Armed Services Technical Information Agency

1 Armed Services Technical Information AgencyARLINGTON HALL STATIONARLINGTON 12 VIRGINIA"".CONTROL ONLFNOTICE: WHEN GOVERNMEZNT OR OTrHER DRAWINGS, SPECIFIATION OR OTHER DATAARE UD FOR ANY PURPOSE OTHER T!AN IN CONNECTION WITH A DEFIYTELY RELATEDGOVERNMEZNT PROCUREMEZNT OPERATION, THE GOVERNMENT THEREBY INCURSNO RESPONSIBILITY, NOR ANY OBLIGATION WH AND THE FACT THAT T!IEGOVERNMEZNT MAY HAVE FORMULATED, FURIHD OR IN AMY WAY SUPPLIEDTHNAM DRAWINGS, SPECIFICATIOINS, OR 0rHER DATA IS NOT TO BE REGARDED D~YIMPLICATION OR OHRWISE AS IN ANY MANNER UCZNZNG THE HOLDER OR ANYPESNOR CORPORATIUN, ORt COINVEYIG ANY aRrsM OR g fRONm TO MANUFUSX OR SELL ANY PATENTED DMWION TZAT MAY 04 ANY WAY BE RELATED T'-DISCLAIMER NOTICETHIS DOCUMENT IS BEST QUALITYPRACTICABLE. THE COPY FURNISHEDTO DTIC CONTAINED A SIGNIFICANTNUMBER OF PAGES WHICH DO NOTREPRODUCE REPOR9'BEHAVIOUR OF POWDERED AND FLAKE MAGNETIC CORES AT TEMPERATURES1-4 UP TO 500007 OCTOBER 1957 <7*Fk, NAVAL ORDNANCE, LABORATORY9T 'WiI3O3#M RLNNAVORD Report 5752 BEHAVIOUR OF POWDERED AND FLAktMAGNETIC CORES AT TEMPERATURES UP TO 00ocPrepared by:John F.

2 IHaben and Edmopd AdamsABSTRACT: The magnetic characteristics of seven high permeability(55-300) powdered and flake core materials were measured at 500 Ctemperature incervals up to 5000C. Three low permeability (13-22)Carbonyl iron "E" cores for higher frequency (>100 KC) operationwere also examined. Of tite seven core materials tested for <100 KCapplications, Sendust powder showed the least irreversible changein permeability and eddy current coefficient after cycling to of the cores tested were stable for operation at elevated tem-peratures, although (JQ) curves showed that the Flakenol core was themost constant over the temperature range. The commercially available2-81 Mo-Permalloy and Sendust powder cores were less constant. Themagnetic characteristics of the high and low frequency carbonyl ironcores insulated with plastic binders were seriously degraded above2250C. Although the cores tested were unstable above 2500C in varyingdegrees, the copper conductor material and its insulation appears tobe the limiting factor for eontinnious 5000C S.

3 NAVAL ORDNANCE LABORATORYW hite Oak, Silver Spring, MarylandINAVO)RD Report 5752 17 October 1957 This report summarizes the changes in the magnetic characteristicsof powdered and flake magnetic cores at intervals up to 5000C. Thisstudy was undertaken for the Information of the Navy under BuOrd TaskNo. 503-725/54025-01040 and BuShips Project NOL-266, because noinformation was available on the temperature stability of powderedmagnetic cores above 2000C for use in high frequency W. WILBOURNEC aptain, USNC ommanderL. R. MAX ELLBy directioniNAVORD Report 5752 CONTENTSPageIntroduction IExperimertal Measurements ID 5"rissimn of Results 5 Permanent Effects of 5000C Exposure 5 PermeabilityEddy Current and Total Losses .5 Quality Factor (JJ) 6 Sumary and Conclusion 6 Appendix IEffect of "emperature on High Frequency CarbonylIron Cores I8 ILLUSTRATIONSF igure 1. Equipment Block Diagram 7 Figure 2.

4 Effect of Temperature on Permeability(All Cores)Figure 3. Temperature Effects (Sendust Powder)Figure 4. Temperature Effects (o-Permalloy)Figure 5. Temperature Effects (Carbonyl Iron) 11 Figure 6. Temperature Effects (Alfenol Powder -16%) 12 Figure T. Temperature Effects (Alfenol Flake -14%)Figure 8. Temperature Effects (Flakenol) 14 Figure 9. Temperature Effects (Flakenol) IFigure 10. Effct of Temperature on Quality Factor ) PIFigure 11. Temperature Effects (Carbonyl Iron, MV")TABLEST able I. Types and Dimensions of Magnetic CoresEvaluatedTable II. Magnetic Characteristics Before and AfterHeating to 5000C 4iiiNAVORD Report 5752 BEHAVIOUR OF POW)ERED AND FLAKEMAGNETIC CXRES AT TEMPERATURES UP TO 5000 CINTRODUCTION1. The principal requirement of magnetic powder cores in most inductorapplications is that their magnetic characteristics remain relativelyconstant with changes In temperature, hunidity, flux level and all these factors, it was considered most urgent to determine theeffects of temperature since many electronic components uay have tooperate at temperatures as high as 5000C.

5 Although some if~ects oftemperature on powdered core inductors have been measured , , thesewere confined to a narrow temperature range, the maximum limit usuallynot exceeding 1900C. Therefore, the changes in permeability, total lossfactor and eddy-current loss czefficient were determined for variousrepresentative magnetic powderea and flake core materials over a tem-perature range of 250 -5000C. Most of the samples represent corematerials for use at low flux densities and at frequencies up to100 KC/S. In addition, three powdered carbonyl iron cores for use athigher frequencies ( -35 mc/s) were also examined. The data forthese cores are presented in the Appendix. Table I lists the corematerials that were evaluated in this MEASUREMENTS2. Because the temperature stability of magnetic powdered cores variesto some extent with core dimensions, whenever possible, samples of equalsize were selected.

