Патент USA US2403999код для вставки
, §94@ T. iA. READ m. ' `?,403,999 SONIC METHOD FOR TESTING METAL Filed Feb. 22‘, 1944 Fi È- l- 6'/ . A 2 Sheets-Sheet 1 A5’ 38 *?‘ . S G CASE + Il “ f Epi§-ä„ E ---__ . Sí h „',M 231-? V-"Íêô , nasal/ecs $5/ 53 AMPLIFIER _. “l l ‘El -5-“ | ‘ ’ I ¿5 ë 34 ‘t ' 'I FREQUENCY-Í» f « /8 (l ¿e Í * E' g5 ‘D , I l y 5 _. 26 O ._ i ' > » 2 È ë , Q 20 M” ‘f 25 f3 l E' I Erg„àà ._ 34 m »fi 2/ fr ' _4 2"' 21/ ïw Ü @ä 20 ' À V/ß@ 2 /6 _ 050 . \/5\§ l y QÈ «L `__`_‘ ¿__Í-_ï- ‘ , /6\ ß > 4 \ 26 34 E _7- 'N È T 2 . ya f 34 ` _ gfggäsrsàämm THDMAELA_READ,HEREERTl-_FJSFELD’ r . EUMNER ÑÁÄKITCHEN, N ` , . m” cfmdymm _ O l u, , " Patented July 16,_ 1946 2,403,999 UNITED. STATES PATENT OFFICE 2,403,999 SONIC METHOD FOR TESTING METAL Thomas A. Read, Herbert I. Fusfeld, and Sumner W. Kitchen, Philadelphia, Pa. Application February 22, 1944, Serial N0. 523,430 4 Claims. (Cl. ’i3-69) (Granted under the act of March 3, 1883, as amended April 30, 1928; 370 O. G. 757) 1 2 The invention described herein may be manu factured and used by or for the Government for ing same are shown by the accompanying draw governmental purposes without the payment to us of any royalty thereon. Our invention relates to the testing of metals and it has particular reference to methods for detecting ilaws, cracks and other defects in metal specimens. ings wherein: Fig. l is a diagrammatic showing of brass car tridge case testing equipment which incorporates the principles of this invention; Fig. 2 is an enlarged section view showing preferred mechanical constructions for Fig. l’s case support and driving magnet; Broadly stated, the object of our invention is Fig. 3 is a horizontal section on line 3-3 to provide improved procedure by which the pres 10 through the top plate and circular air gap of the ence of internal flaws in metal specimens of vari ous compositions and shapes may be detected positively, quickly and reliably. driving magnet of Fig. 2; Fig. 4 is a showing of further details of the strain gauge which is attached to the test speci A more speciñc object is to make special ad men side ; » aptation of our improved testing method to the 15 Figs. 5, 6 and 7 are curves illustrative of cer examination of metal specimens which are cylin tain principles upon which our improved test drical in shape and non-magnetic in character. method is based; ‘ Another object is to provide a method of test Fig. 8 illustrates alternative forms which Fig. l’s ing the brass cases of fired and other artillery case driving source and amplitude measuring cartridges for the presence of dangerous season 20 means may take; cracks. ' Fig. 9 shows further variations of the basic system of Fig. 1; .and A further object is to provide a reliable test for the susceptibility of fired and resized cartridge Figs. 10, 11 and12 show alternative arrange cases to splitting or other fracture on subsequent ments by which the amplitude of case vibration reñring. 25 may be registered and measured. A still further object is to provide a test methodV THE TESTING APPARArtrs 0F FIGS. 1-2-3-4 by which it is possible prior to resizing to reject In Fig. l we have shown our improved testing all cartridge cases which after being resized and facilities as organized for detecting the presence reconditioned will crack on subsequent firing. The improved metal testing method of our in 30 of flaws in a brass cartridge case I5. For clarify ing the description to follow it will be assumed vention is predicated on our discovery that the that this case is designed for use in a 75 mm. “damping” characteristics of vibrated metal artillery piece. As the description proceeds it specimens will sharply reiiect the presence of in will become evident that other sizes and shapes of ternal iiaws if the specimen vibrations have such cartridge cases may be tested with equal facility, intensified amplitude as to set up in the metal and that metal specimens of compositions other peak stresses which are far higher than any here than brass and of shapes and contours other than tofore employed by tests of the “sonic” character represented at I5 likewise lend themselves to here considered. test by our improved flaw detecting method here In practicing our invention we attain the fore in disclosed. going and other objects by exciting the metal Essential elements of the equipment shown in specimen to longitudinal vibration at its natural Fig. 1 include: (ai) a case vibrating magnet M; or resonant frequency; by so building up the (b) a source B-C of alternating current power amplitude of this resonant vibration that the for energizing a driving winding I6 oi the mag peak stress thereby induced in the metal has an net; (c) apparatus including a strain gauge S for exceedingly high intensity (of the order of ñve to measuring the amplitude of vibration that is inten thousand pounds per square inch for brass duced in the test case I5; and (d) current and cartridge case metal); by measuring the damp voltage measuring instruments A and V for giving ing capacity (i. e. the internal friction) at this the ratio of the driving power input to the strain high amplitude of vibration; and by comparing gauge output. the observed measurement with known stand THE CAsE DRIVING MAGNET ards for purposes of classifying the metal in the specimen either as sound or as defective. The case driving magnet M has the construc Illustrative embodiments of our improved test tion best shown in Figs. 2-3. A central core I8 ing method and of apparatus suitable