8/7/2019 DOE efforts 1719645

1/68

-

Final ReportDepartment of Energy GrantNumber DE-FG02-98ER82543, Entitled

The Development of Shortwatch, a NovelOvertemperature or Mechanical Damage SensingTechnology for Wires or CablesBPW, Inc.September 7,2001

Ken Watkins, JackMorris, C. P. Wong,and Shijian Luo

DOE Patent Clearance Grantedp B w q/*Lq.oLMark P.Dvorscak Date_ _630) 52-2393L -mail, mark.dvorscakOch.doe.govOffice of Intellectual Propert LawDOE Chicago Operations OMce

1

http://mark.dvorscakoch.doe.gov/http://mark.dvorscakoch.doe.gov/

8/7/2019 DOE efforts 1719645

2/68

. Ttris was p q a d as an accollnt of work spoasorrd by an agency of theUnited States Gwcmneat Natber tbe United States Oovetnrntllt nor ury -cyth#eof, nor any of their c m p l o v ~ ,makes any warranty, express or implitd orassnmcs ahy kgal k W t y or rrspoasib%ty for the tcaulcy, tampIetenett, or we-fdnets d ny information, apparatus, prcduct, or prcccu ddoscd, or -&tbat its use would not infringe prnltdy owed rights. Rcfaence herein to m y spe.cifi 0ommcrcii-d prodact, procsss, or service by trade name. trademark, manofic-tu=, of othmrisedocs not auxtsarily constitute or imply ts cadonemcot, tccom-mendatha, or favoring by tbe United S 2 t a Govcmmcnt or any agaxy thmof.The riews and opinions of authors cxpwscd bmin do not a##sarily natc orrcfkct hose of the Uaitcd States Govcrpmwt or my agency thtrrof.

.

t

8/7/2019 DOE efforts 1719645

3/68

8/7/2019 DOE efforts 1719645

4/68

Table of ContentsTitleIntroductionStatement ofProject Requirements and GoalsProgram SummaryFirst Year EffortsSecond Year EffortsGoals and Follow-on WorkSummary and ConclusionsAppendix A: Proof-of-Concept TestAppendix B: Georgia Tech Conducting Polymer Materials StudyAppendix C: BPW AgingTrials of Shortwatch SensorsAppendix D: Time Domain Reflectometerand Noise IssuesAppendix E: Certified Test Report on Shortwatch CableAppendix F: References

2

Page

334451 1 13141852606466

8/7/2019 DOE efforts 1719645

5/68

IntroductionThis report documents the Department ofEnergy sponsored SmallBusiness InnovationResearch (SBIR) Grant number DE-FGO2-98ER82543, which isentitled, TheDevelopment of ShortWatch, a Novel Overtemperature or Mechanical Damage SensingTechnology for Wires or Cables.

Statement oProjectRequirements and GoalsIn response to the Department of Energy SBIR solicitation item Number 32. AdvancedTechnologies for Commercial Nuclear Power Plants and Space Power Plants and inspecific subtitle C. Advanced Characterization Techniques and Modeling Methods toPredictAgingBehavior of Electrical Cable Jacket and Insulation Materials, BPWIncorporated is proud to present its final report for the development of ShortWatch, apatented overtemperature or mechanical damage senshg technology for electrical wiresor cables.The ShortWatch technology addresses the problems identified in the above-mentionedDepartment of Energy solicitation, which states, Nuclear plant electric cables are subjectto various types of environmental aging, predominantly thermal and irradiation aging.Cables in critical applications are required to hction during and after an accident. Thefunctional performance of a safety circuit may be lost ifjacket integrity or insulationdielectric strength is inadequate. Grant applicants were sought for technologies thatprovide advanced techniques suitable for in situ testing of a cable to (1) characterizecable jacket and nsulation material condition, (2) predict the remaining life of a cablebased on science-based aging models, and (3) assess the abilityof a cable to perform itssafety function(s) during design basisevents. Nondestructive techniques capable ofcharacterizing the conditiodperfomce of an entire cablerunare of primary interest;techniques that require small samples fiom the endsofcables or very localized samplesare also of interest. For condition monitoring techniques that require samples, methodsfor obtaining samples without compromising cable qualifications are required.ShortWatch developments described in this final report provide an in situ technology,which addresses items (1) and (3) above by providing a real time, mechanical damageand over temperature sensing capability fbr wire and cableused in the nuclear industry.

3

8/7/2019 DOE efforts 1719645

6/68

Program SummaryThis project was completed on time and within its budget. Project goals weredemonstrated in successful ests conducted on prototype Shortwatch cable withintegrated mechanical damage and over temperature sensors at the RockbestosSurprenant Cable Corp. facilities on May 16,2001.First Year EffortsProgram results for the first year include:1. Selectionof Shortwatch overtemperature sensor recipe. Sensor tape extrusion trialsat Rockbestos confirmed mechanical and electrical specifications determined inPhase 1of the SBIR. Of several polymers studied and tested, the blend of HDPE and LDPEutilized in a commercial environmentally qualified wire was chosen as the base polymerto introduceas few new variables as possible. The sensor recipe also included a low-structure carbon black and anti-oxidants. Appendix B documents conductive compositestudies conducted by the School of Materials Science and Ehgineering at the GeorgiaInstitute of Technology as part of thisproject.2. Wire design completion A multi-conductor wire design was chosen incorporating ahybrid design of Shortwatch which features center filament of Shortwatch conductivepolymer sensor (figure 1)that provides overtemperature-sensing capability. A helicalwrap of Shortwatch metallic ribbon sensor provides mechanical damage sensingcapability. This design was chosen due to the broad applications multi-conductor cables.withShortwatch capability will provide in a plant environment. A second designincorporating a co-extruded Shortwatch sensor in the insulation of a wire or cable had tobe deleted fiomPhase2 SBIR due to a flrst year fimding reduction The co-extrudeddesign d l e pursued furtherupon additional funding.3. Shortwatch sensor filament extrusion. A sensor filament approximately0.5 mm indiameter, which meets the dimensional requirements of the selected hybrid Shortwatchcabledesign, was extruded at Rockbestos and subjected to mechanical and electricaltesting at GeorgiaTech and BPW, Inc.

4

8/7/2019 DOE efforts 1719645

7/68

Figure 1: Shortwatch Sensor Filament4. Cross-linking effects. Positive results on sensor te mpera tu re -resi e response atswitching were observed by electron beam irradiationof the Shortwatch sensor. Cross-l i i g liminated the negative temperature coefficient (NTC) response of the matrixobserved at temperatures above the switching (transition) temperature (see Appendix B).5. Fault location investigations. In-house testing indicated that a Time-DomainReflectometer (TDR) is able to determine the locationof mechanical damage toShortwatch-capable wire. The locationof the damaged area canbe determined evenbefore the damage is significant enough to result in an electrical fhult in the wire. Othermethodologies were investigated, including standing wave ratio (SWR) echnology.6. Other investigations. Prelimbary investigations included controller design studies todetermine general electronic circuitry which would be used to monitor Shortwatch overtemperature and mechanical damage sensors, age investigations into accelerated agetesting which would demonstrate the abiliy of the Shortwatch sensors to hc ti on overan extended time period and noise immunity investigations to determine electricalnoisegenerated in Shortwatchsensors iom energized conductors protected by the sensors.

SecondYearEffortsProgramresults during he second year of the project included:1. Fault detection.Fault detection may be broken down into overtemperature fault detection and mechanicaldamage detection.

5

8/7/2019 DOE efforts 1719645

8/68

Overtemperaturehult detection by Shortwatch is distributed in nature by an extrudedstrip or filament of conductive polymer, which switches at the desired overtemperature.The critical Shortwatch design parameters are switch temperature, switch magnitude anddesign resistivity. The switch temperature is determiued by the basepolymers utilizedand, in this case, hebase polymer of a commercial environmentally qualified wire isused to provide a maximum switch temperature of approximately 130C,or about 40Cgreaterthan he design temperature of the wire.The switch magnitude is the ratio of sensor resistance at a maximum switchtemperature divided by the design (90C) resistance. Although a switch magnitude of100 is considered acceptable asfaras providing reasonable resolution of overtemperatureconditions, the design aim isa switch magnitude of approximately 1000. Thismagnitudewill provide anovertemperature resolution of approximatelyan inch in 100 feet ofShortwatch cable.A trade-off exists between switch magnitude and design resistivity (resistivity of theovertemperature sensorat 9OOC). Studies performed at Georgia Tech (see AppendixBand Figure 2) showed a generally higher switch magnitude at high sensor resistivity.Although switch magnitudes of over 1000 were obtained in Shortwatch samples, lower(approximately 100)initialdesign switch magnitudewas selected inorder to reducesensor resistivity in order reduce the size (diameter) of the sensor filament.

Sho~WatchAlrtment SensorTemperatureRespame

Figure 2: ShortwatchFilament Sensor Resistance-Temperature ResponseA helically wrapped metallic ribbon or wire provides the mechanical damage sensing ofthe Shortwatch multi-conductor cable. The ribbon is wrapped around the insulatedconductorsof the multi-conductor cable and inside the outerjacket of the cable. Cuttingor abrading the outer jacket will open he ribbon befbre the conductor insulation ispenetrated. Shortwatch instrumntation senses the loss of sensor ribbon continuity,initiating analarm or control action.2. Fault location.We have experimentedwithaTimeDomainReflectometer (TDR), shown in Figure 3,which determines the location of mechanical damage in Shortwatch capable electricalWiring (see Appendix D). The technology, which is in current use, requires two electricalconductors. We have found that the Shortwatch sensor c a ~ le used as one of the

6

8/7/2019 DOE efforts 1719645

9/68

conductors inthe TDR experiments. The helically wound metallic outer sensor is anideal candidate forthisapplication.

