bh + \ 3.2.7 Semiconductor Strain Gauges RArwt fh tupe Silicon is the basic material for making semiconductor sttain gauge. Usually silicon is doped with boron to make p-type strain gauge. It will be »-type strain gauge if it is doped with arsenic. Resistance of p- type gauges increases with the applied tensile strain and that of n-type ga Resistivity of a semiconductor material is given by et 1 Ake Untrtata Unt Win. where g is the charge of electrons; » is the number of charge cartiers that depends on doping; 1 is the mobility of charge carriers. “Theimparity concentration typically lies berween 10'° and 10” atoms/cm’. The resistivity ofp-type silicon 10°? atoms/cm’ is 500 jum which is much higher than that of copper. The strain gauge jin. This change of resistivity is called che piezo-resistve effec. ied strain, thus causing large change in resistivicy.je is shown i active gauge but it will not be subjected co The bridge unbalance voleage is noi 3.3, LOAD CELL °25 ‘Substituting for initially balanced bridge we get RyRy = RR, AR AR , AR, Pe RRR umn type load cell is shown in Fig. 3.4. Load can be either tensile or compressive. The 4 is on the left side of the column and the axial gauge 3 is on the back face of it. The “1 and 3 are subjected to axial strain and gauges 2 and 4 are subjected to transverse strain. Jn gauges are active gauges. We can measure the load P wich one strain gauge; however, will ive higher bridge output and it will make the load cell insensitive co‘output voltage ¢, can be 1t C is equal3.3 LOAD CELL 927, with high range. The voltage ratio at maximum load (¢/E,)nix foF the column type load eell Wheatstone bridge combination will be given by & Ap 4+) Ban) 28 Load cells are usually fabricated from steel with following specifications: 0684 x 10'! Nim? 3 516 X 10° Nim? 3.5 Ryand Ry are slightly differene for temperature1D TORQUE Mes d at one end an, ithe bottom suri tively. Moreox 37. Now a | auges | and3.4 CANTILEVER BEAM TYPE LOAD CELL shu SGA am wstalld ow the Mnial dusction, = Crlisrn Lap ina, The load P produces a moment of Px, where x is the distance from the mid-point of axial gauges installed to the point of application of load P. The strain developed at the gauges is given by OPx. @ 2-262 ee 2 iipkegiiols tS spe where wis the width of the cross-section of the beam; h is the height of the cross-section of the bears. Therefore the change of resistances in the gauges will be given by AR 2 ARs _ ARs ARg SAPs, R RR Ry Ewh? GAO) Substituting Eq. (3.10) in the expression for unbalanced Wheatstone bridge, that is, Eq. (3.7), the ‘output voltage can be written as oi A See3.5 INTELLIGENT LOAD CELLS 231 IER, is the unknown resistance Ry then Ry Ry eR; Rye? Now if R, is the strain gauge, the ambient temperature compensation can be made by replacing Rz ‘by a dummy gauge. While strain gauge {R,) will be subjected to stress, dummy gauge will not be; ‘however, both will have at the same temperature all the time. ‘A.2 Unbalanced Wheatstone Bridge Then: nheatstone bridge is unbalanced, then the nodes ‘b’ and ‘d’ are not at the same potential. Theee i) ee See See oes See eee 6143 Sources of Error Bonded strain gauges have assumed great importance for application in situations involving both static and dynamic measurements of strain. The size of the gauge permits point-to-point investigation of strains occurring on surfaces of structures. Metallic wire strain gauges have Proved their long term stability. Their application potential has considerably increased with the availability of a range of good quality cements for bonding the strain gauges to the surfaces of Structures, The gauge factors of the wire- and foil-type gauges are seen to be sufficiently Constant for some of the materials and are normally furnished by manufacturers after Considerable testing and experimentation. To enhance the accuracy and reliability in the Measurements of strain, sources of error other than those due to mounting are seen to be due toi ee ae ee ‘The values of strain normally encountered in engineering measurements + 10 to 10°. Taking the gauge factor of 2, and the gauge resistance as 120 0, the « sesistance of the gauge due to a strain value of 10 is as low as 0.24 x 10° 9. Its ‘that there should not be any other interfering input that affects the resistance o! clement. It may be recalled that the resistivity of the material changes with the tem?’ ‘the clement and hence its resistance. The strain gauge material and the material of the s ‘on which it is mounted have their coefficients of linear expansion, which may not :l the same. Thus temperature of the surface and hence of the strain-gauge &le'"* considerable influence on the resistance of the element due to the above-mentions | Sometimes, changes due to temperature effects may °™ Veeen For accurate measurement of strain, it is essen"! ‘ ‘compensated or are taken care of by suitable corres". ‘is bonded to the surface of a structure, and © Voltage developed will be treated as if that contribute to the development °! “"——Selrical Transducer 293 in gauge ee eure Sie they are cemented together, throughout the te etetlnate de iny increase in resistance of the gauge can be considered Spe cirdennotn nies to the strain gauge. In practice, however iia to st ep stots under test, which may be large or small in size, at one eee ' e, throughout eee = Syed, as in earlier cases, in having two identical