for practic of magnetically soft iron is fastened to a lower L 2,403,999 3 4 plate IQ of the same material and substantially square in shape. Extending upwardly from each of the four edges of this lower plate are side plates 2S, also of magnetic iron. Fitted into the square opening formed by the upper edges of these side plates is an iron top plate 2l having in its center a round Opening somewhat larger than described are found to have a natural vibration frequency of the order of about four thousand cycles per second, and in conducting tests on such cases source B-C thus should be capable of sup the top >of the central core I8. plying case driving energy at four thousand cycles per second and further be adjustable through one or two hundred cycles above and below the stated The ’ utility of this adjustment will Ibe come further evident as the description proceeds. The step down transformer T is for the purpose value. ` The annular space between magnet core I8 and the surrounding plate metal 2| constitutes an airl gap through which a direct current winding 24 causes te flow unidirectional magnetic flux from of converting alternating current energy of the the central core I8 in all radial directions to the source B-C into energy of the lowered voltage surrounding top plate 2l. This core Winding 2li and relatively high amperage which is required for satisfactory operation of the case driving moderate-voltage potential readily availalble at is continuously excited by any suitable source of direct current, designated in Fig. 1 by the termi-` nal “-i-” and “-.” Energizing leads 25;?5 are brought from winding 24 to the outside of mag net M in any suitaßlole manner suchv as that indi cated in Fig. 2, i magnet M. ers on an insulating tube about 8 inches in diam 20 eter, and having a secondary made up of a single The outside diameter of the magnet’s central core I8 is slightly less than the inside mouth diameter of the cartridge case l5 to be tested, while the inside diameter of the central opening in the magnet’s top plate 2l is somewhat larger 25 than the mouth diameter of the same case. This permits the case mouth to ht down into the mag net’s annular gap in a manner clearly- indicated In one design which has proven sat- , isfactory this transformer makes use of a primary having 8S turns of No. 18 wire wound in two lay copper strip 2 inches lv'v'ide by .020 inch thick wound as a single turn around the same tube. The particular transformer used is of the air core type. This copper strip of 2 inches by .020 inch cross ’ section is, in the test equipment now described, continued from the transformer T into the air gap of the driving magnet M where it constitutes by Figs. 1-2-3 and to a depth of slightly over 11/2 the single turn driving winding shownV at I6 in inches. In this position there is passed through 30 Figsv 2_3. For producing `cartridge case vibra the ease mouth metal the radial magnetic flux tions of the elevated order required by our test earlier referred to as being set up by direct cur method currents of the order of several hundred rent winding 2li. Preferably this unidirectional amperes are supplied .by the secondary of this magnetic force has an intensity of the order of transformer T. 10,000 oersteds. The load thus presented to the power source Positioned between the case mouth metal'and B--C is found to be predominantly inductive, and the-magnet plate metal 2| is the alternating cur without corrective means results in an objection rent driving winding I5 earlier mentioned. Pref ably low power factor. To compensate for this erably this winding takes the form of the repre use is made of a capacitor 35 series connected in sented single turn of copper strip 'secured inside 4 the transformer primary `circuit as shown in Fig. the opening in plate 2l in any suitable manner l.’ By choosing the capacitive rea'ctance of ele such as shown in Figs.- 2-3. There thin layers of ment 3S to equal the inductive reactance of trans insulation 28 separate the copper strip I5 from former T and its connected load, the power fac the. plate metal and at the same time mechani tor of energy drawn from the alternating .current cally support the Astrip through a «bonding thereof source B-C may be made substantially unity. to the metal 2 l . Under these conditions the driving magnet M is For holding the cartridge case I5 in_the repre observed tol impose a load on the source B-C sented test position use may be made of any suit which approximates a pure resistance of about 12 able means which impart mechanical support ohmsi Y without interfering with case vibration. Such The described magnet M and energizing sources means may take the form of a support plate 33 therefor constitute an “eddy current” metal held at suitable distance above the top of magnet drive by which there is exerted on the case mouth M Iby corner uprights 3d and having a central mechanical forces which alternately act upward opening somewhat larger than the outside diam ly and downwardly This action results from the eter of the cartridgey case VHi. Two or more 55 fact that the alternating 4current in the driving strands of piano wire 3l are stretched at moder winding l5 induces corresponding`- eddy currents ate tension across this opening «between securing in the case mouth metal. The interaction of screws S2. By restraining downward movement these induced eddy currents with the radially of the case head rim, these wires 3l mechanically flowing unidirectional magnetic flux produces on suspend the case in the test position represented. 