Scribd
Upload
10 views26 pages

Tensor

Useful for mathematical physics

Uploaded by

hisanahamza123
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
10 views26 pages

Tensor

Useful for mathematical physics

Uploaded by

hisanahamza123
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
r 3 Tensors ae y In many physical problems vector notation is not sufficiently general to express the between quantities involved, For example let us consider the flow of chectre: carvent erween current density vector J and the electric field vector E is given by vector equation J=0E er) where o is electrical conductivity of the medium. For such a medium the conductivity o is scalar and for the X-component we write Ji = OE; --@ However, if the medium is anisotropic as in many crystals, the current density in x- éiections may depend on the electric field in the y and z-directions as well as on the field in x- rection. Assuming a linear relationship we may replace equation (2) and write Jy = 6y1Ey + 612Ey + 6y3Es -B) and in general =X ouke 4) Thus, for ordinary three-dimensional space, the scalar quantity ¢ must be replaced by a more general quantity 6 (a set of nine elements) to effect the required change in direction and magnitude - SH) oy, O23 On G2 Os 63 03233 This general quantity 0%, (array of nine elements) is called a tensor. The description of tetric, optical, electrical and magnetic properties of anisotropic solid may well involve tensors, The fundamental laws of physics must possess the same form in all coordinate systems. The study of this accepted fact, known as invariant formulation of physical laws is embodied in ‘ensor analysis. Tensor analysis is a generalisation of vector analysis. It paws for the answers ‘0 the questions as: ) Ace all laws of physics expressible in terms of scalars and vectors only ? Gi) With respect to which transformation does a certain physical law remain invariant ? (ili) How should a certain physical law be written if a wide class of transformations is fog 1M ‘ensor formulation became popular when Albert Einstein used it as an excellent tool the presentation of his general theory of relativity. Now-a-days it is invaluable in its ‘Plications to many branches of theoretical physics : mechanics, elasticity, plasticity and “lctromagnetic theory etc. kK Mathemat; 232 Cal Phy, -2. n-dimensional space : : Pein pes et i etme 2 fH mis, i int. For example, (x. int in recta, Clean corn sytem. By analogy if 3 pont i epresented By St ofm reat yan conveniently (x!, x . xXes (1,23) tus 2s Fess aie powers ofthe variables involved), then all the points corre. Coal values of coordinates variables) ar ssid to form an n-dimensional space, denoveq ‘A curve in n-dimensional space (V,) is defined as the collection of points which sat the equations =x (u), (= 1, 2,3...) where u is a parameter and x# (u) are n functions of u, which satisfy certain contin equations. ; A sub-space V,, (m det ) {Also two ot more dummy indices can be interchanged. In order to avoid confusion the same ; 2 index must not be used more than twice in any single term. For example (3 ay)? will not " be written as aya, but rather aya,xx". Any index which is not repeated in a given term is called a real index. For example, His Areal index in glx, A real index can not be replaced by another real index, ¢.g. ah # ag’ x 4. Kronecker delta symbol Kronecker delta symbol 8," is defined as 1 iffey © ve[) ifuey 2 0 233 A Mathemation » ie , ronecker delta Some properties ue independent variables, then (plist at aay property of Kronecker delta symbol is obvious * By AR = AY jth n-dimensions, then 3h =8’=n y vention OS 25, +, +5;°..- +5te1+14+1+..4+1l=n yi w@ 8H 8," = 8g! By summation convention 5,25," = 8! Bo! + Spt 8,2 + By! By? +... + By Bo" + 8," By" = 0 + 0 + 0 Hu. Fbb P+... 