Figure 3: The Biddle CFE 51OETime Domain Reflectometer (TDR)

Time Domain Reflectometers have been used for many years and remain the West, mostaccurate way to pinpoint cabling problems. Due to advances in current technology, theoperation and interpretation of a TDR have been greatly simplified. Because of its abilityto identi@cable problems, the TDR is now rapidly regaining popularity throughoutcommunications industries. If a cable is metal and has at least two conductors, it can betested by a TDR.Standing wave ratio (SWR) technologywas also investigated and demonstrated duringthe project (see Appendix Test). Work by others (Eclypse International) indicates S W Rmay b ve superiorsoftfaultdetection capabilityas compared to TDR echnologies.Future work isplanned to investigate overtemperature hult location byuse ofthistechnology.3. Sensor aging tests.Thermally accelerated aging tests were carried out by BPW and GeorgiaTech todemonstrate mechanical and electrical performance of Shortwatch sensorsover thedesign lifetime of wire and cable utilizing the technology (see Appendix C nd Figure 4).The sensor filament demonstrated mechanical integrity over the design life period of theaccelerated age test. Switch magnitude decreased significantly over the test period,

7

8/7/2019 DOE efforts 1719645

10/68

although this is believed due to loss of crystallinity of the basepolymer (a phenomenonnot expected during natural aging).BPW is currently researching how to utilize accelerated aging results of Shortwatchsensors n predicting remaining use l l ife. Although not originally anticipated in Phase2, demonstration of such a capability would significantly advance the goals set forthinthe Department ofEnergys SBIR solicitation in predicting the remaining life of cable.

Figure 4: Accelerated Age Testing of Shortwatch Sensor Filaments4. Production of Shortwatch capable wire.Figure 5 presents a cross-sectional drawing of the Shortwatch multi-conductor cabledesign selected for prototype production and testing. The multi-conductor jacketed cableutilized three insulated conductors and a Shortwatch conductivepolymerovertemperature sensor filament in the center. A Shortwatchmetallic sensor ribbon,wrapped helically around the insulated conductors, provides mechanical damage sensing.

8

8/7/2019 DOE efforts 1719645

11/68

Figure 5: Shortwatch Hybrid Cable

Figure 6 presents a photograph of a length of the prototype Shortwatch cablemanufactured atRockbestos Surprenant Cable Corporation of Clinton, Mass. The cableis a 600 volt 3 conductor, 12AWG tranded cable utilizing cross-linkedpolyethylene(XLPE) conductor insulation and Hypalon jacket. The cable was manufactured toRockbestos Firewall 3 specifications. FireWall3 is an environmentally qualified(EQ)cable commonly used in the nuclear industry. Environmental qualification of heShortwatch prototype cable was not scoped,nor performed on he prototype cable.

9

8/7/2019 DOE efforts 1719645

12/68

Figure 6: A Length of the Prototype Shortwatch Cable

Figure 7 showsa close-up photographof the prototype Shortwatch cable showingastripped end of the cable with the mechanical damage sensing ribbon and the Shortwatchovertemperature-sensing filament visible in the photograph. The mechanical damagesensing ribbon is nickel-plated copper, 0.002 thick, wrapped at approximately4turns/hch. The overtemperature-sensing filament is a compositeof the conductorinsulation (electronbeam cross-linkedPE) and a low-structure carbonblack filler andantioxidants. The filament diameterwas 0.020 inches.

10

8/7/2019 DOE efforts 1719645

13/68

Figure 7: A Close-up Photograph of the Shortwatch Cable

5 . In-service test of Shortwatch.Integrated testingwas performed on the cable shown in figures 5 '6 and 7. This testingincluded manufacturer's acceptance testing, cable overtemperature tests, cablemechanical damage testing, fault location testing by "DR and S W R echnologies, andmultiplexed fault detection demonstration. This testing is described in Appendix A.

Goals andFollow-OnWorkThisproject provided a multi-conductor cable for the nuclear industry, which, for the firsttime, provides real-time overtemperature sensing and Warning of cable mechanicaldamage befure an electrical tkilure occurs. Thisproject hlly meets the goals of the Phase2 SBIRaspreviously stated. BPW, Inc. has identified several areas for possible follow-on work, which we expectwill be of considerable interest in the Department of Energy'sbroader goal of i de n tw g new technologies relevant to safe and reliable operation of thenation's nuclear power plants:1. In-service test of Shortwatch cable.BPW proposes installation of Shortwatch cable in anoperating plant system test bed to:a. Demonstrateperformanceof Shortwatch cable in an operational environment;b. Reduce perceived "risk" o potential users of this echnology; and

11

8/7/2019 DOE efforts 1719645

14/68

C. Serve as a test-bed for future technology improvements in fault and life-predictivewire and cable.2. Design and manufELcfue Shortwatch wire and cable utilizing co-extruded sensors.Co-extrusion has the potential of reducing the cost and manufktwing complexity ofShortwatch-capable wire and cable by eliminating the step of separately extruding andcabling the sensor filament. A filament of the conductive matrixwould be co-extruded inthe wire or cable jacket during manufacture. This process also has the potential ofutilizing a single filament forboth overtemperature and mechanical damage sensing,eliminating the need for a metallic ribbon in the cable.3. Development of Shortwatch-capable cable protection elements forusewith existingwire and cable.Development of Shortwatch-capable sheaths, conduits, and wrapping tape componentswould allow installation over existing wire and cables to provide mechanical damage andovertemperature sensing capability to installed wire and cable systems. Development ofa demonstration protective device utilizing high-temperature materials used in a i r 4 sbeing done by BPW under an SBIR funded by the DOD.4. Overtemperature fault location.Fault location of Shortwatch mechanical damage sensorshas been demonstrated utilizingTDR and S W R echnologies. Location of overtemperature faults is more difficult due tothe soft fault nature and high material impedance of the sensor material. Future work,especially in S W R echnology may lead to the ability to locate the resistivity change inShortwatch sensorsdue to localized overtemperature conditions. Such a capabilitywould be especially valuable in locating internal or external heating sourcesbeforeinsulation failure, and to locate areas of cable, which may be susceptible to, acceleratedaging conditions.5. Prediction of useful life.BPW has determined that electrical characteristics of the Shortwatch sensor demonstrateelectrical properties, which appear to predict remaining useful life ofthe sensor. Sincepolymers used in wire insulation aretheprimary components of the Shortwatch sensor,thiswork may extrapolate to prediction of wire insulation degradation over time.Further work in this meawill be required to determine the feasibility of the method andpossible correlationwithw4.e insulation life.If found feasible, thismethod would be a significantadvance over current methods ofpredicting remaining life by destructive mechanical or chemical testing.

12

8/7/2019 DOE efforts 1719645

15/68

8/7/2019 DOE efforts 1719645

16/68

AppendixA Proof-of-ConceptTest

The final testing of theovertemperature and mechanical damage sensing capability wasperformed at the Rockbestos Surprenant Facility in Clinton, Massachusetts on May 16,2001. The purpose ofthese tests was o demonstrate the overtemperature and mechanicaldarnage sensing capability of our Shortwatch technology in realistic conditions and with aprototype nuclear-capable low-voltage power and instrumentation cable. Attendeesincluded DOE headquarters representatives including the headquarters TechnicalManager, BPW Principal Investigator and Senior Engineer fbr the project, RockbestosEngineering and R&D representatives, representatives of the School of Materials Scienceand Engineering fiom the &or& Institute of Technology, and representatives fiomEclypse International of Corona, California.The meetings included a project overview by BPW, Inc., and a presentation on theconductive composite work done by Georgia Tech. Tests and demonstrations includedovertemperature and mechanicaldamage &ult detection in the prototype Shortwatchcable, mechanical damage hult location in the prototype cable, and multiplexed faultdetection of multiple cables.The firmace and test cable for the overtemperature test are shown n figure A1. The oventemperaturewas 150C.

Figure Al: Furnace and Test Cable

14

8/7/2019 DOE efforts 1719645

17/68

8/7/2019 DOE efforts 1719645

18/68

FigureA3:Close-up of Wheel and CableThe end result of the mechanicaldamage test is shown in figure A4. The figure presentsthe test cable, which has the insulation abraded away until one of the primary conductorsis exposed.

Tests were a~ completed onthe Eclypse Model 501 Analyzer/multiplexer (FigureA5)to demonstrate that existing equipment could be used to support the Shortwatch-capable16

8/7/2019 DOE efforts 1719645

19/68

8/7/2019 DOE efforts 1719645

20/68

Appendix B: Georgia Tech Conducting Polymer Materials StudyDevelopment ofPTC Conductive Polymer Composite forShortwatch Sensor

Shijian Luo andC.P.WongSchool ofMaterialsScience&EngineeringGeorgiaInstitute of TechnologyAtlanta, GA 30332

IntroductionConductive polymer composites containing conductive fillers such as metalpowder, carboablack and other highly conductive particles in a non-conductive polymermatrix, have been widely used in the electrostatic dissipation (ESD), electromagnetic

interference shielding (EMS). There is a special group among the electric conductivepolymer composites. They are conductive polymer composites with large positivetemperature coefficient (PTC), which in some cases are called positive temperaturecoefficient resistance (PTCR). The resistivity of this kind of composite increases severalorders of magnitude in certain narrow temperature range. The transition temperature (Tt)was defined by intersection of tangent to point of inflection of resistivity vs. temperaturecurve with horizontal from resistivity at 25 "C (pz~). This kind of smart material canchange from conductive material to insulating material or vice versa upon heating orcooling, respectively. The smartness of this kind material not only lies on this large PTCamplitude (defined as the ratio of maximum resistivity at the peak or the resistivity rightafter the sharp increase to the resistivity at 25 "C), but also its reversibility, its ability inadjustment of the transition temperature, low temperature resistivity and high temperatureresistivity. The PTC conductive polymer composite has been successfblly developed inthis project for Shortwatch sensor. The work in this project was summarizedm differentphases.

1. Analysis ofComm ercial Conductive Polymer Composite.Some commercial products based on the conductive polymer that may show PTCbehavior were investigated, and these samples were obtainedand analyzed. Someof themare ContrimL.D. and Contrim V.F. fbm Crystal-X Corporation Actually, these productsare used for antistatic purpose. Their composition, thermal behavior, as well as theresistivity versus temperaturewere analyzed.