60 the case mouth metal an alternating mechanical force »which is 4directed along the axis of this case. SUPPLY or' CASE DRIVING ENERGY This alternating mechanical force reverses its For causing the just described magnet M to direction in step with the frequency of the cur drive the cartridge case i5 at its natural rate of rent from driving source B-C, and when the fre vibration use is made of the facilities which the 65 quency thereof is chosen to match the test case’s left portion of Fig. i represents. These facilities natural frequency of vibration there may be pro include the earlier named source B-C of alter duced in the case longitudinal vibrations of the hating current power plus a step down trans extremely elevated magnitude which the test former T, a capacitor 36, the vearlier mentioned method of our invention requires. amine-ter A and a frequency meter F. Since the case driving forces are directed alter 70 Source B-C may take any suitable form capa nately in opposite directions at a frequency hav ble of supplying up to about one kilowatt of power ing the stated value of around four thousand at a frequency which exactly matches the natural cycles per second, the average force of each of. or resonant rate of longitudinal vibration for the these complete cycles is zero and this makes it cartridge case l5. Cases ofthe 75 mmf size here' 75 possible to support the case in a position in the 2,403,999. 5. 6a magnet gap bythe light Wires'indicated at 3I ment possible use is made ofythe amplifier' 38. in engagement with the rim at the head of the Thisampliñer may be any one of a number >of case. commercially available forms, and for this rea ' \ son no attempt to show details has been made. THE AMPLITUDE MEASURING MEANS Exciting its input terminals is the potential appearing across the resistor 43. .Each change therein is magniñed many times by the ampliñer For measuring the amplitude of the mechanical vibrations thus induced in the test case I5 use i's made of the earlier named strain gauge S plusv and impressed upon a suitable measuring device an amplifier 38 plus the earlier named volt meter V. 4 indicated as voltmeter V. 10 As is more clearly shown in Fig. 4, the strain gauge S consists of a series of back and forth - When-appropriately calibrated, this voltmeter can thus be made to yield a direct indication of the amplitude at which the cartridge case I5 is loops of electrical resistance wire 4!) mechanically being vibrated yduring practice of the improved bonded together by suitable insulating material test method of our invention. Conveniently this> and further bonded to the wall side of the car 15 indication may be in terms of the millivolts of tridge case I5 under test. As a result of this potential which appear across resistor 43, and bonding any change in the length of the test in connection with certain test data later pre metal I5 produces a corresponding changeV in the sented use is made 0f a voltmeter indication ex mechanical length of the‘gauge loops 40. pressed in such terms. These loops are connected in series and the 20 The combination just described is thus so or~Ã ganized that complete absence yofvibration on resistance variations of’ each which accompany changes in its effective length are additively com-V the part of the test case I 5 will cause meter V to bined to give a total resistance change which give what may be termed a “zero” indication; lends itself to ready measurement or detection. that vVibration of moderate amplitude on the part of case I5 will produce an intermediate reading Strain gauges of the type shown at S are commer by voltmeter V; and that vibration of high am-~ cially available, and one design found especially plitude by ca'se I5 will produce acorrespondingly suitable has a static resistance of 500 ohms and high reading by thevoltmeter. a strain sensitivity of 3.5. By strain sensitivity is meant the ratio of the fractional change in re sistance of the gauge to the fractional change in gauge length. ‘ For most effective response to longitudinal vi . bration of the test case I5 the strain gauge S " OPERATION oF THE FIG. 1 TEST EQUIPMENT In applying the just described test equipment of Fig. 1 to ñred ’l5 mm. brass cartridge cases we are able to predetermine with high accuracy all of those cases in the tested group which will rup should be attached at some intermediate point between the two case ends, such as is sho-wn in ture upon resizing, reloading and subsequent fir Figs. 1_2. This particular location was selected after experimentation which showed that in lon» ing, and which of the cases in the group will with stand the reñring without rupture or other fail ure of the case metal. ` gitudinal Vibrations of the character here con The need for such determination has been felt sidered the effect is analogous to simple compres sion and stretching from the two case ends. This 40 in a practical way for a long time. Splits which have been obtained in the firing of resized brass results in a point known as the displacement cases are now ascribed to the presence of season “node,” at lwhich mechanical movement of the cracks. These cracks are caused by the at case divides and further at which there is no tacks of the products of the burning of the pro change of mechanical position. For cartridge cases of the 75 mm. design shown, 45 pellant powder during the period between the ñring of the round and the cleaning of the case. this point is approximately