40558 Wa (iii) If we are dealing wi Ox OxY oe ay © agvaxe™ ax Generalized Kronecker delta. The generalised Kronecker delta is symbolized as nap Bas and defined as follows : () The subscripts and superscripts can have any value from 1 ton. the i sither at least wo superscripts or at least any two subscripts have the same value fe subscripts are not the same set as superscripts, then the generalised Kronecker dela i 22 ‘or example, Ot = = i= 3.0 ‘) If all the subscripts are if re Separately different and are the same set of numbers as the weneTbts, then the generalised Kronecker delta has the value + 1 or — I according 2 whether it requires an. ev 2 Same order as the subscripts. Pca of permutations to arrange the superscripts in te 8123 5123 _ gras po = 8a = Bis = + 1 = 513 bis 8 = 8i452= ea te ea le HMM ow OV Va Vy ns vg "Gono all values of the variables x!, x2, .. x, then show th! Solution. We have Ditlerentiat 40g 8 = 0 0 ‘eatiating eqn, With respect to ‘ agg 0 3? we get ‘a ash * gb 22 _ « “OP + 0168 5,000 ker delta ties of Kroneel ae rae , __ x" are independent variables, then (its',2, ny OY ag (ii) An obvious property of Kronecker delta symbol is it By A = AY a (ii) If we are dealing with n-dimensions, then bts 5%=n 7 By summation convention - BP =8 +8245). t= 14 le leg @) BBY = 8H ie By summation convention 8B’ = BiH Bo! + Bal 8,2 + By! 8,3 +, = 0 + 0 + 9 + oa _ age l=n + + BH Sgt + 8,1 6," eth bo + 405 8H and defined as follows : * De Subscripts and superscripts can have any value from 1 to n 4i) If either at least i : ne ag ie ne me a Pescripts or at least any two Subscripts have the same value For example 88 superscripts, then the generalised Kronecker delta is 7. uw (iii) If a the 5 ii Superscript, then Ser EM and are the same set of numbers as th sete eg ber of ela has the value + 1 or ~1 according 3 order a the subscrpes le Mations to arrange the superscripts inthe aR s B= BFP = and BER = 523 It should be Noted Bre gta Rey Se BB on Bh Hy ab 8 - y oS Oe Sw + dy = 0) Jor an Values of ihe a y that . We ane Variables 1x2 em then show Differentiating eqn. ( let ° with "spect ag, xo O40 {0% we pet | we a ag Ox Pe. or a Fapx98,0 “a S| at 235 doy XB" + aygx° BH =O gy X* + Ayn = 0 rentiating eqn. (ii) with respect to 2%, we get at, oe Aaya w * HB QxY ay BF + yp 8h = 0 i a8) + dyy By = 0 a yy + Ay = 0 ‘ 3S (a Se rs \alue 6 in another system of variables x#, Then if o sefantion 0 is sid to be scalar or invariant or tensor of order zero. Te quantity By = 31+ 82 + By =n ae 16 gw diffe wo (ii) =0 scalars, Contravariant Vectors and Covariant Vectors ‘alars. Consider a function 9 in a system of variables x! and let this function have isascalar or an invariant. (b) Contravariant vectors. Consider a set of 1 quantities A!, A2, A, ... A" in a system ables x" and these quantities have values A!, A?, A3,... A” in another system of of vat saiables # If these quantities obey the transformation relation Zu _ oxt Ab=S 7 Ae ve Oxt (12) ten the quantities AY are said to be the components of a contravarinat vector or a contravariant tensor of first rank. Any n functions can be chosen as the components of a contravariant vector in a system of variables X#. Multiplying eqn.(12) by at and taking the sum over the index jt from 1 to n, we get av ax® axe axP ap xb gy Ox OH OX” go = 88 Aw = AP Se, Meade, a Ate ge te aA 9 ax8 a po au (13) Abas, A Equations (13) represent the solutions of equations (12). ‘The transformation of differentials dH and d=# in systems of variables x" and 3 te *Pectively, from eqn. (Sa), is given by xt 14 = Soa wa (14) B= a dig’ ®@uations (12) and (14) are similar transformation th Pe Tentials dx# form the components of contravariant vector, whose ‘components in any other st conn the differentials dH of that system. Also we conclude that the components of a Variant vector are actually the coniponents of a contravariantrensor of rank ane. equations, we can say that the Mathematica, Phygi cg 236 era fer change of variables from #10 2% then the ney ons given by gm OE 4 ysing (12) am A Let us now C0 components A”™ must Be ax ls ion(12). This indicates that the tr ' ‘Tis equation has the same form as equation( 12) ansfomatny ccontravariant vectors form a group. Note, A single superscript is always us explicitly stated. ie f OG cocen Vectors. Consider a set of n quantities Ay, Az, A3, ... Ay ina system of 5d to indicate a contravariant vector unless the conay is variables x and let these quantities have values Ay, A2, A3..