1.1. TGA analysisThe thermogravimetric analysis (TGA) experiments were performed under NZwithheating rate of 5 "C/min. The TGA profiles are shown in Figure B1. It can be seenthat in addition to polyethylene (PE), there is another component (polymer or otherorganic compound, which accounts for the weight loss in the range of 300-350C) in the

18

8/7/2019 DOE efforts 1719645

21/68

samples. The residue is supposed to be the conductive filler-carbon black. Thecompositions of the two samples are found as follows (they are reported in weightPa=ntage)-ContrimL.D.:LDPE: 53%, CarbonBlack 38%, Unknown component (possibly EVA): 8.8%

ContrimV.F.:HDPE: 66%, CarbonBlack 29%, Unknown component (possibly EVA): 4.6%

Io- ContrimLD20 U--- ContrimWI

Figure B1. TGA profile oftwo conductive polymer composites.

1.2. DSCAnalysisDifferential scanning calorimeter @SC) experiments were performed underNZ ithheating rate of 10 OC/min. The DSC analysis results are shown in Figures B2.Both of thetwo samples showed two melting points (85 "C, and 130 "C), which corresponded to thetwo polymer componeqts. It can be seen that the PE component in Contrim V.F. hasmuch higher crystalhity than hat in Contrim L.D. (The weights of the two samples arealmost same).

19

8/7/2019 DOE efforts 1719645

22/68

0.0 ContrimVF0 -- Contrim-LO5 .5 -

EB -1.0-5 .iiI

-1.5-

-2.0 t0 50 100 150 200 250 ? K)

Figure B2. DSC profile of two conductive polymer composites1.3. M A analysisThe therm0 mechanical analysis (TIM) experiments were performed under N2with heating rate of 1OClmin. The dimension change in the thickness direction wasrecorded. The dimension increased steadily with temperature. As the temperatureapproached the melting point, the dimension increased sharply due to the melting of thecrystalline polymer. However, for sample Contrim L.D. (Figure B3),the sharp increaseoccurred at the melting point of the unknown component. After that, the dimensiondecreased. It did not show increase in the dimension at the melting range of PE. This isbecause that the crystal ismainly contributed by the unknown component. As is shown inthe DSC profile, the melting peak at 85 OC is much bigger than that of 130 OC.

Figure B3. TMA profile of commercial conductive polymer composite.

20

8/7/2019 DOE efforts 1719645

23/68

8.0 -7.0 -6.0 -5.0 -4.0 -

For the sample of Contrim V.F. (Figure B3), the.dimension increased steadilybefore the 110 OC and increased sharply when it approached the melting point of HDPE,and it reached themaximum at 130OC It did not show any sharp increase in the range of80-90 OC as the crystalline portion due to the unknown component is much less than thatof HDPE. (It was shown on DSC profile). It should be noted that there are two shoulderpeaks (in lower temperature side) before the major peak. They should not be attributed tonoise. Actually, they are possibly due to the melting and recrystallization processoccurring before the major melting. As the heating rate was very slow in this experiment,the melting and recrystallization were expected to occur before the major meltingprocess.1.4. Resistivity vs. TemperatureThe samples were cut into small pieces and the resistance was measured withconductive adhesive as wo electrodes (The resistance of the conductive adhesive is muchlower than hat of the samples).

10.09.0

-4-Contrim V.F.+Contrim L.D.

0.0 I I I I0 50 100 150 200Temperature (C)

Figure B4. The resistivity versus temperature ofcommercial conductive polymercompositeFigure B4 shows the resistivity vs. temperature in the thickness direction ofsample Contrim L.D. The samplewas placed in an oven and the temperature of the oven

was raised5 OC at one step and dwelt for 5 minutes, and then the resistance was recorded.The resistivity increased steadily with the increased temperature and increased sharply asit approached to the melting point (85 "C) of the crystal of the unknown components(maybe plasticizer or additive). At 95 OC the resistance reading was beyond the range ofthe meter (100 mega-ohm). The resistivity vs. temperature cure is well matched by the"MA profile of the sample. It means that the PTC behavior of the sample is due tothermal expansion of the polymer matrix.

21

8/7/2019 DOE efforts 1719645

24/68

7.0 -6.0 -

c!Ez 5.0 -P 4.0 -d 3.0

2.0 -1.0 -0.0

0Y

u)u).-8A

I I I I I

Figure B4 also shows the resistivity versus temperature for sample Contrim V.F.in the thickness direction. There are five ranges in the curve. In the temperature rangebeIow 58 "C, the resistivity increased slowly with the temperature. In the temperaturerange of 58-98 "C, the resistivity increased sharply with the temperature. In thetemperature range of 100-120 OC the resistivity increased slowly with the temperature. Inthe temperature range of 120-134 OC the resistivity increased very sharply with thetemperature. Beyond 134 OC the resistivity decreased with the temperature. The sharpincrease of resistance in the temperature range of 58-98 "C was due to the melting of theunknown component. The very sharp increase in resistance in the range of 120-134 "Cwas due to the melting of the crystal of HDPE. The negative temperature coefficient(NTC) effect above 134 "C was due to the reorientation, reaggregation or reassembling ofthe conductiveunits.Figure B5 shows the resistivity vs. temperature for sample Contrim L.D. in thedirection parallel to the sheet. The sample was placed in the oven that was heated at thestep of 2 "C and dwelt for 2 minutes; the resistance of the sample was read every twominutes. The same trend was observed. However, the increase in the resistivitywas lessin the lower temperature range, and the resistively in the direction parallel to the sheetwas different fiom that in the thickness direction. The resistivity was beyond the range ofthe meter (100 mega-ohm) at 93C.I

22

8/7/2019 DOE efforts 1719645

25/68

L.D. and Contrim V.F. is well explained by the thermal expansion of the polymer matrix.The sharp increase in the resistivity of the PTC conductive polymer is due to the meltingthe polymer crystal.2. Characterizationof Commercial PTC Conductive Polymer

The PTC conductive polymer can also be used in the resettable fuses. It waslearned that two companies (Raychem, Bourns) have commercial resettable fuse productbased on the PTC conductive polymer. Some resettable h s ere obtained. Theresistance vs. temperature bebavior of Bourns MF-R800 resettable fbse was investigated.The PTC conductive polymer used in th is product was subjected to the thermal analysis,and possible composition was suggested. The intrinsic conductive polymer (dopedpolyaniline) was analyzed too.2.1. PTC behavior of CommercialPTC materialsThe resettable fuse (MF-R800) was supplied by Bourns. The resistance wasmeasured with four-wire method. The fuse was placed in an oven, of which thetemperature was raised at the pace of 5C and kept for 5 minutes. Around the PTCtransition temperature, the temperature was raised at 2C at each step and kept for 2 min.The resistance was measured at each temperature point. The resistance was also recordedon the cooling cycle. For the cooling cycle, the heating power ofthe oven was turned off,and the oven was allowed to cool ~ t ~ d l y .he resistance at each temperature point wasrecorded.Figure B6 shows the resistivity of the conductive polymer used for MF-800resettable fuse. Obviously, thismaterial shows a PTC transition around 126 "C t heatingmode. The resistivity above the transition temperature is five orders of magnitude higherthan the value below the transition temperature. And this PTC transition is clearlycorrelated with the melting ofthe polyethylene crystalaround 13OoC,which is shown inDSC profile in Figure B7.In Figure B6,the resistivity changes, with the temperature in cooling mode, wasalso shown. The PTC transition is about 8 "C lower than he PTC transition temperaturein the heating mode. This is due to the crystallization temperature is always lower thanthemelting temperature.The matrix polymer polyethylene iscrosslinked. From the profile of resistivity vs.temperature, it was found that there is no negative temperature coefficient (NTC)behavior after the PTC transition; this is evidence that the PE is crosslinked. Theresistivity continues to increaseafter the PTC transition. In the uncrosslinked system, theNTC behavior will occur after the PTC transition, as shown in the case of Contrim V.F.

23

8/7/2019 DOE efforts 1719645

26/68

Resistiityof MF-R800 resettable fusevs temperature1OE+07

1.OE+05 -1.OE+04 -

1 OE+01i --.OE+OO I 1 I I I I I 40 20 40 60 80 100 120 140 160

Temperature (C)Figure B6. Resistivity ofMF-R800 resettable fusevs. temperature.

2.2. Thermal analysis of commercial PTC conductive compositeThe conductive polymer used for the resettable fuse was taken out after thecoating of the resettable was peeled off. The conductive polymer was subjected tothermal analysis. All the thermal analyses were performed according the same proceduresused for the analyses for Contrim L.D. and Contrim V.F.The TGA $rofile of doped

conductive plyaniline was also recorded under N2 with heating rate of S"C/min.

0.0,

OS--E8 -1.0-

-1.5-i

FigureB7.DSC profile of PTC conductive polymer used for MF-R800 resettablehse.

24

8/7/2019 DOE efforts 1719645

27/68

In order to find out the composition of the conductive polyper used in theresettable h e , the TGA experiment was performed on the conductive polymer. TheTGA profile was shown in Figure B8. The sample began to lose weight rapidly around240C. After 255"C, the weight is slow. And fiom 550C, the weight loss wasaccelerated and nothing is left above 624OC.ThisTGA profile strongly suggested that theconductive f&er in this system is not metal or metal oxide; it is not carbon black either.It was proposed that the conductive filler is intrinsic conductive polymer and the weightloss at 240-255C isdue to the dopant in the conductive polymer.

120

100

8025 6040

20

0Temperature (%) UtMWsalW.OGTA hstnments

FigureB8. TGA profile of PTC conductive polymer used for MF-R800 resettable fuse.In order to prove that intrinsic conductivepolymer losesweight completely above600"C, the conductive polymer Versicon Conductive Polymer form Monsanto Lot #170/175/178 as subjected to TGA experiment. It was told that this conductive polymer

is sulphonic acid doped polyaniline. The material began to lose weight at 220"C, whichis due to the sulphonic acid dopant. The weight loss of major component beganat 328"C,which account for 84%. The dopant accounts for 12%, and water accounts or 4%. Theweight loss is complete at 600C with negligible residue.3. Compounding and Characterization of PTC Materials with Carbonblack as Conductive Filler3.1. ExperimentalThe HDPE (04452N)was supplied by the Dow Chemical Company. t has amelting point of 129-134OC, a density of 952 kg/m3 and a melting index of 4.0 &lo

25

8/7/2019 DOE efforts 1719645

28/68

min Five different carbon blacks were used in this study and their characteristics arelisted in Table B1. Conductex SC, Conductex 975 and N660 carbon blacks were suppliedby Columbia Chemicals. Vulcan XC-72 and BP2000 carbon blacks were supplied byCabot Corporation. The structure of carbon black is generally characterized by tintstrengtlq CTAB (cetyltrimethylammonia bromide) absorption, DBP (dibutyl phthalate)absorption, and iodine absorption. Carbon black with small particle size shows high tintstrength. High-structurecarbon blacks contain more void spaces. As a result, a largervolume of DBP is needed to fill the voids between the aggregates. The adsorption ofiodine and CTAB serves as the basis for determining the surface area of carbon black.