one-third of the case Such cleaning is subject to considerable delay length distance from the head end. Mounting by reason of the fact that in many instances the of the strain gauge S at this point results in the cases lay out in the ñeld exposed to weather for maximum movement between the two gauge ends; and while other positions are found also 50 long periods of time after their original iiring, and when subsequently subjected to resizing op to give indications of the case vibration amplitude erations preparatory to a second use there fre their effect is relatively less than the nodal point quently have developed minute cracks, :particu mounting here shown. larly on the side Wall interiors. y For converting variations in strain gauge re These are so'inconspicuous and diflicult to de sistance into corresponding variations in poten 5,5 tect that they go completely unnoticed through tial, use is made of a battery or other direct cur the entire sequence of resizing operations and are rent source 42 of constant potential connected only discovered upon reiiring of the case, when through the strain gauge loops 40 in series with their presence then results in longitudinal splits a resistor 43. . As long as the strain gauge remains static the 60 or other failure of the case. Such failures not only endanger -personnel ñring the weapon but current flow through resistor 43 is constant and by reason of loss in chamber pressure they so alter the voltage ‘appearing thereacross remains un the 'ballistic performance of the fired projectile changed. Each decrease in strain gauge resist that aiming becomes inaccurate and unreliable. ance which accompanies an effective compression of the cartridge case I5 raises the resistor cur 65 In addition the escaping gases erode the weapon’s chamber and firing pin causing early malfunc rent and produces a corresponding rise in volt tion of the weapon. age between the resistor terminals. Similarly, Until the advent of our invention, no method each increase in strain gauge resistance which was available for reliably separating the good accompanies an effective stretching of the test case I5 produces a corresponding drop in resistor 70 from the bad prior to case resizing or at any time prior to case reuse. With our method, however, current and a resultant lowering of the resistor the desired selection can be made quickly, re terminal voltage. liably and effectively. . ' The magnitude of these changes in resistor In Vusing the earlier described equipment of voltage is relatively small and in order to increase Fig. 1_ each of the cartridge cases to be testedis-A it to‘such an >extent as to make ready measure .tabacco 7 subjected to onlytwo preparatory operations. The primer is punched out of the case head, .and the mouth of the case is restored substantially to 8 spending elongation in the side wall metal dabove the nodal point. During resonant vibration, therefore, a simple of the case mouth ñts into the `annular gap of .the “bellows” action takes place wherein the case metal at the nodal point and on both sides there of (with the exception of the case extremities) alternately shortens under compression and lengthensr under tension. Although the nodal driving magnet M and is there supported by a point completely lacks longitudinal motion, some> its original circular shape. , l _ . So prepared the case then is lowered through the ‘opening Vin the support `plate 33 (see Fig. 2) to the `represented position where the’wall Vmetal resting of the case head rim on the supporting 10 transverse expansion and contraction of the case circumference appears there to be present. wires 53|. The strain gauge S is now attached to For the relatively low intensities of induced case the case >side wall >at the location between the vibration which have heretofore been used in at two ends 0f thev case determined as earlier described. - Direct current is now applied to the magnetM’s tempting to test material specimens, it is ob served that a defective specimen exhibits prac tically the same “damping” capacity as does a central core winding 2e, and alternating current sound specimen. In the curve of Fig. 6 such rela from source B-C is applied to transformer T and tively low vibration amplitudes are indicated by thence tothe case driving winding IS. ri'his di the vertical line 45, and typically they have re rect current supply is adjusted to a value prede termined as suitable for the test and which yields 20 sulted in peak stresses in the test metal of the order of five hundred pounds per square inch or the required intensity of uni-directional flux in lower. y the circular gap surrounding the magnet’s central We have discovered that when the amplitude core head I8. of the longitudinal test case vibrations is in The frequency of the alternating current from creased to a substantially higher value, such as source B-C is, by the aid of meter F, roughly ad is designated at 46 in Fig. 6, a defective case will justed to what is expected will match the reso show a substantially higher damping than does a nant frequency-of vibration for the case I5. An sound case. By damping is meant the internal observation at meter V of the resulting amplitude friction of the case metal, and one convenient of case vibration is taken, and holding the case driving current as shown by meter A at a con 30 measure thereof is given by the ratio of the case stant value, the frequency of the alternating driv ing current is varied'in small steps in both direc