- Aq in another system of variables 3. If these quantities obey the transformation equations 7 2 age Aa (16) then the quantities A, are said to be the components of a covariant vector or a covariant tensor of rank one. Any n functions can be chosen as the components of a covariant vector in a system of A variables x and equations (16) determine the n components in the new system of variables 7 Multiplying equation (16) by & and taking the sum over the index j1 from 1 to n, we get axe axe axe = 3 pzu 40> Sp Aa= Ap Thus Ap= wl) ame ax Equations (17) represent the solution of equations (16). Let us now consi a a A’, mast be vanes irther change of variables for x# to x’#, Then the new components ee A= oe ask ‘ 4a api’ Fra An (Using (16)] #* axe fervent vecarto ae frm “S equation (16). This indicates that the transformations “ awn i BS Bet ay / : Ox Axe it follows from (16) that 2Y. Me : a8 rm the components of a covariant vector whose componen's im any other system are the correspond ‘Ssponding partial derivatives a Py ard y. ae ‘This covariant vector is called 237 A single subscript is always used to indi i on gated; but the exception occurs in the notation of cosets tess COMERY is iit n of coordinates, on ys, Tensors: of Higher Ranks 7 he laws of transformation of vectors are Contravariant = Abs 712) Covariant : a= (16 (a) Contravariant tensors of second rank. Let us consider (n)? quantities A¥Y (here jt and take the values from 1 to n independently) in a system of variables x and let these quantities have values AMY in another system of variables X#. If these quantities obey the transformation equations Suv. OF OF pv O24 OXY aay A 3x oP AxB (19) then the quantities AMY are said to be the components of a contravariant tensor of second rank. The transformation law represented by (19) is the generalisation of the transformation law (12) Any set of (n)? quantities can be chosen as the components of the contravariant tensor of second rank in a system of variables x* and then equation (19) determines (n)* components in any other system of variables X# . (b) Covariant tensor of second rank. If (n)* quantities Ayy in a system of variables x are related to other (n)? quantities in another system of variables 2" by the transformation equations + (20) then the quantities Ayy are said to be the components of a covariant tensor of second rank. The tansformation law represented by (20) is a generalisation of the transformation law (16); Any ‘et of (n)? quantities can be chosen as the components of acovariant tensor of second rank in a Sysem of variables 2 and then equations (20) determine (n)® components in any other system of variables ZH (c) Mixed tensors of second rank. If (n)? quantities Ay in a system of variables x# 4e related to the other (1)? quantities AM, in another system of variables x# by the ‘ansformation equations : Oh Age vs 21) ponents of a mixed tensor of second rank ker delta 8)", is; while in covariant tensors ntravariant vector with then the quantities Ay! are said to be the com ‘An important example of mixed tensor of second rank is Kronec Note. In cont tensors the indices are placed as superscript the indices are laced ng subscripts, The mixed tensor AyH transform Ike 4 60 "spect to index j1 and like a covariant vector with respect (o index V rf Mathematical Phyg; eg 238 : Rank of a tensor ‘ensors ranks * a are nai ks are defined bY similar i 110 it per component. For ws. The rank of @ tensor only indi ‘tes + xampel Ay? are the components of 7 The tensors of “hed jant of rank | ; if th the number of eo ara carn of