Table B1. CharacteristicsofcarbonBlacksCTAB Tintstrength DBP Iodine Iodine AverageAbsorption (%lTRB) absorption absorption absorption particle(m2/gram) (d100gram) (mdgram) (mdgm) diameter(nm)

ASTMMethod D3765 D3265 D2414 D1510 D1510ConductexSC 130 123 115 220 220 -Conductex975 140 87 169 250 250N660 35 57 90 36 36 70VulcanXC72 143 87 174 253 253 30BP 2000 635 163 330 1412 1412 12

HDPE (or any other polymer) and carbon black were mixed in the mixer unit of aHaake Rhecorder at 2OO0C, 60 rpm for 20 min, and the equilibrium torque was recorded.The composite was taken out of the mixer after the mixing was complete, and thencompression-molded at 200C into a sheet with a thickness of 0.5 - 1.0m. A smallsquare specimen with dimension of 2.0 cm in each side was cut fiom the sheet. The exactthickness of the small specimen was measured and recorded. An electrically conductiveadhesive (silver flakes filled epoxy) was applied onto the upper and lower surface of thespecimen, which was connected to the two probes of a multimeter. The electricalresistivity was measured in the thickness direction of the composite sheet. The sampleswere put in an oven, and heated at 5C every five minutes in the temperature range below110 "C, nd 3 "C every three minutes when the temperature was above 110 OC Theresistance of the specimen was read for each temperature point at end of each step. Theresistivity was then calculated. All the resistivities reported in this work are DCresistivities, and the carbon black concentrations are expressed in weight percentage.3.2. Effect of carbon black on the processingDuring compounding of the composites, the weighed amount of HDPE was firstmelted in the twin-screw mixer unit at 200 "C, nd then the weighed amount of carbonblack was added into the mixer. The mixture was mixed for 20 minutes at 200 "C withrotation speed of the screw of 60 rpm The orque r e q M o keep the screws rotating atcertain speed and temperature was recorded by the machine. The torque readingdecreased with the experiment time during fist 5 to 6 minutes of mixing and finallyreached an equilibrium value. Figure B9 shows the equilibrium torque readhg of the

26

8/7/2019 DOE efforts 1719645

29/68

mixer for different experiments. The equilibrium torque reflects the viscosity of thepolymer melt system. The higher torque readingmeans that the polymer melt system inthe mixer has higher viscosity. Without the addition of carbon, the pure HDPE melt wasmixed in the mixer, and the torque required to drive the screws at 60 rpm was about 6.0Nm. As the carbon black was added into the system, the torque needed to drive thescrews increased. This shows that the carbon ncreased the viscosity of the polymer meltsystem, which is true for all carbon blacks. However, the extent to which the carbonblack affected the viscosi ty of the melt diffaed fiom type o type. This difference is notso obvious at low loadings, but become more obvious as loading increases. Addition ofBP-2000 arbon black increased the torque most dramatically, while the addition of N66Ocarbon black increased the torque reading the least compared to the other carbon blacks.Thisdifference isdirectly related to thecharacteristics of carbon blacks.

0 20 40 80 80Carbonblack content (weight%)

Figure B9. Torquereading during compounding of the composite at 200 OC 60 rpm.

As is shown in Table B1, BP 2000 carbon black has the highest CTABadsorption, tint strength, DBP absorption and iodine absorption among all these carbonblacks. This means BP 2000 carbon black has the smallest particle size (its averageparticle diameter is 12 nm), highly aggregated structure and highest surface area. Theseproperties contributeto themost obvious torque increase.N660 carbon black shows lowest CTAB adsorption, tinting strength, DBPabsorption and iodine absorption among he five carbon blacks. It has the largest particlesize (70 nm), lowest structure and lowest surfi-ice area. The addition of N660 carbonblack increased the viscosity to the least extent, thusN660 carbon black was the easiestone to disperse.Vulcan XC72 and Conductex975 are fiom different manufactures, but they havesimilar properties as shown in Table B1.Thus hey showed similar behavior as showed inFigure B9. Conductex SC has properties between those of the Conductex 975 and N660,

27

8/7/2019 DOE efforts 1719645

30/68

so the torque readings of polymer melt loaded with Conductex SC were between thoseofpolymermelt loadedwithConductex 975 and N660 at same loadings.3.3. Effect of carbon black on thebase resistivity

Figure B10 shows the resistivity of the conductive composite at different loadingsof carbon blacks at 25 "C. It is not surprising that the resistivity decreases with theincrease of carbon black loading; and the trend is true for all carbon blacks. Polymercomposite with different carbon black at same loading showed different resistivity, andthis is also related to the structure of carbon blacks. BP-2000 carbon black has thesmallest particle size, highly aggregated structure and high Surface area, and theseproperties lead to low resistivity at same amount of loading of carbon blacks, as it iseasier for the BP-2000 carbon black to form conductive path in polymer matrix. WhileN660 carbon black has large particle size, he less aggregated structure and surhce area,the resistivity of the composite filled with N660 carbon black is higher than hose of thecomposites loaded with other carbon black at same loading. As discussed above,Conductex 975 and Culcan XC72 have similar structure, thus they showed similarbehavior of esistivity versus carbon black loading at 25 OC The resistivity of ConductexSC carbon black filled HDPE was between those of HDPE composite filled with N660black and Vulcan XC72 carbon at same oading.

1.E*

1OEm

l.E+03r.4?f3 l . O E 9 20:

l.OE+ol

1.E40

FigureB10.Resistivity of different carbonblack filled HDPE at 25 O C .

28

8/7/2019 DOE efforts 1719645

31/68

3.4. PTC behavior carbon black filled HDPEThe resistivities of those polymer composites fded with different carbon blackversus temperature are shown inFigure B11 to FigureB15.The resistivities of all the composites at different carbon black loading increasedwith temperature. The resistivity increased steadily with temperature fiom room

temperature, and increased more rapidly after 120C as the polymer crystal started tomelt. The esistivity reachedthepeak valuearound 133C. The resistivity decreased withM e r ncrease of temperature, which is called negative temperature coefficient effect.The NTC behavior was very obvious for composite with low carbon black loadii, whileit was not so obvious for system with high carbon black loading. Thisobservation is sameas reported in the literatures. The PTC anomaly is due to the thermal expansion ofpolymer matrix during the melting, and the NTC behavior is due to reorganization ofcarbon black in the mobile polymer melt phase. This reorganization is easy to occur forcomposite with lower carbon black loading due to its lower viscosity,while it isdifficultfor systemwith high carbon black loading and high viscosity. It has been reportedthatthe NTC behavior can be removed by crosslinking of the polymer network It wasobserved by Tang et al that the PTC intensity as well as the base resistance decreasedwith the thermal cycles. The same phenomenon was observed in this study. Thespecimen always showed lower resistance after being cooled down to mom temperaturecompared to the same specimen before the resistance versus temperature experiment wascarried out.Different carbon black filled composites did not show exactly the same PTCbehavior. For BP-2000 carbon black filled HDPE (Figure B1l), the PTC amplitude isclose to lo'.' for composite with filler loading of 10%. Further increase of the loading ledto the decrease of the PTC amplitude. With loading above 20%,the PTC amplitude wasless than 10, which means that the BP 2000 carbon black filled PE had quite goodthermal stability in its resistivity.

29

8/7/2019 DOE efforts 1719645

32/68

l.OE+03

it8 l.OE+oZ-0:

l.OE91

l.oE+oo 420 40 60 80 100 120 140Temperature (C )

FigureB 1. ResistivityofBP-2000carbonblack filled HDPEversus temperature(The weight percent ofcarbonblack is showed for composite).Compared to BP 2000 carbon black filled HDPE, Vulcan XC-72 carbon blackfdled HDPE had greater PTC behavior (Figure B12). At loading of 15% and 25%, thePTC amplitudes were about Id.With further increase of the loading, the PTC amplitudedecreased. The PTC amplitude of the composite with 50% carbon black loading is less

than 10.The Conductex 975 carbon black filled PE showed similar behavior as those ofCulcan XC-72 carbon black filledHDPE (Figure B13). The PTC behavior for conductexSC carbon black filled PE (Figure B14) is similar to that of Conductex 975 and CulcanXC-72 carbonblack filled HDPE.

30

8/7/2019 DOE efforts 1719645

33/68

Y

Y

Y

j

1oEwIOE+Oi

1OE*l.OE+05l.OE+O4

8 l.oE+o3lOEM2l .OE91

I .OEm

E

5E

20 40 66 80 100 120 140Temperature(C)

Figurq B 12. Resistivity of Vulcan XC-72 carbon black filled HDPE versus temperature(The weight percent of carbon black is showed for composite).1OE+07

l.oE+Oo I 120 40 60 80 100 120 140

Temperatwe (C)Figure B13. Resistivity of Conductex 975 carbonblack filled HDPE versus temperature(The weight percent ofcarbon black is showed for composite).