tions `until .the amplitude meter V shows a maxi driving force to the resulting vibration amplitude as indicated through the strain gauge S by Fig. i’s voltmeter V. During the remainder of this . description a quantity proportional to that ratio will be designated as the “sonic test coefficient.” This maximum indicates that the frequency of 35 Observations made by us show that a sound the alternating driving current now exactly test specimen has a damping-amplitude curve of matches the resonant frequency of vibration for the relatively flat character shown at 41 in Fig. 6; the test case i5. This indication follows from thatspecimen metal having defects of moderate the well known resonance curve illustrated in Fig. quantity has a relatively steeper curve such >as is 5 which shows how the test case I5 responds in 40 shown at 48; that the presence of fewer defects vibration amplitude to driving current frequen causes the curve to be less steep, as shown at 49; cies above and below the resonant value which and that defects present in larger number result the dotted vertical line designates. in a curve of the greater steepness shown at 50. Explanation has already been given of how the Accordingly, in operating the Fig. 1 equipment eddy 'currents induced in the case mouth metal the driving current supplied from source B-C interact with the uni-directional flux iiowing to the magnet winding I6 is next increased to through lthat metal to set up in the metal mechan such an extent that the resulting resonant vi ical- forces which alternately act upwardly and bration of the test case l5 rises to an amplitude downwardly instep with the reversals of current of the high order shown at 46 in Fig. 6. mum reading. induced in the case mouth metal. At frequencies ~ of the order of four thousand cycles per second the duration of each of these forces is extremely That high order amplitude is known to be at short. Each upwardly acting force pulse tends tained when meter V gives a reading correspond ing to about 12 millivolts input to the amplifier. It is accompanied by peak stresses in the case to compress the side wall metal above the case _ metal of the order of from live to ten thousand mouth, while each downwardly acting force tends to stretch or elongate it. pounds per square inch. Correlation of the read ings of meter V' with such stresses can be made The inertia of the total case mass prevents the in any one of a number of manners so well known complete case from following these pulsations, to the metals testing art that description at'this point isdeemed unnecessary. andv even though applied at one end only their At these high stress amplitudes of longitudinal effect is to set up longitudinal vibrations toward (Si) vibration there is a relatively wide variation in and away from an intermediate “nodal” point the damping capacities of sound and defective along the case length. Were both ends of a, sim cartridge cases. As Fig. 6 indicates, the greater ple vcylindrical test specimen to be open this point the defects the higher the damping and hence the would `be substantially midway. Closure of the more driving force that is required to induce the top end by the case head shifts this neutral point high vibration amplitude. upwardly to about two-thirds of the total distance Since for a given Vibration amplitude there is from the case mouth. a more or less direct relation between the speci Our observations show that at resonant fre men damping and the driving power require quency for Vthe case each compressive action in ments, those cartridge cases which are defective the case side wall below the nodal point is accom- ` require considerably more power from driving panied by a similar compressive action in the source B-C than do cases which are sound. One case side wall above the nodal point; likewise, measureof this power is the reading of ammeter that each side wall stretch or elongation below_ _ A. It is found that the impedance of ¿the case; the nodal point ls also accompanied by corre ' driving circuit remains substantially constant re 2,409,999 v9 10 gardless of whether the case is sound or defective, and for this reason the reading of ammeter A may be assumed to be directly proportional to This data shows an excellent correlationÍ be tween the sonic test coefiicient and the incidence the driving power input. the group was a result of its selected components for it was expected that most of the cases with of ñring splits. The high percentage of splits in ANALYSIS oF TEs'r RESULTS a coemcient above 0.50 would split.l f In order that some variation in the values of test vibration amplitude may be permitted, we prefer to analyze comparative test results through a ratio earlier termed as the “sonic test coefficient.” This coeñîcient is the quotient of the ampere current reading by meter A to the milli volt potential appearing across resistor 43 as read THE MODIFIED TEST SYSTEM 0F FIG. 8 In Fig. 8 we have represented alternative forms of case driving power supply and of vibration amplitude measuring means. These diiîer fromthe corresponding elements earlier described in connection with Fig. 1 in the manners now to be by meter V. For test system constants having pointed out. one particular set of values, new o-r sound cases exhibit a coeiiicient value of approximately 0.30; fired cases having defect contents within toler ances acceptable for retiring show coeñicients of 0.50 and below; and fired cases having defects suiiicient to cause rupture after resizing