rank 3 and covariant 0) ey transform mixed tensor of rank sraing to the equation - according ame aw’ a3° x* 4 ,oBy (22) Ay = aya gx? Ox” 0x* , the number of dimensions gives the num| The rank of a tensot whe i eso ° Prank rin n-dimensional space has an of compen Sank ‘of a tensor gives the number of mode of changes of a physical Ce ig from one system to the other which is in rotation relative to the first, Bd venity that remains unchanged when axes are rotated is a tensor of zero rank, Cron 8 Od wad similarly the tensors of rank one are ‘The tensors of zero rank are scalars or invarian| vectors. SOLVED EXAMPLES Ex. 2. Show that the transformations of tensors form a group. (Meerut 1998, Rohilkhand 1998, Kanpur 2004, 1992, Banaras 2003, 1994) Solution. Let Ay be a mixed tensor in a system of variables x". According to tensor transformation law the components Ay! are related to the components Ay! of the tensor in the system of variables X# as ox ax wali) Let us now consider a further change of variables from X#to x’#, then the new components acne 1 OFT aH AFP AFH ay axe ax¥ Ox axe = ax’ OX4 Oy Ox © O%¢ Ox! az axv (i) Ex. 3. A covariant tensor has compo) ts covarians covariant components in spherical coon oO 2) ~®, rdinates, » 32 in rectangular coordinates. Find (Calicut Uni Solution Here ™V., 2003, Bharthi Hen eye, ar 1985. Madurai 1986, Kanpur 200!) @ 9 1 peransformation equations between coordinate systems are ® x! = ¥!sin ¥? cos 33 x= Fl sin ¥ sin 8 x= Feos # ~@ Then he required covariant components, from eqn, (), are given by | ax? ax3 | re gg dat gers = (sin cos X3) (x! x2) + (sin 32 sin 8) (2x2 . ~ G)4) + (Cos ¥%) (xl x3) | = Gin 8 cos 4) 7? sin26 sin $ cos 4) + (sin 6 sin 6) (2r sin 6 sin 9-7 cos?) | + (Cos 8) (r? sin @ cos @ cos 6) = @! cos X? cos 33) (x!x2) +X! cos ¥2 sin 1) (2x2 ~ (x3)2) 4 Cl sinx?) (x2) = (rcos @ cos 4) (r? sin?® sin cos ) + (cos @ sin $) (2r sin @ sin $ — 72 cos? 6) + Crsin 6) (7? sin 8 cos @ cos 6) = (Cx! sin ¥? sin %) (xlx 2) + (@! sin 2 cos #) (2x2 ~ (x3)2) +0 = (rsin @ sin 6) (7? sin? sin 6 cos 4) + (r sin 8 + cos 6) (2r sin @ sind — r2 cos) Ex. 4. Show that in Cartesian coordinate system the contravariant and covariant components of a vector are identical. . (Meerut 1989, Madras 1986, Banaras 1973, Rohilkhand 1997) Solution, Let (x, y, z) and (X, y, Z) be the coordinates of a point P relative to orthogonal Cesian coordinate systems $ and 5” respectively. If (li, my ;), (lp, ma, ny), (Igy ma, ns) are “retin cosines of axes 2, y, z respectively, we have Bekjxt my + mz F = bx + may + mz @ Zalyx + myy + nz From which we have x=L¥ +hy+ he y = my z+ m2y + m3z “@ Zk + my +32 z; so that (i) and (if) may be written as [yx! + myx? + nyx3 vn (iii) Pre = Igx! + mpx? + nyxt = : 53 = Myx! + myx? + myx OMEN Aion al Pp, "Ybig * z eee i ale Aah sai ey ye mit mye? + ms * cs oy? a em ange +m ty jant vector AM is given by = Q equations (v) may be written as Using equations (ii), Als Alt mA? + nA LA! + myA2 + m2A> + (vi) Bo = Al + yA? + 348 ‘The transformation equation for covariant vector Ay, is given by 5 oot 4 -2e e+ (vi) Using equation (iv), these may be written as Are Ay + miAy + mAs a + il = hAy + moA2 + nAy Ay = Ay + msAy + ngAy Comparing ations (6) and (8) we © see that there is no distinction between contravariat ‘and covariant vectors ) coordinates. when we restrict ourselves 10 rectangular cartesian transformation ¢) 310. Algebs diane etnies of Tensory the case of tenon apabtrmetion, ‘The addi contravariant and covenant cant ‘and vain nd subtraction of tensor is defined 08) fp ices, The addin Same if ol sation or subeaton oases ke y | 241 n ts. To add 1 india elements. To add or subtract wo tensors the coresponding clement are jo 6 ret ame of two tensors of the same rank and same type is also a tensor of the 0.7 opt if there are two tensors Ag! and Bg | ee eition and subtraction are given by ‘yo Agi’ + Bg!” = CoH” (addition) Ad” - Bo” = Dg” (subtraction) erm ae the tensors