31

8/7/2019 DOE efforts 1719645

34/68

Figure B14. Resistivity of Conductex SCcarbon black filled HDPE versus temperature(The weight percent of carbon black is showed for composite).Among all the carbon blacks tested, N660 carbon black has the largest particlesize, lowest surface area and low congregated structure. Thus he composite filled withN660 carbon black shows the lowest melt viscosity, and highest electrical resistivitycompared to other composites at same loading, as mentioned above. However the PTCamplitude of the composite with N660 carbon black was the highest compared to those ofthe composites filled with other carbon blacks. The PTC amplitude of the compositeswith loading of 25%, 30%, 35% and 40% were all above lo5(The resistance reading ofsamples containing 25%, and 30% carbon black were beyond the range of the multimeter

of 120 MC2. So the actualPTC amplitude of those composites was even higher). Even atloading as high as 50%, the PTC amplitude was still as high as lo4.As the loadingincreased to 60%,the PTC amplitude dropped to

32

8/7/2019 DOE efforts 1719645

35/68

l.OEtl1 -

l.oEc(rz

Figure B 15. Resistivity ofN660 carbon black filled HDPE versus temperature m eweight percent of carbon black is showed for composite).From the comparison of PTC behavior of different carbon black filled HDPE, itcan be found that large particle size, smd1 surface area, and small extent of aggregatedstructure lead to the polymer composite with great PTC behavior. This is because theunaggregated individual carbon black particles can be separated more easily underthermal expansion of polymer matrix,which leads to increase of the distance betweenadjacent particles and decrease of the contact area between the adjacent particles. While,

the particles in the aggregated structure are difficult to be separated by thermal expansionof polymer matrix However, the great PTC behavior can not be simply attributed to thethermal expansion of the polymer matrix,which leads to the conductive filler volumeloading fhll below the percolation volume &actionas suggested in the literature [2]. As isshown in the PTC behavior of N660 carbon black filled HDPE, the resistivity of thecomposite with 50% carbon black loading reached IO6 o h c m , which is much higherthan the resistivity of the composite with 25%, 30%, 35%, and 40% weight loading atroom temperature. However, the volume fraction of conductive filler in the compositewith 50% weight carbon black loading at the peak resistivity temperature is still higherthan hat of he compositewith loading of 40% weight at mom tempedture. So the largePTC behavior of N660 carbon black filled polyethylene is due to some microscopicmechanism under the macroscopic phenomenon of large thermal expansion during themelting of the polymer crystal. Although carbon black particles are dispersedhomogeneously in the polymer melt during mixing, carbon particles are pushed out of thecrystalline region as impurity in recrystallization process during cooling. Thus, attemperature below melting point of polymer crystal, carbon particles are dispersed in theamorphous region only. As he crystalliie melts, those carbon black particles re-disperseinto the polymer melt, thus the inter-particles distance increases so significantly that

33

8/7/2019 DOE efforts 1719645

36/68

electron tunneling between conductive particles is hampered, and resistivity increasesAmong the composites filled with different carbon blacks, theN660 carbon blackfilled PE showed the greatestPTC behavior. Large particle size, small Surface area, and

small amount of the aggregated structure lead to great amplitudeof PTC behavior. Thegreat PTC behavior is due to some microscopic mechanism under the macroscopicthermalexpansion ofthepolymermatrix during the melting ofpolymer crystal.

greatly.

3.5. PTC behavior ofPolymer Composite Pilled with Two carbon blacks.In order to find out the synergetic effect between two totally different kinds ofcarbon black, the composite filled with bimodal carbon black was prepared. Figure B16and Figure B17 show the PTC behavior of polyethylene composite filled with bimodalcarbonblack ofN66O andBP2000.It canbe seen fiomFigureB16hat with the addition ofBP2000carbonblack intothe composite of N660 illed HDPE, the resistivity of the composite decreased greatly.The resistivity of compositewith40% N660 carbon black and 10% BP2000 carbonblackis 150 times lower than that of with 50% N660 carbon black alone. With furthersubstitution of N660 carbon black with BF2000 carbon black, the resistivity decreasedslightly. The PTC amplitude of the composite with bimodal carbon black is much less

than the composite filled with N660 carbon black alone at the same amount of loading.With 10% N660 carbon black replaced by BP2000 carbon black, the PTC amplitudedecreased fiom more than lo4 to less than 100 for composite filled with 50% carbonblack.1 OE+071 OE+061 OE+051 OE+041 OE+031 OE+021 OE+Oll.OE+OO

0 50 100 150 200Temperature (C)

FigureB16.Resistivity ofHDPE composite filled withbimodal carbonblacks versustemperature.

34

8/7/2019 DOE efforts 1719645

37/68

Figure B17 shows the PTC behavior of LDPE composite filled with 40% carbonblack. It is clear that with 40%N660 carbon black only, the composite showed high PTCamplitude. While with 5% N660 carbon black replaced by BP2000, the base resistivitydid not change much. However, the PTC amplitude decreased greatly. Originally, thePTC amplitude was more than lo4, with the substitution of 5% N660 carbon black byBP2000 carbon black, the PTC amplitude decreased to less than 100. With 10% N660carbon black being replaced by BP2000 carbon black, the base resistivity decreasedgreatly, and the PTC amplitude was even lower, less than 10. Figures B16 and B17suggest that the addition of BP2000 carbon black in place of N660 carbon black is notbeneficial to increase PTC amplitude, and at themean time it reduces the base resistivity.

r0W3 1 OE+041 OE+03I8U 1.OE+O2

1OE+011OE+OO

0 20 40 60 80 100120140160Temperature (C)

Figure B17. PTC behavior ofLDPE compositefilled with 40% carbon blacks.3.6. Effect of Polymer Matrix on PTC Behavior.Figure B1S shows the PTC behaviors of polymer composites containing 40%N660 carbon black in different polymer matrices. It can be seen that among differentpolymer matrices, polyethylene (PE) can provide the composite with 0 large PTCamplitude. Between HDPE and LDPE, the HDPE can provide polymer composite witheven higher PTC amplitude. EVA resin is usually introduced to polymer composite tomaintain the desired mechanical property at high filler loading. The introduction o fEVAto PTC composite led to an increase in room temperature resistivity and a decrease inPTC amplitude. Use of polypropylene instead of PE as polymer matrix led to polymercomposite with higher base resistivity and lower PTC amplitude. The PTC transition ofpolymer composite with PP as polymer matrix also took place at higher temperature, asthe meking point of PP (1 66C) is higher than hat ofpolyethylene.

35

8/7/2019 DOE efforts 1719645

38/68

I I I I0 50 100 150 200

Temperature (C)FigureB18. Resistivity ofcomposite containing 40% N660 carbonblack anddifferent resin versus temperature.

3.7. Effect of Carbon Black on the Crystdllinity of Polymer Matrix.The DSC analysis was used to investigate the effect of carbon black on thecrystallinity. Figure B19 shows the recrystallization peak of HDPE, 30% N660 carbon-black-filled HDPE and 50% N660 carbon-black-filled HDPE. It cafl be seen that therewas a slight difference in peak temperature. With the addition of carbon black, therecrystallization peak temperature increased by about 2.5 OC. The carbon did not affect

the melting temperature of polymer crystal (Figure B20). From the areas of therecrystallization and melting peaks, the relative crystallinitiesof the polymer resin weightpercentage canbe calculated, and they are listed inTableB2.The relative crystallinity ofHDPE was taken as 100, he relative crystallinity of HDPE filled with 30% N660 carbonblack is around 103, and that ofHDPE filled with 50% carbon black is around 97. Thereis only slight difference in those samplesin term of crystallinity. The added carbon blackdid not significantly affect he melting and recrystallization temperatures, nor did it affectthe crystallinity significantly.

36

8/7/2019 DOE efforts 1719645

39/68

3- 0-HDPEa--- HDPE+30%NBBO2-

-21 ' I ' I . I , I , L , I , I40 60 80 100 120 140 160 180TemperatureCC)I

Figure B19. Comparisonofrecrystallization between HDPE and HDPE filled with N660carbon black.

0.5

0.0hP3 5IA2I-1.0

-1.5

0 HDPEL. DPE+50%NW0 - -- HDPE+30%NBBC

V60 80 100 120 140 160 120

-0UI Temperature (%)40

Figure B20. Comparisonofmelting peaksofHDPE and HDPE filled withN660 carbonblack.TableB2.Relative crystallinity ofHDPE

Materid Reaystal imtion MeltingHDPE 100 100102.796.2 104.598.4

37

8/7/2019 DOE efforts 1719645

40/68

Carbon black with large particle size, small surface area, and small amount of theaggregate structure leads to the polymer composite with high base resistivity and greatamplitude PTC behavior. Carbon black with small particle size, large amount ofaggregate structure leads to polymer composite with low base resistivity and low PTCamplitude. The composite filled with bimodal carbon blacks with both large and smallcarbon particle does not show both low base resistivity and high PTC amplitude. Carbonblack can lead to a slight increase of the recrystallization peak temperature; however, itdoes not significantly affect the melting temperature nor the crystahi ty . Furthermore,PTC behavior also depends on the polymer matrix.4. Compounding and chsaracferization of Shortwatch tape4.1. Resistivity measurement in thickness direction for tape materials

It was desirable to use the base resin of cable insulator material for theShortwatch conductive composite. The materials were compounded in Rockbestos withtwo basic resin and conductive carbon black N660.Among all the samples, the followingphenomenon was observed. The rehistance reading was not as stable as commercialavailable resettable fusemade of PTC conductive polymer composite.For DFD-6040 esin (standard resin for cable insulation) and NA-317 esin (amuch less branched, with a much sharper melting transition) filled with N660 carbon(40%, 45%, 50%, 55%) uncrosslinked and crosslinked sample, their resistance wasmeasured together with their behavior versus temperature. The PTC amplitude (definedas the maximum resistivity after PTC transition to resistivity at room temperatme) wasfound to be above lo5.Lower carbon black loading gave higher base resistivity andhigher PTC amplitude.

8/7/2019 DOE efforts 1719645

41/68

1.OE+02 -l.O+Ol -.

+4S% carbon, uncrosslinked50% carbon, unaosslinked

+40% carbon, croculmked+45% carbon. croculmked .- 4 4 0 % carbon, -linked

I.OE+OO t *

Figure B21. Resistivity in thickness direction ofDFD-6040 esin filled with N660 carbonblack versus temperature.