and sub must be capable of generating a frequency which matches the resonant vibration frequency of the test case I5 and should in addition be adjustable through a small range on either side of theres onant value. It feeds into the power ampli fier 53. This power amplifier 53 corresponds to the source B-C of Fig. 1 and should have an output capacity of approximately one kilowatt; in ener gizing transformer T it performs exactly the same sequent firing have test coefficients above 0.50. This relation is best indicated by the curve of Fig. 7 wherein the horizontal dotted line indi cates the point at which acceptable cases should be separated from unacceptable ones on the basis of proven test results. One set of data which establishes the foregoing sonic test correlation is presented by the accom panying Table X. TABLE X Some test correlation Sonic test Firing Sonic test Firing Cabe ., ' coeíicient result Case ` coefficient result 161 0. 368 0 202 0. 760 :I: 166 108 79 116 . 400 . 400 . 417 . 433 0 0 0 0 200 102 127 115 . 778 . 779 . 784 . 796 0 0 x + 158 205 153 117 . 437 . 442 . 450 . 450 0 0 0 0 . 451 . 460 . 467 . 468 . 500 0 0 0 0 0 . 800 . 804 . 825 . 842 . 847 . 850 17 I O m 7C 159 111 100 131 204 64 169 137 150 129 160 . 886 . 888 . 919 . 933 . 940 66 62 77 67 165 . 500 0 . 516 . 544 . 550 . 568 . 577 0 0 a: 1: I 103 . 591 . 592 . 616 . 636 0 0 :t 0 121 1. 02 z 126 136 140 1. 14 1. 32 1. 33 :t-l :t-I :r :L 201 208 162 74 ' 125 141 . 952 . 955 172 206 . 989 . 990 In this Fig. 8 arrangement adjustment of the power driving frequency, as measured at meter‘ F, is eifected at the driving oscillator, while ad justment of the driving power current, as meas ured by meter A, is effected at the power ampli fier. Both of these adjustments are manual and 40 . 650 x 122 1. 35 . 651 0 173 1. 37 0 . 659 0 134 1. 38 :t 112 . 660 0 130 1. 44 + 71 73 . 661 . 666 z + 138 139 1. 50 l. 60 z :l: . 668 0 132 164 . 676 I 135 1. 64 81 . 679 -i- 168 2. 00 l. 60 :c+ z 156 203 133 . 685 . 705 . 708 1 0 0 114 72 68 2. 35 2. 66 4. 00 ‘i I+ I+ 157 . 728 I+ k70 5. 34 I+ 123 . 7 54 I - I+ O-No defect. x~Longitudinal crack. -I--Tran sverse rupture. The seventy-five ñred brass cartridge cases of 75 mm. size which are identified under the “case” column were examined for the presence of season - I . ' > In order to obtain an amplitude measurement for the vibrating case I5, switch 60 is first thrown to the upward position wherein its output circuit feeds into resistor 43 and the tuned-plate ampli ñer 55.» That amplifier is manually adjusted for the resonant case frequency, and under these conditions it functions to magnify the voltage -I 119 are made to meet the requirements earlier de scribed in connection with Fig. 1. Looking next at the frequency amplitude meas uring means of Fig. 8, these employ a tuned-plate amplifier 56, a cathode ray oscillograph 51, a beat frequency oscillator 58, a potential comparing de vice 59 and a transfer switch 60. z x+ I + 0 170 167 ' function as does the Fig. l source. z I + a: I+ 171 . 106 120 101 151 - Look first at the case driving power supply means of Fig. 8, it makes use of a power ampli ñer 53 and a driving oscillator 511. This oscillator fluctuations across resistor 43 which the strain 50 gauge S produces. ` ‘ - y So magnified, these fluctuations aretransmit ted to the cathode ray oscillograph 51 where they cause to be traced on the oscillograph screen (not shown) a visual showing of the case vibration 55 wave form and amplitude. In the system of Fig. 1 only the latter quantity can be indicated. The former occasionally is of value and can be pro videdby the Fig. 8 arrangement, The amplitude of the case vibrationwave hav 60 ing been noted on oscillograph 51, transfer switch 60 is shifted to the downward position where the amplifier 56 has transmitted thereto “compari son” oscillations from the beat frequency oscil lator 58. This oscillator is manually adjusted to the resonant frequency of the test case I5 and it causes oscillograph 57 to trace a wave of the same frequency as is present in case I5. ‘ cracks by our improved sonic test method. This examination gave the “sonic test coefñcients” oscillograph may be varied by device 59. Adjust which Table X lists. These eases were then fired at 12% excess pressure in a 75 mm. weapon hav ing a worn chamber. 53% of the cases split. None of the cases with a sonic test coeiü‘cient less ment from that device is now so made that the wave traced by oscillograph 57 has the same am plitude as did the wave there traced when ampli ñer 55 was connected with the strain gauge S. than 0.55 split. All but one of those with. the coefficient above 1.00 failed during firing. The amplitude of this wave as show by the Under these “matched wave” conditions the 75 voltage reading at indicator V of the comparing 2,403,999 12 ll device 59 is now noted. of the strain gauge S earlier described as. being mechanically attached to the case side wall and This reading corre sponds to that directly obtained by voltmeter V varying its resistance in step with the elonga in the organization of Fig. 1. tions and contractions oí the side wall metal. For purposes of computing “sonic test coeiì cients” the modiñed organization of Fig. 8k thus Cl While exceedingly satisfactory such a strain gauge is not, however, the only device which is the full equivalent of the basic organization lends itself to registration of amplitude measure shown in Fig. 1. ments. THE REGENERATIVE ORGANTzATroN or FIG. 9 During its vibration, a 75 min. case emits a It has been seen that Fig. 8’s driving. power sup ply facilities require that the frequency of the supplied driving power be adjusted at oscillator 54 to match the resonant frequency of the tested case I5. The need for such manual adjustment may be dispensed with through use of driving power equipment organized as shown in Fig. 9.