'26 PHiLIPS ',fECHNICAL' REVIEW Vol.' 5, No. 1

THE INVESTIGATION OF ;RAPIDLY CHANGINGMECHANICAL STRESSES WITHTHE CATHODE RAY OSCILLOGRAPH

by S. L. de BRUIN.

By means of a specially made resistance strip which is firmly fastened to a ~art of thestructure to be investigated, the changes in shape caused in the structure by rapidly chang-, mechanical stresses are converted into electrical voltage variations. These latter can~. . hlthen be made visible by means of a cathode ray oscillograph. The very sensitive, porta ecathode ray oscillograph type GM 3 152 can be used in this way as a quick-reaction in-dicator in the study of certain mechanical problems. -

Mechanical stresses.. ~hich occur in materialsand parts of mechanical structures can be investiga-ted by measuring the accompanyll;tg changes inshape. These changes are usually only extremelysmall, but they may nevertheless be investigatedwith special instruments constructed for thatpurpose. Such instruments, however, generally havetoo much inertia to allow the measurement ofsmall, rapidly changing mechanical deforma-tions, such for example as occur upon vibration insteel girder structures, high speed turbine shafts,aeroplane .wings, etc. Moreover, the parts of thestructure which must be tested are often difficultto reach so that mechanical instruments can not, ,be brought to bear. --A solution of this difficulty is offered in the pos-

sibility of replacing the mechanical, measurementby an electrical one in which one may take imme-diate advantage of the already highly developedelectrical measuring technique. It is for examplepossible' in this way to use a cathode ray oscillo-graph 1) as a very sensitive indicator with almostno inertia. _

For this purpose a simple method has been worked.out whereby .the mechanica] deformation in astructural element can be converted into variationsin ·voltage. A specially constructed resistance stripwhose electrical resistance varies practically linearlywith its length is firmly attached to the outer surfaceof the part of some mechanical structure to beinvestigated. By an appropriate electrical circuitthe changes in resistance are converted into changesin voltage 'which can he made visible directly on'the screen of a cathode ray oscillograph. In orderto obtàin reproducible results, it is very importantto take 'care that the resistance strip is attached asfirmly as possible to the surface of the structuralelement being tested, so that one may be sure thatthe strip in its entirety undergoes the rapidly alter-nating expansions and contractions.1) For instance the cathode ray oscillograph GM 3 152

which is described in Philips technoRev. ~, 198, 1?39 ~ndthe employment of which in meehaaical engmeeringis discussed in Ingenieur 54, W 29, 5 May 1939.

621.317.755: 620.178.5

The resistance strip

The measuring strip consists of a piece of flexibleinsulation material (fig. 1), upon which a line isdrawn with a suspension of powdered charcoal.

:I425fi

AgIc AgIp

Fig. 1. Sketch of a resistance ~trip made ?f fle;cihleinsulationmaterial (p) and provided with conductmg silver ends (A_g).C is the carbon line whose resistance changes upon stretchingor contraction.

The electrical resistance of this carbon line changesas . th~ strip is stretched or shortened, becauseof the fact that the mutual contact of the carbonparticles in it becomes poorer or better as the casemay .he. Just as in the case of the carbon micro-phone, where use is made of the fact that the resist-.ance of a mass of carbon powder depends uponthe pressure exerted upon it, we can employ thisphenomenon in order to measure the deformationsto which a mechanical construction is subjectedat different conditions of loading. The great sen-sitivity of this method of measurement may b,econvincingly demonstrated by a very simple test.If the measuring strip is included in a sufficientlyaccurate resistance meter, a measuring bridge forexample, and if the strip is then bent with the carbonline on the convex side of the resulting curve, sothat it is stretched, the resistance is actually foundto have increased compared with that in the non-deformed state. If the strip is bent so that the·carbon line is on the inner, concave side, the resist-ance becomes smaller.The resistance strip is prepared in th~ following

way 2). On a strip of insulation material 0.3 mmthick, 50 mm long and 8 mm wide (see fig. I),between the ends, which have been rendered con-ducting by means of a layer of silver, a line isdrawn with a drawing pen, using very finelydivided. carbon powder suspended in a liquid

2) The necessary components bear the type number GM4 470.

JANUARY 1940 THE INVESTIGATION OF RAP1DEN CHANGING MECHANICAL STRESSES 27

binder. The binder IS dried by heating. A stripis obtained in this way with a resistance of forinstance about 10000 ohms.

Fig. 2. Tie rod to which a resistance strip is fastened.

The resistance strip can easily be cemented tothe surface of the object to be examined, for in-stance a tie rod (jig. 2) by means of celluloidlacquer. According as the temperature of the ob-ject tested is higher or lower the strip glued to itmust be dried under pressure for 1/2 to 1 hour. Thecopper wires, by means of which the strip can beincluded in an electrical circuit, are then fastened tothe silvered ends of the resistance strip by means of acement containing highly conducting metal. Finally,to avoid the absorption of moisture from the at-mosphere, it is advisable to cover the whole resis-tance strip with a layer of celluloid lacquer beforethe measurements are begun.