1 OE+l 11 OE+l01 OE+091 OEW1OE+071 OE+061 OE+051OE+O41 OEM3

--(E-&% arbon. unoosslinked50% carbon,uncFosslinked*40% carbon.crosslinked

I - -1OEM21 OEM1I o m20 40 60 80 100 120 140 160

Temperature (C)Figure B22. Resistivity in hickness direction ofNA317 resin filled with N660 carbonblack in thickness direction

39

8/7/2019 DOE efforts 1719645

42/68

Crosslinking normally can reduce the NTC behavior after the PTC transition.This effect was confirmed in the system under investigation. However, crosslinking doesnot significantly affect the base resistivity of the conductive polymer composite. Thephenomenon that the resistance reading at individual temperature point (lower range) wasnot stable suggested that the carbon black may not be mixed very uniformly.There is one very important phenomenon to mention. After one temperature cycle(heating up to temperature above PTC transition temperature and then cooring down toroom temperature), the resistance of same piece of specimen increased (two samples ofNA317+50%N660-crossld (increased 50 times) and DFD+50%N660-crossld(increased 7 times) were tested in this experiment). This strongly suggested that the agingtest is necessary for the electricity behavior of material4.2. Electrical propertymeasnrement in length directionThe resistance was measured in length direction at Merent temperature of PTCmaterial compounded in Rockbestos. The results were shown and compared in thefollowing figures. There is no significant difference in the resistivity between the tworesins. However, the measured resistivity in length direction seemed to be two orders ofmagnitude lower than hat in thickness direction.

+40% Catbon, crosslinked

l.OE+OO ! I I I I0 20 40 60 80 100 120 140Temperature(C)

Figure B23. esistivityofNA-317 esinfilledwithN660carbonblack m length direction

40

8/7/2019 DOE efforts 1719645

43/68

1 OE+07

1OE+06

1OE+05nEI!E 1.OE+O4c0Y.->g 1.OE+03.-@

1OE+02

1 OE+01

8/7/2019 DOE efforts 1719645

44/68

Table B3. Mechanical property ofN660 carbon black filled polyethyleneAverage elongation (%) Average Strength(Mpa)resinDFD-604040%, uncrosslinked40%, crosslinked45%, uncrosslinked45%, crosslinked500x3, uncrosslinked50%, crosslinked

30.86111.823.370.016.633.0

14.816.7515.217.818.419

resmNA-31740%, uncrosslinked 16.4 14400, crosslinked 35.0 1545%, unaosslinked 13.1 15.7545%, crosslinked 20.3 16.7550%, ung.osslied 9.4 15.755004crosslinked 22.5 19

4.4. Effect of antioxidant on PTC behaviorSome PTC materials were compounded in Rockbestos with antioxidant (3%Irganox 1010) added into the formulation. And the resistivity of the PTC material was

measured both in the thickness direction and in length direction. The results are shownFigure B25 and Figure B26.The resistivity in length direction is less than he resistivityin the thickness direction.1 OE+lO1 OE+091 OE+081 OE+071 OE+061OE+051 OE+O41 OE+031 OE+02

--t A317+45%C

1OE+Ol1 OE+OO 0 20 40 60 80 100 120 140 160

Temperature (C)FigureB25.Resistivity in thickness direction vs. temperature ofN660 carbon black filledcomposite (3% Irganox 1010 added).

42

8/7/2019 DOE efforts 1719645

45/68

1 OE+071 OE+06

nE l.OE+05Y5 1.OE+040W$ '-OE+03l.OE+02

1 OE+0121OE+OO I I I I I I I , I

0 20 40 60 80 100 120 140 160Temperature (C)

Figure B26. Resistivity in length direction ofcomposite filled N660 carbon black versustemperature (3% Irganox 1010 added).The antioxidant added did not affect the resistivity of composites and their PTCbehavior. However, the NA317+50% carbon black system became brittle after theaddition of the antioxidant. It should also be noticed that there is not much difference inthe base resistivity betweenNA317+50%C and NA317+45%C.It should also be noticed that with same amount of carbon black loading,composite with DFD6040 as matrix showed significant higher resistivity than thecompositewithNA317 as matrix. Also the PTC amplitude of composite with DFD6040

as matrix is significantly higherthan hat of compositewithNA317as matrix.

5. Comparison of PTC behavior of Shortwatch tape and filamentTwo diffixent geometries of materials were manufactured fiom ShortwatchPTCmaterials: filament and tape. Shortwatch filament showed lower switching magnitudethan the Shortwatch tape prepared earlier. It was suspected that this is maybe due tolower crystallinity of polymer phase in filament due to the W ooling.Two approachesmight solve this problem. One is cooling the filament slowly, and the other one isannealing the filament at elevated temperature. Both of these two approaches should beable to increase the crystallinity.Thermogravemetric analysis (TGA) experiment was performed to check if thecarbon black (CB) oading was at 50%. The TGA curve ofthe filament (Figure B27)showed the weight loss of the filament versus temperature. Polymeric material startsdecomposing at high temperature under nitrogen, while carbon black does not decompose

at high temperature. The residue at temperature above 500 O C is carbon black. From thecurve, we can say that thecarbonblack loading is about 50%. The carbon black loading

43

8/7/2019 DOE efforts 1719645

46/68

z 60-P540-

ao-

for tape sample ofDFD6040+50%CBwas also checked, and the result is shown in FigureB28. The carbon black loading is also 50%. So there is no significant difference incarbon black loading between the two samples. The difference in switching magnitudemay be caused by polymer resin or processing.

'I-D

Figure B27. TGA profile of Shortwatch filament (heating rate: S"C/min, under N2)

100

0

Figure B28. TGA profile of Shortwatch tape @FD6040+50%CB) (heating rate:S"C/min, underNz)

44

8/7/2019 DOE efforts 1719645

47/68

100000

Measurement of resistivity versus temperature was performed h Georgia Techtoo. The results are shown in Figure B29. The lower switching magnitude is alsoobserved. The filament was annealed at llOC for 10 hours and slowly cooled down inan oven. The resistivity versus temperature was m&Ured again after annealing,and isalso shownin Figure B29. No improvement in switching magnitude was observed. Thisindicates that the low switching magnitude cannot be solved by mealimg the filamentA narrow strip (0.54 mm in thickness, 1.20 mm inwidth, and 40.81 mm length)was cut h m the tape sample prepared earlier @FD6040+50% carbon black,crosslinked). The resistivity of this strip versus temperature was also measured, and theresults are shown in Figure B29 too. The strip showed a little bit higher cold resistancethan the filament, and the switching magnitude was much higher than the filament. Itshould be noticed that the resistance of this strip at 115 OC was beyond the range of themultimeter of 120 MJn, but 120 M!2 was taken as the reading at 115 O C . The actualswitching magnitude is higher than t is showed here. This shows that the geometry isnot a &tor affecting the switching magnitude, as the strip is close to the geometry of thefilament. The difference is due to the processing, which may introduce Merentdistribution of carbon black in the composite.

+filament, before annealing+filament, after annealing-- +strip cut from tape P

n 10000{000

100

10

10 20 40 60 80 100 120 140 160Temperature (C)

Figure B29. Resistivity of Shortwatch filamenttemperature. (The resistance readingfor strip at 115&C vas actuallyout of the range ofmultimeter 120 MQ, but 120MQ was taken as it% edmg)

w strip cut fkomtape versus

45

8/7/2019 DOE efforts 1719645

48/68

Differential scanning calorimeter (DSC) was also used to analyze the melting, andrecrystallization of filament as well as the tape by heating the sample to 150 "C thencooling it to 25C. Figure 30 showsthe DSC profile during heating cycle. The meltingpeak of filament was around 106.5 "C. After the filament was annealed at 110 "C for 10hours and then oven cooled, the melting peak temperature (105.7 "C) is not changedsignificantly. However, above the major meltingpeak, there is a small melting peak withpeak temperature at 113C. This showed that there was some change during the heattreatment. Some crystal became more perfection, thus showed another melting peak at ahigher temperature.

The melting peak of the tape sample was at a higher temperature 109.4 "C. Afterthe tape sample was heated to 115 "C for resistance measurement, it was kept at 100 "Cfor 1 hour, and 90 OC for another one hour, and then oven cooled. The DSC profileshowed no significant difference fiom that of the sample without heat treatment.

I O

Figure B30. DSC profile of filament and tape (heating rate: 5 "C/rnin)The DSC cooling profiles for those four samples are shown in Figure B31.Thecrystallization started at higher temperature for the tape specimen than for filamentspecimen. This indicates that there may be some subtle difference in the microstructure ofthe tape and filament specimen. There is no sign&ant difference in area ofrecrystallization peak, which are compared inTable B4. However, significant differenceexists in peak anxi among different samples. Peak area of tape sample is obviouslygreater than peak area of filament sample, indicating the tape specimen has higher

crystallinity than he filament. Annealing of the filament did not significantly increasethe crystallinity.

46

8/7/2019 DOE efforts 1719645

49/68

0

Figure B31. DSC profile of filament and ape (cooling rate: 5 "C/min)

Table B4. Peak area for melting peaksand recrystallization peaks (J/g)filament, not annealed 42.8 38.4filament,annealed 43.4 38.4tape, not annealed 47.6 38.5taDe. annealed 47.2 39.2

melting peak recrystallization peak

It was mentioned that the resistance never returned to the original value after thematerial is cooled down &om high temperature. This is related to the structure of thepolymer composite. At high temperature, the structure of carbon black is broken, and thebroken stnrcture cannot be repaired completely after the material is cooled down. Thecontinuous slow decrease of resistance at room temperature is due to the molecularmotion of polymer chain of amorphous region (This kind of motion takes place attemperature above the glass transition temperature of polymeric material). Highcrosslinking may help restrain the motion of polymer chain of morphuus region,However, high crosslinking will also increase brittleness of material. Several sampleswith different crosslinking density (with different E-beam dose) may be prepared andtested if fbther optimization of the processing condition is desired.It was observed that the filament shrank during PTC test. Shrinking of thefilament is normal. During extrusion of filament, orientation of polymer chain isintroduced. During heat treatment, these oriented chains disorientate, and thus shrink.This shrinking problem can be solved by annealing the filament at elevated temperature.Different heating rate will give different behavior of resistivity versustemperature. The PTC behavior is related to themelting of polymer crystal, and meltingof polymer is heating rate dependant. Usually, high heating rate leads to high meltingpoint, thus higher PTC ransition temperature. There is no significant difference in the

47

8/7/2019 DOE efforts 1719645

50/68

100,000nE 10,000 -E 1,000-r0 100 -

10 -1 -0

W0$>v).-.cI.-a:

transition temperature for measurement taken at heating rate of 1.0 "Clmin and 0.1"C/min. Thisprobably indicates that there is no significant influence in heating rate in therange tested (O.l"C/mino 10"C/min). The difference in transition temperature obtainedat heating rate of 10 "C/min is possible due to experimental error or non-uniformityamongdifferent samples.