tion for such seen wai an active gauge, in a unity-ratio eee oe a dummy gaug: raulge is cemented to the surface of an iden n in Figure 6.24(a). The dummy g and the body on which it is cemented294 Transducers and Instrumentation ee any mechanical stress. When both structures are ture, kept free from struct F a le . the bridge is balanced. An equal r f and at the same temperature, the qual rise in te Tee eee ‘structures raises equally the resistance values of both the active anc “tres: and the bridge still femains balanced. When the test structure is strained. resis Auge, up and the unbalance voltage V, is proportional to the strain, of thes So ace ean is far less than unity, V, is given by Ep _EGe coe 4 In case the test surface is wide enough to admit another strain gauge, by the s active gauge, as shown in Figure 6.24(b), temperature compensation is as tase provided the second gauge is mounted so as to be sensitive to strain occur the other strain. If the strain in the transverse direction is zero, the second gat resistance effected by temperature only and the effect of temperature 1s the sar gauges, as they are proximate to each other. The above relationship, as given in Eq. (6 ‘unbalance voltage, still holds good. In most cases, the transverse strain is of opposite sign and is equal to we. where i Poisson’s ratio of the material of the structure. Accordingly, unbalance voltage will be modifi as EGe eee ama lt 1) tes sensitivity can be doubled while having temperature compensation by usit} seve Aug Pountadon the test siscture but located in such a way thatthe two gates ees sch but in opposite sense. Such an arrangement is possible only with pesos nis aa which may have the strain gauges mounted on opposite sides surfaces ofthe cantilever ‘oe 6.24(c), However, it is essential that both the top and bottl eats cantilevers and, always at the same temperature, and such a condition is possi of the oes that are physically small in size. Any difference 1” ed pant in measurement. The unbalance voltage _ will be twice as much as the voltage sive? basi bridge, while the other tw arms. It may be noted, in this context, that ° enable the location of the strain gauges: |" the © 18 at the same temperature, the arrangement © the resistors for the ratio arms. AS th Investigations oflaterials for resistance strain gauges are not s used to make strain gauges. The table also Composition (Cu 55%, Ni 45%: Ni 80%, Cr.20% 1Ni 75%, Cr 20%, others 5% ‘Ni 36%, Cr 8%. Mn-Si- Mo 4% and rest iron, Pure platinum, Pt100 Platinum-tungsten ___Alloyed with small tungsten quantity _ important factor and should be given due consideration Advance has a f value of 0.03 x 107/°C, for nichrome it is as high as 0.1 10°/°C, platinum has a Value 40 ‘times more than that of advance, platinum alloyed with tungsten has a value somewhat lower lnainiyieasesa ‘Advance Nichrome Karma Isoelastic Wire strain gauiges are of two different type: a the unbonded type the resistance wire is stretched between two frames as shown in Fig, 23, ‘The rigid pins on the two frames are insulated. The rela- tive motion between the two frames produces a strain in the stretched wire which in tum causes a change in its resistance value. This change in resistance value is actually ‘measured to measure the strain or the force causing it ° foil ceme paper or thie! 0.0075 em thick. The sketches of such gauges are shown in Fig. 2.4. The . it consists it 7 the resin at a high temperature. The measunins leads are either soldered or welded to the gauge wire. The bonded strain’ gauige with the or plastic, bakelite ss WEAVE Cit seer be used as backing material for which id be used as the epoxy or ceramic cement wou! io atlas oe eeSnail stra gauge. vl TT “Temperature fluctuation is a major cause of error inmeasurement by strain gauge. Tempo & ea e defi le (O alters the gauge resistance, and (ii) causes differential expansion between the gauge ms aes ‘material or the test piece causing variation in the measurable strain. In general it is a a roa rs Out the correction for the temperature effect. The usual method of compensation 1S toil esl "uges, one on the test specimen and the other on unstrained material of the same compo on oe *dummy, Both these materials ure kept close together for attaining the same tempetukS ETT eo, temperature variation would equally affect both, thereby compensating each other. omg ‘an, however, be made as shown below.is necessary gauge 1Displacement, Force, Velocity, Acceleration and Torque 49 helps minimize these effects. The coating materials are chosen from waterproof compounds such as microcrystalline silicon wax, bitumen, rubber compounds etc. + Thermoelectric effect arising at the leads sometimes causes error but this is very small. + Magnetostrictive effect in strain gauges of ferrormagnetic materials is observed when the gauge is in use. Itproduces a disturbing voltage as large as 2 mV. The gauge, if properly conditioned by repeated straining and passing current while on strain, may be free from this effect. ‘The gauge must be stable in operation. Instability occurs due to. zero drift and hysteresis effects, when on load. Adopting proper fixing techniques with adequate compensation, the stability can be increased to a large extent. he eee ene cecictamre ctrain oases. foil type. bonded gauge; vacuum deposited thin metal-| Unbonded Strain Gauges wire strain gauges are