~ There Fig'. S’s driving oscillator ‘54 is replaced by a phase shifting network 62. The input termi nal's of this network are directly connected with the output terminals of an amplifier 'I2 having 20 highly audible sound, and the intensity of this sound has been observed to vary in direct pro portion to the amplitude of the vibration. In the arrangement of Fig. l0 advantage of this fact is taken by the use oi a microphone 3G placed to receive the sound waves induced by the longi tudinal vibrating movements of the case. This microphone 65 may be of the conven- . tional carbon-granule type used in commercial telephones, or the condenser type used in radio broadcasting, in which case itr sets up resistance automatic volume control, while the output ter variations analogous to those produced by the minals of phase shifter 62 lead directly to the strain gauge S. This similarity makes possible input terminals of power ampliñer 53. a direct substitution in the electrical circuits of The power ampliiier 53 of. Fig. 9 corresponds Fig. l, for example, of the microphone 66 for to the similarly identiiied amplifier of Fig. 8 and 25 the strain gauge S and results in the apparatus constitutes the source of alternating current drive organization which Fig. l0 shows. ~ energy impressed upon transformer T and there The organization of Fig. 10 employs battery by transmitted to the driving winding I6 of mag ¿52 in the same manner as does the strain gauge net MÍ The strain gauge ampliner `33 of Fig. 9, S of the earlier Views. It is possible to eliminate in turn, corresponds in all respect to the simi 30 this battery by employing a dynamic type of larly marked ampliñer of Fig. 1 and indicates the microphone in the manner shown at 57 in Fig. 1l. strain gauge output at meter V. Such a device generates its own potential and In operation ofthe Fig. 9 system, each vibra for this reason is suitable for directly exciting tion of the tested case I5 varies the resistance `the ampliiier 38 without recourse to the earlier of strain gauge S, produces a corresponding shown battery 42 and resistor 43. change in the voltage across resistor 43, causes Fig. l2 shows a further arrangement for con an ampliiied measure of this change to appear at verting the case vibrations into amplitude pro the output terminals ofV ampliñer ‘I2 and hence portional changes in potential and supplying that at the input terminals of phase shifter 62. There potential to amplifier 33. In Fig. l2 use is made such displacement and timingV ís introduced as 40 of an electrical pick up device 63 analogous to proves most effective for exciting the power am pliñer 53. Since this excitation is in the form of pulsations which recur at the case’s resonant frequency, the cycles of case vibration once set up are self-propagating through a feed back or that used in phonographs. Such. a device 69 also generates its own voltage and hence is suit able for direct connecting to the input terminals regenerative action. Once, started, this action continues indeñnitely and it automatically adjusts the frequency of which vibrate. at above the audible sound range may also have their vibration amplitudes detected of amplifier 3.8. It should be pointed out that test specimens by the just described apparatus of Figs.A 10-11-12. the case driving voltage from amplifier 53 to an SUMMARY exact matching relation with the resonant fre 50 quency of the vibrated case I5. From the foregoing it will be seen that we have In order to start this regenerative vibration provided improved procedure by which the pres control, it is merely necessary to set up some ence of internal flaws in metal specimens may electrical or mechanical disturbance, such as clo be detected reliably, quickly andeasily; that we sure of a switch 64 in the supply circuit for the have made special provision for examining metal magnet’s direct current winding 24. Once the specimens which are cylindrical in shape; that sequence of regenerative actions above discussed, we have provided a method for testing the brass has been started, these actions `continue until cases of ñred artillery cartridges for the presence some break is made in the power supply or feed back circuits. The automatic volume control desired in am pliñer 'I2 is such that oscillation of the system in the manner described above is possible when the sonic test coeinicent of the cartridge case is sub stantially above the acceptance value; it further is such that the amplitude ofV oscillation is suc cessively greater for cases of lower sonic test co efficients until the maximum permissible power from amplifier 53 is attained for a case with a 60 of dangerous season cracks; that we have devel oped a reliable test for the susceptibility of fired and resized cartridge cases to splitting on subse quent rlring; and that we have perfected a test method by which it is possible prior to resizing to reject all cases which after being resized and reconditioned will‘fracture on firing. The testing yapparatus for '75 mm. cartridge cases which we have shown by way of illustra tion may with very slight modiñcation be adapted to the testing of artillery cartridge