It was also observed that there is variation in PTC behavior of the filamentmanufactured at different batches shown inFigureB32.

+ 0% C /00 Filament

Figure B32. Variation of PTC behavior of Shortwatch filament.

6. Aging Test on Shortw atch M aterialsPTC filament was subject to aging test at 160C. The mechanical property and theelectrical property were measured after different aging time. It is easy to understand thatdurometer increased steadily to near hard state. This is due to the crosslinking ofpolymeric matrix during aging, which also leads to the increase in density. The slowdecrease in the cool resistivity of the sample versus storage time at room temperatureafter aging at high temperature is due to the slow crystallization process at momtemperature. Crystallization can take place the temperature range betweenglass transition

temperatureand melting temperature.The increase in resistivity in the initialaging test is probably due to the wetting ofpolymer resin on the surkce to carbon black andor the even m e r eparation of carbonblack particles. The drop of the resistance in later aging stage is possibly due to theincrease of volume fiaction of carbon black. The crosslinking of polymer matrix willreduce the crystallinity; possibly increase the density of polymer matrix, as thecrosslinkii increase the density in the amorphous region. PTC conductive polymercompositewith higher crystallinity polymer matrix will give lower base resistance andhigher switching magnitude. Higher density of polymer matrix (after aging) will lead to

48

8/7/2019 DOE efforts 1719645

51/68

decrease of the base resistivity as the volume kction of carbon black increases. Also,long time, high temperature aging in anoxidative environment may lead to the oxidationof polymer matrix and possible oxidation of carbon black surf8ce. All these possiblefactors will af fec t the resistivity and switching pgnitude. It may not be a simpleprocess. In order to better understand what happens in this aging process, morefundamental study with TEM, FM,XPS, nd DSC can be done. Also, t should be keptin mind that there is no direct correlation between electrical property of sample at longterm with the accelerated aging at high temperature and mechanical property. Actually,the total loss of PTC behavior of composite after 900 hours aging at 160'C is veryinteresting behavior (Figure B33).

1 .OE+081 .OE+071 .OE+061 .OE+051 .OE+041 .OE+031 .OE+021 .OE+011 .OE+OO

Sensor R-T ResponseI - tUnaaed Fi lament , 50% C I

0 20 40 60 80 100 120 140 160Temperature,C

Figure B33. PTC behaviorofShortwatch material after aging.Electrical property after long time high temperature has not been reported for PTCconductive polymer composite in the literature. People have just done aging for a fewthermal cycles rather than ong time high ternperahre aging. For the material to be used,how long time the product will be exposed to thehigh temperature should be understoodfor the high temperature aging. The material during aging is at totally different state

(above the melting point of polymer crystal, with no crystalline phase) from the workingstate (below the melting point of polymer crystal, with crystalline phase in thecomposite). We suspect that the crosslinking takes place in the mlt phase of origi.milcrystal. After 900 hours aging at 160C, there is no crystal phase left even after thematerials cool down, thus there is no PTC amplitude (switching magnitude) observedafter 900hours aging. DSC can be used to find out the change in crystallinity after agingat high temperature.

49

http://origi.mil/http://origi.mil/

8/7/2019 DOE efforts 1719645

52/68

Differential scanning calorimetery (DSC) analysis was performed on the agedShortwatch filament samples. The Figure B34 shows the DSC profile of the sampleduring DSC heating scanning. In the heating scan, the melting of crystal canbe detected.From here we can find out the melting peak temperature and relative crystallinity(percent of crystal phase in polymer). Before aging, the melting peak temperature is106"C, and the melting enthalpy is 42.7 J/g (the peak area). After the filament was agedat 160C for 286 hours, the melting peak moved a low temperature of 96.7OC, and themelting enthalpy is 32.2 J/g, indicating that the crystallinity decreased to 80% of thecrystallinity of filament before aging. While after the filament was aged at 160C for 883hours, no obvious melting peak was detected (No crystal left!). This analysis resultssuggested that aging at 16OOC has significant effect on crystallinity and meltingtemperature: it leads to lower crystallinity and lower melting temperature. The reason forlower switching magnitude after aging and eventual disappearing of switching magnitudeis exactly what we proposed: decreashg and eventual disappearing of crystallinity of thepolymer after long time high temperature aging. The DSC results confirmedour previousanalysis. In addition, the lower melting temperature will lead to lower switchingtemperature, which can be observed on one of charts ''Sensor R-T Response" (FigureB33).

0- aged for 883 hrs at 16OC0.4

0.2-zy"B 0.0-E

-0.2

-0.4-

74.49%t

t

1ffi.43C

Figure B34. DSC heating profile of Shortwatch filament. (heating rate: S"C/min)

The crystallinity of commercial PTC materials used for resettable h e howed asimilar loss with the thermal aging. it was also noticed that the switching temperatwe ofaged Raychem materials moved to lower temperature accompanied by a lower switchingmagnitude. This is what happened in our PTC materials, which is due to lower meltingtemperame (smaller crystal size) after high temperature aging.

50

8/7/2019 DOE efforts 1719645

53/68

ConclusionPTC conductive polymer composite for Shortwatch sensor has been successfbllydeveloped with carbon black as conductive filler. Carbon blacks with large particle size,small surface area and small amount of aggregate structure lead to composite with high

base resistivity and great PTC behavior. The great PTC behavior is due to microscopicmechanism under the macroscopic thermal expansion of the polymer matrix during themelting of polymer crystal. It was found that there is anisotropy of PTC behavior inShortwatch sensor materials, which is mainly due to the processing. Aging of PTCmaterial for Shortwatch sensor at high temperature (16OOC) can eliminate thecrystallinity, thus the PTC behavior.

51

8/7/2019 DOE efforts 1719645

54/68

Appendix C: BPWAging Trials of Shortwatch Sensors1. Introduction

IEE 323 and IEE 383 establish environmental qualification(EQ) requirements for wireand cable. Although EQtesting was not mthe scope of work of thisSBIR project,thermal accelerated agingtrialswere conducted to (1) demonstrate mechanical integrityof the Shortwatch conductive composite sensor over a typical accelerated age test periodused for thermal aging ofwire and cable in the auclear industry and (2) demonstrateacceptable electrical performanceof the conductive composite sensor after thermal aging.

2. TestApproach24 test samples were prepared fiom Shortwatch sensor filament received fromRockbestos Surprenant Cable Corp. on May 30,2000. Sensor filament specifications:Nominal extruded diameter: .020Base polymer:Conductive filler:Measured Resistivity:

LDPE/HDPE blend (Rockbestos TireWall3)N660 carbonblack (50% by weight)0.94 n-cm @ 23CAnti-Oxidant: 3% Irganox 1010

The filament was electron-beam cross-linked after extrusion.

Figure C :Sample extrusion h e Ftockbestos)

52

8/7/2019 DOE efforts 1719645

55/68

Figure C2: Sample filament coil

Sample filaments (Figure C2) were cut to 8.0 cm and inserted into 2.0 mm diameterglass tubesopen at both ends for support of the 16ilaments during aging.

Figure C3: Sample filament in glass support tubeSample glass support tubesFigure C3) were inserted into insulative glass fiber sleeves(Figure C4) to reduce quenching effects ofair-coolingand to improve coolinguniformity

53

8/7/2019 DOE efforts 1719645

56/68

when removed fiom the aging oven. The following photograph(Figure C4)shows theassembled sample and container.

Figure C4: Assembled filament sample and containerAll filament samples were aged at 160C in an aging oven (Fisher Scientific Isotemp@800 series, forced air) for times ranging fiom 1hour to 883 hours. &-age measurementsincluded sample length and diameter (for volume measurements), DC resistance in thelength (extruded) direction, and initial hardness (durometer). Ovenage time, agedresistance, length, diameter, and hdness were recorded for each sample.

Figure C5: Prepared samples

54

8/7/2019 DOE efforts 1719645

57/68

Figure C6: Prepared samples in aging oven

Figure C7: Durometer for measuring sample hardness

Electrical resistance Figure C8)was measured at one hour and 24hours &er removalfiom the oven to compensate for resistance changeswith time due to re-crystallizationeffects.

55

8/7/2019 DOE efforts 1719645

58/68

Figure C8: DC resistance measurements

Figure C9 shows the resistivity and sample durometer vs. oven age time for sample timesfiom 1 hour (arbitrarily assigned as unaged) to 880 hours. The 296-hour aging timecorresponds to an extrapolated 40% reduction of elongation-at-break@AB) for the basepolymer after40 years of aging at 9OC. The tests were extended to 880 hours todetermine the eventual failure mechanisms of the sensor filaments.The selection of 160C for accelerated aging was partly necessitated by time available inthe project. Lower temperatures (ideally below the transition temperature of 121C)would have been preferable, but would have required excessive aging times.As canbe seen fiomthe graph, the durometer (hardness) increases fioman nitial valueof80.5 to 83at the defined end of life (296 hours). Hardness at 880 hours was 88,corresponding to sufficient brittleness to cause breakage fiom normal handling.