Conneetions and sensitivity of the measuringarrangement

The resistance strip, prepared as described inthe preceding section, is included in an electriccircuit as indicated in the diagram of jig. 3. Witha sufficiently high series resistance Rv, the voltagevariations between the terminals of the measuringstrip Rw, which is in parallel with the built-in am-plifier of the cathode ray oscillograph giving anamplification of about 1 500 times, are in the firstinstance practically proportional to the resistancevariations of the measuring strip. When the defor-mations are not too great, 0.05 per cent at the most,

Gt13152~o~0000000GO 00 0

34254

Fig. 3. Measuring arrangement with the cathode ray oscillo-graph 3 152. Rw is the resistance strip, R; a series resist-ance, B a battery and C a condenser which serves to preventdirect current from passing through the oscillograph.

for instance, these resistance variations are alsoapproximately proportional to the deformations,and the latter in turn, according to Hook e 's law,are proportional to the mechanical stresses whichcause them in the material tested. We thereforeobtain as result a deviation on the fluorescent screenof the cathode ray tube which is practically pro-portional to the mechanical tensile stress to be mcas-ured.

Fig.4. Oscillogram of the variation with time of the mechan-ical tension in the outermost layers of a rod clamped at oneend which is struck at a given moment and then allowed tocome to rest unhindered.

Thanks to this proportionality it is possibleto judge the vibration phenomenon to be investiga-ted directly from the fluorescent screen of theoscillograph. As an example an oscillogram isreproduced in fig. 4. which shows the change oftension with time for a rod clamped at one endwhich is set vibrating at a given moment and thenallowed to return to rest without interference.If it is desired to measure the mechanical defor-

mations in absolute values, the relation betweenthe changc in resistance and the ehange in lengthof the measuring strip must be established by cali-bration. For this purpose a resistance strip is gluedto a tie rod for instance (fig. 2) and included in a

c

Fig. 5. Circuit of an electrical measuring bridge with whichthe resistance strip can be calibrated. T tie rod, S resistancestrip, R resistance box, T fixed comparison resistances, J pen-tode amplifier with a tuuing indicator.

28 PHILlPS TECHNICAL REVIEW Vol. 5, No. 1

bridge circuit 3) as in fig. 5. The tie rod can then bestretched in a machine for testing tensile strength(fig. 6).

Fig. 6. Measuring arrangement for the calibration with thehelp of a machine for testing tensile strength.

Between the specific change in resistance of themeasuring strip and the specific change in length(! of the tie rod, measured with the above-mentionedmachine, a satisfactorily reproducible 4) relationcan be found of a character such as that shownin curves I, IJ and III of jig. 7 which were deter-mined with several test objects. Itmay be seen fromthis that for deformations (! of not more than0.03 per cent, the relation between the variationof electrical resistance À and mechanical tensionIS in fact practically linear. In this region the

3) For example the "Philoscope", type GM 4 140 which wasdiscussed in Philips techno Rev. 2, 270, 1937.

4) For reproducibility it is necessary, as is usual in stretchingtests, that the rod should have been prestretched.

quotient Àj (! has a value of about 15. With largerdeformations thc electrical resistance is found toincrease more rapidly than proportional to thestretch.

/IV 1/is / /'

/' /'

P /' //' /'

11~

r/ /

/' /' //'

//' /'

/' ,/1.0 /' »:

/ /'/ /' /

/'/ /' /'/ »: .-'

// /// »: /'V"..,.... .....-/

/.-' ,/

0,5 .......-:::~./' ..... ..,....---'",A ..... / _ ...............

...... ::::- ....... _;,..-.#-:;.-:::.:.----_À.

0'?~':s;.~--2 3 4 5 6 7 8 9 x10-4

0 iooo 2000 _kg 300034255

Fig. 7. Relation hetween t.he specific resistance variation(i of the measuring strip and the load on the tie rod in kg, orthe specific stretch Ä of this tie rod.

For the calibration of the measuring stripit is by no means necessary that one use a machinefor testing tensile strength, it is only necessaryto be able in some way or other to cause an ac-curately measurable change in length of the strip.In order to test a strip it may for instance he gluedto a test rod (fig. 8) which is supported at twopoints at a distance 2 I from each other. The rodis then bent between its points of support into acircular form over a distance d. The changes inlength of the upper and lower surfaces of the rodthen follow from the formulae for the circle. Inthis way the specific change in length of the rodcan be determined and the relation between thatand the specific change in resistance of the strip,measured in bridge connection, can be found.

.34291

Fig. 8. Circularly bent test rod, supported at two points at adistance apart of 21 and with a bend of d.