1、c_(D-T6.- Master yield and rupture curves for7075-T6 employing the linear temperature parameter (2) are shown in fig-ure 23. Good correlation was found particularly for yield temperaturesat 40 ksi and 60 ksi where the test points are closely superimposed. Thecorrelation for rupture temperatures is f
2、airly good except at 40 ksi wherethe scatter of the data is considerable.The master curves for yield and rupture are linear from about 17 ksito 65 ksi. The yield stress _y, which corresponds to the yield tempera-ture Ty, may be expressed asI Ty + 2O071Oy = 133.5 - 2.55iog h + iThe rupture stressbe g
3、iven asor at which the rupture temperature_Tr + 2007)Or = 143.5 - 2-69_iog h +Tr occurs may(6)Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACA TN3462 iiIn equations (5) and (6), the stresses Sy and _r are in ksi, thetemperatures Ty and Tr are in
4、OF, and the temperature rate h is inOF per second. These formulas are limited to stresses from 17 ksi to65 ksi.The correlation of the data for yield and rupture temperaturesemploying the reciprocal temperature parameter (4) was almost identicalwith that obtained with the linear parameter (2). Conseq
5、uently, themaster curves based on the reciprocal parameter are not included.The validity of the linear parameter is shown by the correlation ofthe data with the master curves (fig. 23). A more critical evaluationof the accuracy obtainable by the use of the master curve and the param-eter may be had
6、by making a comparison of predicted or calculated yieldand rupture temperatures with the test results at different stress levels.Calculated yield temperatures (fig. 12) agree within 10o F with the testresults. Calculated rupture temperatures (fig. 16) are also in closeagreement with the test results
7、 except at 40 ksi where there is a maximumdifference of about 20 F. Similar calculations, based upon the mastercurves, using the reciprocal temperature parameter (4) were in very closeagreement with those shown in figures 12 and 16.Master curves for 2024-T3.- Master yield and rupture curves for2024-
8、T3 usingthe linear temperature parameter (5) are shown in figure 24.The correlation of the data is not so good for this material as thatobtained for 7075-T6 (fig. 23). The correlation is very poor at 50 ksifor yield temperatures and only fair at 40 ksl for rupture temperatures.The master curves for
9、yield and rupture can be assumed to be linearover part of the range. The yield stress ay under rapid-heating con-ditions can be given asThe rupture stressrY + 20091_y = 106.0 - 2.03 og h + 1_r can be expressed as(7)(8)Tr+ 20091_r = 127.5 - 2.38iog h + iIn equations (7) and (8), the stresses _y and _
10、r are in ksi, thetemperatures Ty and Tr are in OF, and the temperature rate h isProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-12 NACATN3462in OF per second. Equations (7) and (8) are limited, respectively, tothe 16-ksi to 40-ksi and 16-ksi to 50-ks
11、i ranges.As in the case of 7075-T6, the correlation of the data for yieldand rupture employing the reciprocal temperature parameter (4) waspractically the sa_e as that obtained with the linear temperatureparameter (2). Master curves b_sed on the former are therefore notincluded.Calculated yield and
12、rupture temperatures based upon the use ofthe linear temperature parameter are comparedwith the test results infigures 13 and 17. Fair agreement is obtained except at 50 ksi for yieldtemperatures and at 40 ksl for rupture temperatures. The agreementbetween calculated and experimental results_ howeve
13、r_ is not so good forthis material as that for 7075-T6 (figs. 12 and 16). It is not surprisingthat the sameparameters do not work so well for 2024-T3 as for 7075-T6because aging of the former markedly alters the general pattern of theresults. Calculated yield and rupture temperatures_ obtained by me
14、ansof the reciprocal temperature parameter (4), were in close agreementwith those shownfor the linear temperature parameter (3)-CONCLUDINGIn the rapid-heating tensile tests of 7075-T6 and 2024-T3 aluminum-alloy sheet under constant load and temperature rates from 0.2 F toi00 F per second, yield and