56

8/7/2019 DOE efforts 1719645

59/68

The in

EYE0ci5B8dPI

2520151050

Filament Sensor Resistivity andDurometer Vs. Aging

908886 8

nE84 s8280

-Sensor msistiity, 1Hr. afler ternowl+ wometer

0 500 1000Oven Hours, 160C

ial increase in resistivity of the sensor fiaments s not completely understooc butmay be due to several factors including (1) rewettjng of the carbon particles duringlong time periods at melt temperatures, (2) re-distribution of the carbon particles undermelt conditions and (3) loss of crystallinity due to cross-linking of the polymer above thetransition temperature. (see Appendix B fiom Georgia Tech). This cross-linking locksthe polymer and prevents re-crystalhation upon cooling below the transitiontemperature. The eventual reduction of resistivity at the extreme end of life is believeddue to thehigh level of cross-linking resulting in reduction ofvolume h t i o n of he basepolymer ascompared to the inert (carbonblack) conductive filler. We believe that theinitial increased level of resistivity isprimarily a function of the elevated agingtemperature (above themelt temperature) andwould be at a significantly reduced levelunder naturalaging conditions. Inthis sense, the testing was conservative in nature.Figure C10 showssensor resistance vs. temperature (switching) for age times of 1hour(unaged), 96 hours, 296 hours( n o d nd of life) and 880 hours (hilure) at the agetemperature of 160C. The switching magnitude, defined as the ratio of sensor resistanceat 130C to 9OC (design temperature) is reduced fiom slightly over 2 decadesat thewed conditionto aboutone-half decade at nominal end of life. Switching magnitudeis reduced to nearly zero at hilure (880 hours). The reduction of switching magnitude isbelieved to be related to loss of crystallinity of the sensor material duringaging above thetransition temperature. The loss ofcrystallinity was confirmed byDCS measurements

57

8/7/2019 DOE efforts 1719645

60/68

carried out at Georgia Tech (see appendixB). Loss of switching magnitude would resultin the reduced resolution of overtemperature sensing by the filament.

Sensor R-T Response100,000,000

Q, 1o,ooo,o0O1 1,o0o,ooo(DtnIcli 100,000p1 10,Ooo

1,000100

0 50 100 150Temperature,C

-6- ulaged Fhment, 50%+Aged Filamnt 96

C

HoursAged Filament286Hours* ged Filament883HoUrS+ nagedTape, 50% C

Figure C9: Resistivity vs. temperature (switching magnitude)Naturally aged samples measured during the project did not demonstrateany initialincrease n resistivity (and subsequent loss of switching magnitude) of the accelerated agesamples. Figure C11 shows resistance vs. time for naturally aged samples over a ninemonthperiod of the project. These results further suggest the conservative nature of theaccelerated aged sample results.

58

8/7/2019 DOE efforts 1719645

61/68

21.81.6

E" 1.4m&g 1.2P 10.8

0.60.40.20

.-U

Natural Aging o f Insulation Com posite

-4-SSanplel- 1, nealed13/15/00 6/23/00 l O A f l 0 1/9/01

Date

FigureC10: Resistance vs. natural age3. ConclusionsAge testing carried out by BPW, Inc. demonstrates mechanical integrity of the conductivepolymer sensor filament for the design life of the cable. The eventual failure mechanismisbrittleness of the sensor. Accelerated aging also resulted in a significant loss ofswitching magnitude. This loss of switching magnitude is attributed to loss ofcrystallinityof thebase polymer (electronbeam cross-linked polyethylene) as a result ofaging above the crystalline transition temperature. Overtemperaturesensing capability atreduced resolution is maintained over life of the sensor.4. FutureworkThe extreme sensitivity afforded by changes in resistivity of conductive compositesas aresult of small changes in volume fraction of the base polymer appears to provide apowerfir1tool for investigating aging effects of the polymer. Although unrelated to thedevelopment ofa distributed overtemperature and mechanical damage sensing wire andcable completed under this SBIR, ging studies of compositesas their use as a conditionmonitoring approachwill bepursued in subsequent work and research proposals byBPW, nc.

59

8/7/2019 DOE efforts 1719645

62/68

Appendix D: Time Domain Reflectometer and Noise IssuesTim?DomainReflectometerO R )One of the goals of this project has been fault location, that is, the determination of theexact location of an electrical fault after the existence of a fault had been discovered. Weconducted many early experiments with a Tektronix 1503 time domain reflectometer(TDR) (see FigureD1)to ascertain the TDR'susefblness in fault location This TDRserved us well in the early proof-of-concept testing.

FigureD1:The Tektronix 1503 Time Domain Reflectometer

Any TDR requires two, more or lesspardlel, conductors to operate. If a hult occurs inone of the two conductors,theTDR can determine, quite accurately, the distance fromthe TDR to the hult. Basically the TDR sends a short duration electrical pulse down theconductor and measures the time required for the electrical reflection from the detectedelectrical hult to arrive back at the instrument to determine the hult location. Thedistance to the hult is calculated fromthemeasured time delay and the velocity ofpropagation of the electrical impulse in the considered wire. The velocity of propagation(VOP) of the electrical impulse is a specific characteristic of the electricalwjring underconsideration and must be either measured or predetermined. The easiest method todeterminetheVOP stomeasure the actual distance to a knownelectrical fault in theelectricalwiring under consideration and to calibratetheTDR reading to the appropriatedistance. TheVOP can hen be read directly from the TDR. The 1503TDR s anolderunitdesigned for longer ranges and of limitedportability(butwhich canbe purchasedused at very reasonable cost), so after we determined that a TDR could be ofuse westarted a search for a more practicalunit. After considering several different TDRs weselected the Biddle CFL 51OE (figureD2).This unit is very portable and alsoreasonably priced (about $1500).

60

8/7/2019 DOE efforts 1719645

63/68

Figure D2: 3ecta Imel terThe Biddle CFL51OE has:

Six ranges for testing power, cellular,CATV, nd telephone cablese Large high resolution backlit LCD

25, SOy75, and 101)ohmoutput impedanceGain, balance,and contrast controls fiont-panel accessibleAuto rangesaround he cursor for best view of fault

Tbisinstnune t ism advanced instnUnent capable of identming awide range of cablefaults usingTimeDomain Reflectometry. It offers exceptional featuresanda rangecapability n o d y ssociatedwith larger, more expensive instnune ts. Themeasurement range spansfrom30 feet to 9,000 feet, with an advertisedminimumresolutionof four inches.The Model CFLSlOEcanbe used on any cable consisting of t least two insulatedmetallic elements, one ofwhich m ay be the sheath or shield of the cable. The CFL51OE

61

8/7/2019 DOE efforts 1719645

64/68

has internal matching networks to allow testing of 25-, 50-, 75, nd 100-W cables.(These correspond to power, cellular, CATV, and telephone cable).In figure D2, the maximumrange under consideration is 90 feet (lower left) and themeasured range is22.1 feet (lower center). The device is set for 1OO-ohm cable with avelocity of 0.6 times the speed of electrical propagation ina vacuum (the speed of light).The first group of reflections (reading the signalfiom left to right) is from theconnections between the DR and the cable under consideration. The second largereflection is fiom a hult 22.1 feet from the TDR. The TDR is calibrated for arbitrarycable, as described above, by using a similar cablewitha know fiiult at a specificdistance. The controlsare adjusted to set the correct (known) distance and these controlsettings are used for determining faults in similar cable with fzlults at unknowndistances.It should be noted that for a helical conductor (like the Shortwatch metallic sensor)thedistance measured by the TDR to a fault in the Shortwatch sensor is the distance alongthe helixThe experimentswith both TDRs demonstrated that a fault in the metallic outer sensorof Shortwatch can be located very accurately (within the accuracy of theTDR), fthegeometry ofthemetallic sensor is known and if one of the other conductors in theelectrical wire or cable under consideration isused to complete the requirement for twoconductors stated earlier. The conducting polymer sensor cannot be used as he secondconductor because of the enomus difference in impedance between the metallic sensorand the conducting polymer.Noise TestingDuring thedevelopment of Shortwatch therewas a concern that the metallic Shortwatchsensor would act as anantenna and pick up external electronic noise fiom 60 Hertz ACequipment and the higher frequency electronic hum fiom florescent lights and electricmotors. We simulated the metallic Shortwatch sensor and locatedthis simulation closeto 60-hertzelectrical wires and to florescent light and found that a very simple low-bypass filter couldeliminate any noise concerns. Initially we used a R-C low bypassfdter but later found that an even simpler 0.1 -microW capacitor to ground was all thatwas required. The equation governing the low bypass filter is:

1Vout = yin Xi( J F G FWhere:Vout =output voltage asa h ct ion of fkequency (in volts)Vin = inputvoltage (AC noise or hum) as a hction of frequency (in volts)Xc = 142xPI x Frequency in Hertzx C)C= capacitance of the grounding capacitor (in Farads)R =the inline resistance (inOhms).This equationcanbe used o determine the ratio of R/C to get particular voltageattenuation at a specific frequency. We chose R andC so the Vout was five percent of

62

8/7/2019 DOE efforts 1719645

65/68

~~

Vin at 60Hertz. AMicrosoft Excel spreadsheet model o f the above equation was used todetermine the appropriate resistors and capacitors. Later tests indicated that the resistanceR could be eliminated for our purposes and we simplified the circuit to just a 0.1microhad grounded capacitor.

63

8/7/2019 DOE efforts 1719645

66/68

Appendix E: Certified Test Report on Shortwatch Cablel o f 2

n ms

64

8/7/2019 DOE efforts 1719645

67/68

2of2 . -A

.- .-*...--

I

65

8/7/2019 DOE efforts 1719645

68/68

AppendixF: References1. Linda Smith, Kimberly Long, 1993 Residential Fire Loss Estimates, U.S. ConsumerProduct Sdkty Commission, Wwhington, DC 20207.2. Kimberly Long, EHHA, Memorandum entitled 1992 National Estimates ofElectrocutions Associated With Consumer Products, dated July 27,1995, U.S.Consumer Product Safety Commission, Washington, DC 20207.3. CharlesV. Schwab ndiana State University ExtensionPub#Fact Sheet Pm1265a,Jm pr y 1992.4. Milton S.Greenhalgh,U.S. Patent No. 4,707,686, Over temperature Sensing Systemfor Power Cables, issued November 17,1987.5. EdwinR Milk,U.S. Patent No. 4,607,154, Electrical Heating Apparatus ProtectedAgainstm Overheating Condition and A Temperature Sensitive Electrical Sensor for UseTherewith, issued August 19,1986.6. Karl F. Schoch, Howard E. Saunders, Conductive Polymers, IEEE Spectrum, June,19927. Kenneth S. Watkins, Jr., Ralph E. Pope, PCT application W09504757 entitledElectrical Safety Device, published 14 September, 1995, World Intellectual PropertyOrganization.8. DonaldG. Fink, H. Wayne Beaty, Editors, Standard Handbook For ElectricalEngineers, Twelfth Edition, McGraw-Hill, 1