cases of other70 sizes and forms. Such adaptation consists in as 0.30. selecting mechanical dimensions of the case sup THE ALTERNATE AMPLITUDE INDICATORS or' porting frame and the driving magnet M to ac F1os. 10-11-12 commodate the particular size of cartridge case In all of the thus far shown arrangements for desired to be tested, and in choosing electrical indicating vibration amplitude', use has been made characteristics of the driving power supply and test coeñicient below the acceptance value, such 2,403,999> 13 vibration measuring facilities which are appro priate for the selected case size. Our improved sonic testing method further lends itself to use with metal specimens other than cartridge cases and may with comparable success be applied to the detection of internal ilaws in cylindrical specimens which are open at both ends. Nor are cylindrical specimens the only type 14 men exhibits a damping capacity substantially different from that of a sound-brass specimen, and measuring the damping capacity of said test ed specimen at said high stress producing ampli tude for purposes of classifying the brass thereof either as sound or as defective to an observed degree. 3. In a method of testing metal for the pres ence of defects, the steps which comprise induc which can be tested, for upon the making of 10 ing in a specimen of said metal mechanical forces modiñcations immediately apparent to those skilled in the art, specimens of other forms and shapes may also be subjected to flaw determina which repeatedly reverse themselves and by which said specimen is excited to longitudinal vibra tion, adjusting the frequency oi' this vibration to the specimen’s natural frequency of resonance, tion tests by the here disclosed method of excit ing the specimen to resonant frequency vibra 15 measuring the amplitude of this resonant fre quency vibration to indicate the resulting peak tion at amplitudes sufliciently intense to reñect stresses which are set up in the specimen metal, specimen iiaws, measuring the “sonic test co intensifying this resonant frequency vibration’s efficient” under these conditions and comparing amplitude until said indicated peak stresses at the observed value with standards established for specimens of the same form and material. 20 tain a given high value at which a defective metal specimen exhibits a damping capacity Our inventive improvements are therefore eX sharply differing from that of a sound-metal tensive in their adaption and are not to be re specimen,v and measuring the energy that is re stricted to the speciñc form here disclosed by way of illustration. We claim: 1. In a method of testing metal for the pres ence of defects, the steps which' comp-rise excit ing a specimen of said metal to longitudinal vi quired to produce said given-stress-valve ampli 25 tude of resonant vibration whereby to indicate the relative damping capacity of the specimen and therefrom to determine whether the metal of said bration, adjusting the frequency of this longitu specimen is sound or is defective. 4. In a method of testing metal for the pres specimen metal, intensifying this resonant fre longitudinal vibration, adjusting the frequency of men of cartridge case brass for the presence of required to produce the intensified amplitude vibration by which said given high value of meas ured peak stress is yielded, and dividing said given-value stress measurement into said electri dinal vibration to the specimen’s natural fre 30 ence of defects, the steps which comprise elec trically inducing in a specimen of said metal me quency of resonance, measuring the amplitude of chanical forces which repeatedly reverse them this resonant frequency vibration to indicate the selves and by which said specimen is excited to resulting peak stresses which are set up in the quency vibration’s amplitude until said indicated 35 this vibration to the specimen’s natural frequency of resonance, measuring the amplitude of this peak stresses attain a predeterminedly high value resonant frequency vibration and the resulting at which a defective-metal specimen exhibits a peak stresses which are set up in the specimen damping capacity sharply differing from that of a metal, intensifying this resonant frequency vibra sound-metal specimen, and measuring said damp ing capacity at the so intensiñed amplitude of 40 tion’s amplitude until said measured peak stresses attain a given high value at which a defective vibration whereby to determine whether the metal specimen exhibits a damping capacity sub metal of said tested _specimen is sound or is defec stantially differing from that of a sound-metal tive. specimen, measuring the electrical energy that is 2. In a method of testing a cylindrical speci defects, the steps which comprise exciting said specimen to longitudinal vibration, adjusting the frequency of this longitudinal vibration to the specimen’s natural frequency of resonance, meas uring the amplitude of this resonant frequency vibration to indicate the resulting peak stresses which are set up in the specimen brass, intensify ing this resonant frequency vibration’s amplitude until said indicated peak stresses attain high val ues typiñed by several thousand pounds per square inch and at which a defective-brass speci -cal energy measurement whereby to indicate the relative damping capacity of the specimen for purposes of classifying the metal thereof either as sound or as a defective to an observed degree. THOMAS A. READ. HERBERT I. FUSFEID. SUMNER W. KITCHEN».