15、rupture temperatures were found to increaseapproximately in proportion to the logarithm of the temperature rateexcept in certain regions for 2024-T3 aluminumalloy where aging affectedthe results.Under rapid-heating conditions_ yield and rupture stresses maybesubstantially greater or about the samefo
16、r a given temperature as corre-ponding stresses obtained from elevated-temperature tensile stress-straintests for i/2-hour exposure, depending upon the temperature rate andmaterial. The increase in yield and rupture stresses with temperaturerates for a given temperature becomesfairly small at rates
17、above60 F per second.Linear and reciprocal temperature-rate parameters madeit possibleto take into account the effect of the temperature rate and to constructsingle or master curves of stress against the parameter. These curvesprovide a convenient method of obtaining yield and rupture stresses andPr
18、ovided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACATN 5462 15temperatures for a given temperature rate. Good correlation of the datawith the master curves was obtained except where aging altered theresults.Langley Aeronautical Laboratory,National Advi
19、sory Committee for Aeronautics,Langley Field, Va._ March 25, 1955-Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-14 NACATN3462APPENDIXDESCRIPTIONOFTESTEQUIPMENTFORRAPID-HEATINGTESTSLoading EquipmentThe general arrangement of the loading equipment is
20、 shown in fig-ure 2. The load was applied to the specimen through a 5:1 beamor leversystem by meansof weights. The fulcrum of the beamwas supported on aplate mounted near the top of the three columns of a 120,O00-pound-capacity two-way hydraulic Jack. The weight-loading system had a maxi-mumcapacity
21、 of lO,000 pounds on the specimen. Knife edges were employedat the fulcrum and other loading points of the beam. With this system,the beamcould be positioned and the weight cage could be lifted fromthe floor by lowering the ram of the two-way Jack.The method of connecting the specimen to the loading
22、 system is shownin more detail in figure 5. The load was applied to the specimenby meansof yoke-and-pin connections. In order to prevent grounding at the top andbottom of the specimen, which was part of the electrical circuit, theloading bars were electrically insulated from the rest of the system.T
23、his insulation was provided by the two-plece rectangular load insulatorsshownabove the top column plate and above the ram.Because a discontinuity of slope was obtained in the tlme-temperaturerecords at the beginning of heating, the possibility of inertia effectsin the load system was investigated. L
24、oad variations were measuredbymeansof a calibration bar in series with the specimen. The variationin output of two wire strain gages mounted on the bar was measuredwitha pen-type recorder with a 100-cycle-per-second response. No detectablevariations in load occurred within the range of temperature r
25、ate coveredby the tests. Small load oscillations were obtained, however, at ratesabove 190 F per second.Strain-Measuring EquipmentThe extensometer system consisted of two pairs of extensometerframes mountedopposite each other and 1 inch apart, and two strain-transfer units, each of which actuated a
26、variable linear differentialtransformer-type gage. Details of the extensometer frames and contactends of the strain-transfer units are shownin figure 4.Each extensometer frame was madewith a double-edge knife edge anda wire arm serving as a stabilizer for the knife edge and keeping itnormal to the s
27、pecimen surface. Spring clips were used to hold the knifeProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-_ACA TN 3462 Z5edges in firm contact with the specimen surface and were mounted in sucha manner as to keep the frictional forces at the ends of t
28、he arms at aminimum in order to avoid knife-edge slippage. Co_nercial lava blockswere mounted at the ends of the arms so as to insulate the arms fromthe specimen and prevent possible burning of the knife edges duringheating and at rupture. The globular material visible at the ends ofthe frames (fig.
