for the same S/P ratio; thus, the probability of each type of failure may set additional constraints for ascertaining the optimal architecture of the electrodes .

A proposed topology for electrodes assembly is shown in Figure 3. This specific topology consists in: (i) a coated linear wire is mounted on the surface of the cable; (ii) a coated wire is welded to the linear wire and wrapped around the cable; (iii) a third coated wire is wrapped around the cable and the previous wires, and welded to the wire placed in the middle. In this topology, the open-parallel fault can be neglected given that the parallel electrode is protected by the series electrode. In addition, the fault due to series short can be ignored as wiring malfunction would unlikely produce a short behind the series electrodes. Therefore, it is possible to access the ideal P/S ratio that maximizes the minimum difference between faults in order to ease measurement and to better distinguish faults. Figure 8 shows the normalized minimum impedance difference between possible faults for the P/S ratio in the range 10^~6-10⁶. The minimum impedance difference between faults can be increased considering meshes with variable resistance (Pi/Si≠P_j/S_j , V i≠j ) ; however, this performance increase hinders sensor versatility.

Electric circuits are frequently quite intricate in vehicles and aircraft because circuitry comprises many independent branches. In the case of aircraft, for example, cables can be quite long, too. To allow cable branching efficient identification, different strategies can be employed. One option is connecting a DASPU system in each fork, which means the concept described above is replicated multiple times, one on each branch. Figure 9 illustrates the architecture for detecting faults in branched harnesses, where boxes identify specific DASPU elements . Another option could be adding extra resistors (R_c) for delivering branch indexation. Figure 10 presents architecture for detecting faults in branched harnesses, along with the equivalent circuit considering branch indexing resistors . The box identifies the DASPU element; these two circuits are similar to that of Figure 2 . Yet another option could be combining those exemplified in Figures 9 and 10 ; this is the recommended solution for complex, long, branched circuitry.

The strategy discussed thus far is suitable for new cabling and harness, where the electrodes can be easily integrated during cable manufacturing. However, alternative solutions must be found for cabling already in operation, because the approach discussed above would be ineffective. Instead of embedding the electric sensors in cables, a more convenient solution is developing the passive sensing network separately and then wrapping it around the cables to be monitored. Although more expensive and less versatile, this solution extends the present invention to wiring already in operation. Figure 11 presents a thin resistive electric sensor embedded in a flat band configuration, used for wrapping cables and harnesses. For wide sensing blankets the strategy developed for wiring with multiple branches can be applied, i.e., a blanket may combine many electrode sensors and indexation resistors. The distribution of the electrodes on the sensor can be engineered according to specific needs to optimize detection capabilities. Figure 12 shows a cross section of the resistive electric sensor in a flat configuration wrapping a group of cables .

The decision making process is based in Equation (05) . The algorithm needs to assess the measured impedance at a given time and compare it with a gauge resistance (R_0k) · Although more sophisticated programs can be used to classify signal patterns, e.g., neural networks or other high level algorithms with flexible training capabilities, a comparison between measured and nominal impedance provides a robust, straightforward approach to characterize wire aging. The system stores a matrix with the resistances expected for each failure and measures the default value obtained in nominal conditions, which is set as the gauge impedance. Over time, the system contrasts the initial and measured values to determine whether faults have occurred. More elaborate algorithms can be chosen when the number of mesh elements in the circuit increases or the noise level in the circuit is high.

The wire health monitoring system requires easy integration in cables without affecting their performance. The monitoring system shall not interfere with wiring. For this reason, the electrodes comprise the outer layers of the sensing system. This architecture allows for detection of cable degradation before extensive chafing - or other phenomenon - affects the wiring core, which would lead to a major fault. The electrodes must therefore possess appropriate mechanical and electrical properties. The electrodes should be thinner and more fragile than the inner wires. The dielectric sleeve that covers the electrodes should be more susceptible to chafing than the inner dielectric material protecting the cable. As a rule, the outer layers that comprise the wiring health monitoring system (electrodes and insulation layers) should not be more robust than the cables/harnesses they are monitoring. Whenever an external body breaks the protecting sleeve, through continuous mechanical chafing or other phenomena, it exposes the electrodes to the external environment. This situation produces voltage anomalies that can be detected by the wiring health monitoring system.

