A MACHINE FOR BEND TESTS - Research | Philips Bound... · A MACHINE FOR BEND TESTS ... through 180° by the machine shown in fig.1 ... The jaws of the vice are moved toward each other.

246 PHILIPS TECHNICAL REVIEW VOL. 17, No. 9

collector of N-germanium - the theory is complete-ly analogous. In principle, the theory can be deve-loped without mention of holes and electrons, bydealing simply with minority and maj ority chargecarriers: the theory then becomes identical for bothtypes of transistor. Practical differences in thebehaviour of the two types are caused by the factthat the diffusion constant for electrons is abouttwice that for holes.

Summary. The action of a junction transistor is based on theproperties of PoN junctions in a germanium crystal. Such atransistor is built up of a P-1V junction from "emitter" to "base"

followed at a very small distance by an N-P junction from"base" to "collector". The action is explained, starting fromthe fact that the current at a junction is determined by thediffusion currents of the minority charge carriers which caneasily be calculated at the boundary planes of such ajunction. Itis pointed out that the voltage applied across a junction deter-mines the concentrations of the minority charge carriers in theboundary planes. At one of the junctions this concentration isfixed at zero; a small change of voltage across the other junc-tion then results in a strong variation of the concentrationgradient, and thus of the diffusion current of the minoritycharge carriers, between the two junctions.

Emitter and base efficiency are discussed and calculated forthe one-dimensional case; this is also done for the currentarnplification factors for circuits with common base and comm-mon emitter. The difference in the input and output imped-ances for these two circuits is explained and thc characteristicswhich follow from the theory, are deduced.

A MACHINE FOR BEND TESTS

A requirement frequently imposed upon a metalis that a specimen of certain form should not crackor break when bent through a prescribed anglearound a mandrel of prescribed diameter. Theresult of a bend test, however, depends upon themanner in which the test is carried out, and this isnot always specified. The consequence may he adifference of opinion between supplier and customeras to whether the metal satisfies or falls short of thespecification.

With a view to overcoming such difficulties theMaterials Research Group of the Research Labera-tory in Eindhoven has developed a special bendingmachine fOTtesting sheet and strip metal (fig. 1).By means of a single hand-operation the test speci-men can be bent in this machine quite smoothly andwith the correct radius through any angle betweeno and 180°. The design is shown in fig. 2. The testspecimen (4) is held between the clamping piece (1)and the mandrel in the form of an interchangeableradius plate (3), which is accurately rounded on thetop edge. The upper face of the clamping piece isshaped in such a way as to SUppOl'tthe specimen asclose as possible to the bending point, thus pl'event-ing unwanted deformation. The bend is effected bythe thrust block (7), which is moved by the level'(9). Three rollers (6) mounted in the thrust blockensure that, when the lever is pushed over, the blockslides along the specimen almost without frictionso as to exert practically no tensile load upon it.This is important, for if there were any appreciabletensile load on the specimen it would not be sub-jected to a pure bend and premature cracking mightoccur. Allowance can be made for the thickness ofthe specimen by adjusting with wing-nut 8 the

620.177.3

posltIOn of the thrust block in its groove on thelever. The lever 9 turns about an axis whichcoincides with the centre of curvature of theradius plate; this is essential to the correct function-ing of the machine. The interchangeable radius

Fig. 1. The newly developed machine for testing sheet or stripmetal for its bending properties. The test specimen is insertedbetween rollers and a radius plate. At the right can be seen twoof the interchangable radius plates over which the specimenis bent.

plates are ground in such a way that the centre ofcurvature always lies at the same position in themachine. A support bar (2) takes up the horizontalcomponent of the force exerted upon the radiusplate during operation and thus prevents it frombending under the strain. When the angle of bendexceeds 90° this component reverses direction andthe clamping piece (1) then takes upthe horizontal

MARCH 1956 MACHINE FOR BEND TESTS 247

force. The support bar is then no longer necessaryand, indeed, it forms an obstruction to furtherbending. For this reason it is pressed down againsttwo springs (12) by the cams (5) on the end of thelever (9), which bear on both ends of the pin (11)when the angle of bend becomes greater than 90°(fig. 2b). In this way the bend test can be continuedunhindered up to 180°. The angle of bend can heread from one of the cams (5) which is calibrated indegrees. Even at an angle as small as 60° to 70°,the outside of the bend becomes visible with thespecimen still in position in the machine. It can thusreadily be perceived at what angle of bend crackformation begins.