29、 4) consisted of a ceramic cement which held the lavablocks securely to the wire arms. The frames were mounted and spacedi inch apart on the specimen by means of a gage block.The strain-transfer units were mounted independently of and atright angles to the specimen so as to keep the effect of temper
30、ature onthe strain-transfer unit and the differential transformer at a minimum(figs. 3 and 4). The tubular arms of the strain-transfer units weremounted on flexure plates about at their midposition. A vertical dis-placement of the specimen with respect to the strain-transfer units didnot result in a
31、n indicated strain except for slight variations due toextensometer-contact irregularities and small angularity effects. Theoutput of the two gages was added so that an average value of strainwas obtained. Each end of each straln-transfer arm was in point contactwith the flat horizontal surface of on
32、e of the knife-edge units (fig. 4).The arms were weighted in such a manner that positive contact was main-tained throughout the test. In order to avoid damage to the knife edgesand strain-transfer units under rupture conditions, end pieces supportingthe contact points were attached magnetically to t
33、he ends of the tubulararms of the strain-transfer units. Release of the joint occurred onlyif more than the normal contact pressure was exerted. The contact pointswere insulated electrically from the end pieces.Heating EquipmentHeating was accomplished by passing a high-amperage low-voltagealternati
34、ng current directly through the specimen. Continuous voltagecontrol was achieved by means of a 400-volt induction voltage regulatorwhich was used to regulate manually the primary voltage of a 75-kilovolt-ampere transformer having an output of 3-8 to 20 volts. The transformerand regulator can be seen
35、 mounted on the wooden stand in figure 2.Two connections, one on each side of the specimen, were used betweeneach end of the specimen and the bus bars from the transformer. Eachconnection consisted of five thin, flexible braided copper leads silver-soldered togetherto form a flat terminal which prov
36、ided a good electricalcontact with the specimen surfaces. These flexible leads were clamped tothe specimen with small clamps at one end, and each lead was bolted tothe bus bars at the other end.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-16 NACAT
37、N 3462Temperature-Measuring EquipmentTemperatures were measuredwith No. 28 gage chromel-alumel thermo-couples which were flattened and held in firm contact with the specimensurface by meansof special mountings and clamps. The thermocouplejunction was fused by meansof a mercury arc. T_o thermocouples
38、 werelocated at the mldposltion directly opposite each other on the faces ofthe specimen so as to provide a check on the temperature measurementsin the region in which the strains were measured. Single thermocoupleswere also clamped at points 3 inches above and below the midpositionto provide inform
39、ation on the temperature gradient. Tworecordingpotentiometers, each having two channels, were used and had responserates of full-scale travel in 1 second. Inasmuch as one channel wasrequired for strain measurements3 three of the four available channelswere used for temperature measurements. Four tem
40、peratures were recordedon three channels by alternately switching manually the output of theupper and lower thermocouples to the input of a single channel duringthe test.The method of clamping the two thermocouples at the midposition isillustrated in figure 9. Each thermocouple was held in positive
41、contactwith the specimen surface by an asbestos pad cementedto a thin backingplate, an extension of which held the thermocouple in a fixed positionon the asbestos pad. The backing plate consisted of a flattened portionof thin Inconel tubing. The pad insulated the thermocouple from the airand the bac
42、ki_ plate. An adjustable clamp provided the pressure neces-sary to insure reliable temperature indications. This method of clampingthermocouples has proved to be quick and convenient. Consistent resultsare possible for the aluminum-alloy specimenswith this arrangement ifintimate contact of the flatt
43、ened thermocouple and specimen surface isobtained, the asbestos pad is in good condition, and adequate pressureis applied. Thermocouplesmounted in this manner can be used repeatedly.In order to determine whether the results obtained with clampedthermocouples were consistent with those obtained from
44、thermocouplespeened into small holes in the surface, comparisons were madebetweentwo clamped and one peened thermocouple located at the sameheight onthe specimenfor a temperature rate of lO0 F per second. The compari-sons showedthat the maximumvariatlon in response rate between the twotypes of therm
45、ocouples was about 3/_ percent. The scatter at 800 F atthe sameinstant was about 5 F for the two clamped thermocouples. Tem-peratures measuredby the peened thermocouple fell between those obtainedfor the c_ed thermocouples. Indicated temperatures obtained with thetwo clampedthermocouples generally a
46、greed within a few degrees. If theinstallation was imperfect, however_ low or inconsistent temperatureswere obtained from either peened or clampedthermocouples.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-3GNACATN 3462 17Specimen Temperature Gradi
47、entsTypical temperature gradients obtained along the length of thereduced section of the rapid-heatlng specimen for temperature rates of2 F and llO F per second are shown in figure 6 for center temperaturesof 400 F to 500 F. Temperatures were measured at the positions shownby means of peened thermoc
48、ouples and an oscillograph recorder having18 channels. In order to duplicate test conditions, the measurementswere taken with the specimen mounted in position for testing and withthe extensometer frames and clamped thermocouples in place. The resultsshow the marked effect of the extensometer frames
49、and clamped thermo-couples on the local temperature distribution. The normal gradient with-out equipment mounted on the specimen is a smooth curve which is concavetoward the specimen. The addition of the extensometer and clamped ther-mocouples depresses the normal curve as the result shows. The tempera-ture variation within the 1-1nch gage length is fairly uniform for 2 Fp