available in two forms: unbonded and the bonded ones. aiejed type, a fine wire of 251 diameter is kept stretched between two insulating posts ‘on two different platforms that can move relative to each other under external ye euch a system forms the simplest assembly of Figure 6.21(a), actual transducer mploying four saphire posts and holding four equal lengths of tungsten-platinum wire er, are common as the four wires can form a Wheatstone bridge circuit providing » compensation. The four posts, as shown in Figure 6.21(b), are mounted on a star- ” When the centre of the star spring is subjected to the force under measurement, Diaphragm Force rod ‘Strain gauge windings (1 & 2)gauge elements on opposite Sides voltage of the bridge circu; it is Prop is for particular applications ‘make them suitable for application, bonded stain ‘The resistive element may be in the form of type is the wire-resistance SaUgE, With the to form a flat grid in one plane The Closely spaced as to minimize the lengh on t0 that along the longitudina! he strain gauge sufficiently insensitive vity” becomes very lo BS shown in Figure 6.22(b), may ‘Alternatively, the wire may be in a helix on a hollow cylindrical 6.22(c). Both of these wrap- of thickness from 10, 10 ‘of circular section. The fleets of and are nyPassive Ectical Trrscicars 294 process of manufacturin, : them has by rougl $ are also developed | oad down the gauge length to 3 mm DY va op Suum deposition of films on thin insulating base factors slightly higher than those of the above resistance elements are ceme: resin, to two sheets of flexible insulating ented, the combination becomes an inte ‘The bonded strain gauges, thus construc! data concerning gauge factors. pr accurate measurement of the surface strain of any structure, it is necessary that the in gauge become integral with the test surface, For each type of gauge and test er cementing agents should be used. The area where the strain gauge is to be is first polished by emery cloth or polishing wheel and then all traces of greasy film ed by cotton swabs dipped in acetone. Similarly, one of the surfaces of the strain ; also cleaned for mounting on the structure. For mounting paper-backed gauges, cement is generously applied to the surface of the structure as well as the gauge then the gauge is mounted on the surface, such that the gauge length (axis) is along in which the strain has to be measured. Excess cement is squeezed out by = i i iti til it is fully dried up. After gluing the jon of gentle pressure. It is held in position un din potion tnt packed gauges with phenolic cement (araldte) they are helt tt PO ih and then slowly heated up to about 170°C. Foil-type and cement, after test surface by neg 2 eer ice or 3 é i with n with trichlorethylene. It may . nt to become integral ve nting thus enables the strain gauge on test surface and the gauge eee that the gap between INE Tal surface of the con with the depth of the i the above types. led securely by nitrocellulose cement or material, one on each side of the element. ‘gral component and the element cannot be ted, are available from manufacturers along £ test S fin gee available as tans! backing to the———— Yacuum deposition and sputter deposition thin film gauges are produced seween the elastic element and the gauge is not necessary for bonding. In suitable elasti ressure measureme! © metal element which can be adapted for strain generatior nt, is placed in a vacuum chamber with a suitable dictect fewer vapour point than the metal, With application of requisite amoun material yapourizes and then condenses and finally forms a thin layer on thé template of a suitable shape is now placed over it and the evaporation-d Thus, the gauge is formed over the insulator t repeated with the gauge material ; In the sputtering-deposition process, t he strain member. In the second step, without u over the entire substrate and the gauge pattern is defined by 1d photosensitive masking all unmasked tep is nearly the same layer on t ng a temp material is sputtered imaging techniques an sputter-etching is used to remove als from outside the cha ers inside the vacuum chamber insulating carrier fi reinforced phenolic having a thickness of about 0.002 cm polyimide and gla fnuge materials with corresponding gauge factors and temperature coefficient of resistance along- with the resistivity values. “Materia! < nominal Gauge Thermal coefficient Nominal resistivity composition (%) factor _of resistance (%I°C) (2 cm) ‘Constantan, Ni 45, Cu 55 21-22 2x 10° 0.45-0.48 ‘Advance, Ferry Karme Ni 74, Cr 20, Fe 3 21 2x 10° 1.25 cus Nichrome V Ni 80, Cr 20 10? 1,00 ‘soetastic Ni 36, Cr 8,-Fe 52, 1.75 x 107 1.05 Mn-Si-Mo 4 RW soy Pt 92, WB 36-45 24 x 10% 0.82 ve Ni 100 12 0.68 0.65 Mangenin Cu 64,Mn 12,Ni4 0,3-0.48 2x 10° = Patron Pt 100 ot te tied y $ aja iment '9 the environment, specifically the metal-paper and metal-disleetric | 4Specially made for bonding of adhesives Max. strain at A room temp. (%) | purpose adhesive, Epoxy €an be used in a thin, 1 at high temperature ‘ohm at 50 V de, If this g is susceptible to erro, dity which deteriorate e been produced in hniques have also In the flame- e wires of theBae (a) Fla: 241 Rosetied strain gauges: (a) three-element stacked type, (b) two-elerrent right-angled, (©) three elements at 45° to each other, (d) three elements at 120° to each other, (e) gauge pattern on a diaphragm.