Wiring used to make the electrodes is coated with a thin layer of insulation. One or more layers of polymer film insulation, often with different compositions, can be applied to provide a robust, continuous insulation layer. The wire coating shall be made of an insulating material with good dielectric properties, e.g., polyurethane , polyamide, polyester, or polyimide. Other types of insulation such as fiberglass yarn with varnish, aramid paper, polytetrafluoroethylene , and polyester film are also widely used. For ease of manufacturing, most wiring can use insulation that acts as a flux when burnt during soldering. This means that electrical connections can be made without stripping off the insulation first, simplifying electrodes soldering. The S and P electrodes can be made of single-piece coated wires revolving around the cable in a concentric, preferably reversed, zigzag with appropriate pitch. At specific locations, coating is stripped and the wires welded together or to a third linear wire (in case of P) that provides grounding, nesting an electric circuit similar to that shown in Figures 2 and 3. Most likely, the easiest way of welding the electrodes to each other is the following: (i) a coated, linear wire is mounted on the surface of the cable; (ii) a coated wire (P electrodes/resistors) is then wound around the cable and welded to the linear wire in specific locations; (iii) finally, a third coated wire (S electrodes/resistors) is twisted around the other two wires and welded to P nodes at designated locations. In a different embodiment, the wire mounted on the surface of the cable to provide a voltage reference can be replaced with a thin foil, which would also have guard purposes. Along with the cable, the electrodes are subsequently jacketed with a thin insulating layer. Electrode coating involves two complementary processes. First, the conductors must be insulated to avoid accidental short-circuits between them. This usually means coating electrodes with varnish or another insulating thin film, similar to the processes used in transformers. Then, a second layer, e.g. polyimide or fluorocarbon plastic tape, is applied to insulate the electric sensor from the environment. Ideally, the polymeric layer should be fire-resistant and do not possess/release harmful chemical species such as chlorine.

Claims

CLAI S

1. A method for monitoring wire aging, forecasting and locating electric faults in cables and harness, and characterizing circuit failure, each fault producing variations of impedance in the circuit, the method comprising :

Selecting an impedance meter to measure the impedance difference of the device under test, where the device under test can be a wire, cable, or harness;

Choosing appropriate characteristics of a signal to be injected in the electric . circuit of the device under test; and

Selecting a suitable measurement range of the impedance meter .

2. The method of Claim 1, further comprising:

Selecting appropriate architecture of wiring elements designated electrodes - to define resistance of the same;

Choosing suitable length, diameter, and resistivity of the electrodes to achieve optimal resistance values; and

Picking materials with appropriate electrical insulation properties to cover each electrode with a cladding thin film.

3. The method of Claim 1, further comprising: Deciding efficient distribution of electrodes, connected in series and parallel, to discriminate open- and short-circuit faults;

Picking electrodes of specific resistances to create the electric circuit wiring mesh by means of soldering series and parallel elements to each other or to the ground; and

Jacketing the device under test and the set of electrodes with a thin cover layer of insulating material .

4. The method of Claim 1, further comprising:

Measuring the equivalent resistance of the electric circuit of the device under test to compare results against reference values; and

Selecting proper decision making algorithms to forecast and locate potential wiring faults.

5. A system for monitoring wire aging, forecasting and locating electric faults in cables and harness, and characterizing circuit failures such as those from open- and short-circuit, each fault producing variations of impedance in the circuit, the system comprising:

A set of electrodes attached to the device being monitored; An impedance meter;

A processing unit; and

A piece of software.

6. The system of Claim 5, wherein the set of electrodes are connected to each other in a pre-determined cylindrical geometry, the electrodes are made of thin wire with low temperature coefficient of resistance, and each electrode is covered with an insulating film.

7. The system of Claims 5 and 6, wherein the electrodes are associated in series and parallel, and resistors of specific resistance are used for indexing purposes in the electric circuit.

8. The system of Claims 5 and 6, wherein the set of electrodes that envelope the cable under test is wrapped with a thin dielectric layer.

9. The system of Claim 5, wherein the set of electrodes are connected to each other in a pre-determined configuration and embedded in a thin dielectric layer.

10. The system of Claim 5, wherein the impedance meter comprises analogue- to-digital conversion, signal conditioning and processing, memory management, and communication capabilities .

11. The system of claims 5 and 10, wherein the piece of software handles, processes, compares, and contrasts realtime data with reference impedance measurements stored in memory, and yields information to de user about wiring quality of the cables under test .

12. The system of claims 5, 7, 10, and 11, wherein resistors with specific resistance are included in the electric circuit to splitting and offsetting the voltage measurement range, and distinguishing between faults in different branches.

13. The system of claims 7, 10, 11, and 12, wherein various sets of electrodes, impedance meters, processing units, and pieces of software are combined together to assess, simultaneously, the wiring quality of multiple cables or branched systems .