Fig.3a shows the cross-section of a specimen ofchrome steel with a tensile strength of approximat-ely 65 kg/mm2 and a thickness of 1.9 mm, whichhas been bent through 180° over a diameter of 0.95mm. The diameter of the radius plate in this casewas therefore half the thickness of the specimen. Itcan be seen that, even under such very adverseconditions for a bend test as these, the specimennevertheless follows the radius plate quite faithfully.Fig. 3b shows that under more favourable conditions(mild steel with a tensile strength of 40 kg/mm2 anda thickness of 1.0 mm, bent over a diameter of 1.20mm) the inner profile follows almost exactly thecurvature of the radius plate.As stated, the machine described serves for testing

sheet or strip metal, but it is also possible to designa machine on the same principles for testing metalin wire or bar form.

baFig. 3. a) Section of a test specimen of chrome steel, bentthrough 180° by the machine shown in fig. 1 (tensile strength ofspecimen approx. 65 kg/mm''; specimen thickness d = 1.9 mm;diameter of radius plate D = 0.95 mm; thus Dld = 0.5).b) Section of a test specimen of mild steel (tensile strengthapprox. 1·0kg/mm"; d = 1.0 mm; D = 1.2 mm, and Djd = 1.2).

In order to make clear why this rather elaboratemachine has been developed for performing such asimple test, some comment sare called for on the rath erprimitive methods often employed for the bend test.

98

5- _

4_

3-

I

III'

I10

Fig. 2. Structural design: 1 clamping piece; 2 support bar; 3 radius plate; 4 test specimen;5 earn, which depresses support bar; 6 rollers, mounted in thrust block 7, which is fixedby wing-nut 8 to lever 9; the excentric 10 serves to hold clamping piece against specimen;11 pin by which the support bar is depressed against spring 12. a) Machine ready foroperation after insertion of test specimen; b) after bending through 145°.

24·8 PHILlPS TECHNICAL REVIEW VOL. 17, No. 9

A method in common use is to secure the testspecimen together with a mandrel in a vice (fig.4a)and to hend the specimen over the mandrel by handor with ~ hammer. If the specimen is to follow themandrel closely, the hending force must he appliedclose to the mandrel (fig. 4b. and c). The use of ahammer is to he deprecated, as it subjects the spec-imen to shock-loading of unknown magnitude,which may result in premature cracking or hreaking.What is more, hammering close to the ~andrel mayproduce undesirahle 'work hardening of the regionssuhj ected to impact. 'If an angle of hend greater than 90° is specified,

it is necessary, when using this simple method, toremove the test specimen at a 90° hend from the

p

~s

Fig. 4.. Specimen and mandrel clamped together in a vice for aprimitive bend test (a). If the bending force P is applied nearto the mandrel, the bend radius follows the curvature of themandrel (b); if P is applied at some distance from the mandrel,the bent radius willbe too large (c). .

vice, to secure it again in a different way (fig. Sa)and then to continue the hend test by tight-ening the vice itself. It is now almost impos-sible to hend the specimen closely around themandrel, hecause the hending forces are neces-sarily applied at some distance from the mandrel,and the mandrel always slides hack a little as theangle of hend increases. The consequence is asharper hend than specified (fig. Sb). The secondohjection might he overcome hy supporting thespecimen directly heneath the mandrel (fig. Se),hut then the metal is no longer suhjected to a purehending load. Moreover, the force exerted by thesupport may be-so considerahle as to cause unwanteddeformation of the specimen at the hend radius.Nor is it possible with this methöd=to determine thepoint at which cracks hegin to form.

£Fig. 5. Specimenfrom fig.4 arranged in vice for bending throughangles greater than 90°. The jaws of the vice are movedtoward each other. a) As in fig.4c, unwanted bending occurs;the mandrel also becomes displaced (b) causing too sharpa bend. Supporting the mandrel (c) avoids displacement butthe specimen is then no longer subjected to a pure bending.

If an angle of hend of 180° is specified, a methodfrequently used is to press the test specimen into adie by means of a punch with a curvature of theprescribed radius. The aperture of the die is thenequal to the thickness of the punch plus twice thethickness of the specimen (fig. 6). In this case anunwanted load is exerted upon the specimen hy thefriction occurring between it and the die. This

Fig. 6. Bend test employing a die. and a radius plate in theform of a punch. .

friction can he limited by rounding-off the edges ofthe die and possibly by providing the aperture witha taper, hut these measures cannot he taken farwithout entailing the drawhacks illustrated hyfig, Sa and b. The method has no practical value ifthe punch is thin in relation to the test specimen orif punch and specimen are hoth thin. In the firstcase the strains set up in the punch are so great asto cause it to collapse. In the second case thespecimen suffers deformation as a result of theshearing forces arising.

Use is sometimes made of special tools incorporat-ing several of the elements of the machine describedhere. The test specimen is secured in a clamp, oneside of which is given the curvature over which thespecimen is to he hent, that is to say one clampingpiece acts as the mandrel (fig. 7). The metal is hent

Fig. 7. Bend tcst employingspecial tools. The axis of rotationof thrust block 1 coincideswith the centre of curvature M ofthe curved member over which the specimen is bent.

over this surface hy means of a thrust hlock attachedto an' arm, whose axis of rotation coincides with thecentre of curvature of the mandrel. With this arran-gement, however, the specimen cannot he hentthrough more than ahout 110° (depending upon therequisite strength of the mandrel). Moreover' atensile load is exerted by the thrust hlock as itmoves over the face of the specimen. These draw-hacks have heen substantially eliminated in theapparatus described here.

M. J. W. GEENEN.