6 The dimensions are shown in Table The toroids were wrapped with glass tape insulation and wound with25 turns of AWG 930 "Silotex" (Anaconda Copper Type DX Class H) is the trade name arolied to magnet wire insulated with alkali-free glass fiber yarn and bonded with silicone varnish. Although it israted at 1800C "hottest spot temperatures", motors wound with Silotexmagnet wire have operated continuously at over 2000C without impairivugthe The corve were placed in a laboratory box furnace capable ofoperating up to 12000C with a heat zone of 3" x 4" x 10". They werearranged so that each core and respective leads were separated fromits neighbors. Since the maximum temperature selected (5000C) waoutside of the wire manufacturer's spetcifications, every effort wasmade not to disturb the cores and leads after positioning. Aftercompletion of the high temperature measurements, it was; found that theinsulation had deteriorated and the copper wire was extremely brittleand Report 5752 TADLE- ITYPES AN) DIENSIONS OF MAGNETIC (DRES EVALUATECORE _____ DIWN~sO N (C )No MATERIAL Ufa.

7 Sendust Powder WI 5 .0005.~A,9,%SiIa1Fe2 Mo-Permali, 1Y * Az 120 112% Mol 81 M, $81 re3 Carbomyl Iron GS-6 jNOL TO 10 AX0. 514 16-Alfenol Powder MOL Al* Bal. Fe5 Flake NDL 3M , All 581. Fe ).6 Fiakemol p0L 220 (Seadust flak*)7 Flakepol, 0 NOL 1 * 7 ;N o. 72(Seadust Flake)a Carbovnyl Iron~ E c(m 13 PA9 Carbonyl Iron -lE" E')-1 L&2 1.#010 Carbonv4 Iron "E0m Tc-l 21 I 1JA*Temperature Stabilifi5 Manu facturer:Tmi Tohoku M0tul 1uwiustrios, 3,paAAE -Arnold togieerin Cmpsny, Xaron, IlliAcigNLmNaval Ordssnco* toratory, 010 )' ** ftryllsoEQ-1) Furniskod by Navy, ftroov of ShipsTC- 1)NAVORD Report 57525. The magnetic characteristics measured with respect to temperaturewere (1) permeahility (2) total loss factor (R/uLf) and (3) Legg'seddy-current coefficient (e). It has become standard practice tomeasrfe these properties in compacted ferromagnetic powders and ferritesby means of a single-layer winding upon a toroidal specimen of thematerial, The permeability, total loss factor (R/uLf), and eddy currentcoefficient (e) can be determined from bridge measurements of thecurrent, inductance, and resistance of the toroidal winding.

8 The valuesthat appear in this report were obtained by methods based on work byElraen, Legg and 0wens5. A block diagram of the measurement circuitappears in Figure 1. The equipment used in this study was as foll(ws:(1) Maxwell Bridge, Western Electric W-10004-6(2) Beat Generator, Boonton, Type 140-A(3) VTV, Hewlett Packard, Type 400H(4) Modified Burrell Globar Type Box rn obtaining the a-c loss component of the core material itsilf bythe above method, the shorted inductanc( of the bridge must be subtractedfrorm the inductance of the toroldal winding. In a similar manner, theshorted bridge resistance must be subttacted from the a-c resistance ofthe toroidal winding. The d-c resistince of the toroidal winding mustalso be subtracted from the a-c bridge resistance value. Because ofthe chanqe in resistance with temperature, a d-c resistance reading wasobtaiued at each test temperature,7.))

9 The temperature was raised in 500C steps and the cores were allowedI to soak 4or approximately one hour at each temperature. Since the*easure-ents were made over a period of a week, at the close of eachworki i day the cores were allowed to remain overnight at the last* temperature. The next morning the inductance of each core wasro~astired; no significant changes were found over the previous ewh temperature the inductance and a-c resistance were determinedior each toroid at three frequencies; 25, 50 and 75 KC/S. (Figs. 3-1).The teeperature was increased until a substantial decrease in permeabilityociurred at around 00 C. This temperature was approximately the Curiett p#rature for most of the test samples except for the carbonyl iron core%hich itas a Curie Temperature of 7700C. At the end of the study thefurnace was al'oed to cool to room temperature overnight. The resistance of the cores were then remeasured at 25, 50 and 75 <C/Sto' detormine the teount of temperature hysteresis which occurred.

10 Sincet, porrentage chanq was about the same at each frequency, the permea-lzity and loss data shown in Table II are for KC/S Report 5752 TABLE IIMAGNETIC CHARACTERISTICS BEFOREAl) AFTER HEATING TO 5 oC' Percentage ChaneNo. Core p0 RxlO6* e x 109 F RxlO6 e x 109 Material JLf --lI Sendust Powder 264 237 13 02 Mo-Permalloy ** 117 452 16103 468 19 Carbonyl Iron 254 16600 890. 6400 188004 16-Alfenol 114 1240 33 Powder 100 1350 35 Alfenol 306 1550 53 Flake 232 1590 56 Flakenol I 222 270 294 Flakenol I ** 154 395 587 KC/; B = 5 gauss* Temperature StabilizedNOTE: The data on the first line for each core material gives the valuesbefore heating; the second line gives values after Report 5752 DISCJSSION OF RESULTS8. Permanent Effects of 5000C Exposure. The amount of irreversiblechanges in the magnetic properties of the powdered and flake coresafter exposure to 5000C Is summarized in Table IT.