of the same rank and same type as given tensors, pe rassformation laws forthe given tensors are Of the same rank and same type, then | prs Oee ORY Oat ay axe ab age AY --@) Bova FE ORY Out BoM = ox* x8 axe he 1 A) ‘xdding (3) and (4), we get Gyn OFM RY xt Og | Co = Bea OeB azo O ) ‘eaisa vansformation law for the sum and is similar to transformation laws for Agi” and |p sven by (3) and (4). Hence the sum Ch (= Agi” + Bg!) is itself a tensor of the same aiand same type as given tensors. Subtracting equation (4) from (3), we get = yO" OEY axt | Aol Bol = aes 3,8 age (MPP BY) | sich is a transformation law for the difference and is again similar to the transformation laws AZ and By”. Hence difference Dg!” (= Ag!” ~ BoM) is itself a tensor of the same "stand some type as the given tensors. The addition of tensors is commutative and associative. (i) Equality of Tensors. The two tensors of the same rank and same type are said to be ‘al iftheir components are one to one equal ie., if AMX = BB} for all values of indices mo) ‘ensors are equal in one coordinate system, they will be equal in any other coordinate Thus, if a particular equation is expressed in tensorial form, it will remain invariant inate transformations. | il Outer Produet. The outer product of two tensors isa tensor whose rank is the sum ‘anks of given tensors, ‘epi ifr and 7 are the ranks of two tensors, their outer product will be a tensor of rank Hmm ‘em, Sig "erexample,if Ag and 1B, are two tensors of ranks 3 and 2 respectively, then *aensor Ad By! = Capl* (say) ~@ | Of rank (3 +2) = 5, ~~ 242 transformation equations of given tensors as ite the e For proof let vs WH axt ax” et 4,08 Ad's 38 xh OF? 8) _. deh art ps B= war 2) . yet uring O46 WOW sy ag aut OF* Oe 4 opps Ag" B= Fax ae axe OM axP 7 age az” OE" Ox? Ox* Oops Ysi Eynts ee Spas arose Ue) (ly 5, Hence the outer product of two tensors ong wh of rank 5 (= 3 + 2). o < i ff rank which isa transformation law for a tensor 0 Ag™ nank 3 and other B,} of rank 2 is a tensor Cop! ‘The outer product of tensors is commutative and associative. (iv) Contraction of tensors. The algebraic operation by which the rank of a mixed tensor is lowered by 2 is known as contraction. In the process of contraction one contravariant index and one covariant index of a mixed tensor are set equal and the repeated index is summed over, the result is a tensor of rank lower by two than the original tensor. For example consider a mixed tensor Ajg!° of rank 5 with contravariant indices }1,v,6 and covariant indices 2, p. ‘The transformation law for the given tensor is age OF" 95° a art Oxt x8 OxY OFA xP To apply the process of contraction we put p = 6 and obtain oF as HE hy = OF IXY aS Oxt OES 4 py Ox8 Ox! axt age = dem ax gzh axe axt Aig = Ages (2) Aygh? = te et xY Oxd a 5 85 Ay,07 ( since 2 9F% _ ax 5 xP aga Or Aa’ (since SS = 5 =5;) = 27H FY as Ox BaP aga Ani (13) Process In thi and obtain the contravariant yesrs Cna™PIe We can further apply the contractio® enables ie Mt . 1 0 obtain a tensor of rank (r~ ee Avo!” or AM. Thus the process of contraction As another example consider ie 2 8 ied tensor of rank transformation laws is the contractio : % "of the mixed tensor Ay! of rank 2, HOS Ty 2 OF ad 4 To apply Ox av Api ™ 88 We DULY =u ang Abn Fa, obtain 243 formation law of a scalar of an invariant. As by contraction Process the rank 2 want i nsor Ayt has been lowered to zero, itis justified to call a scalar a tensor of rank aie #® contraction process can not be applied with respect to two indices of the same nature i.e. oe ariant or both covariant, Inner Product. The outer product of two tensors followed by a contraction results in ‘ensor called an inner product of the two tensors and the process is called the inner a1" ication of two tensors. ‘xamples. (2) Consider two tensors Ag!” and B,’ rhe outer product of these two tensors is Adi By* = Caph* (say) Applying contraction process by setting p = 1, we obtain Ag! B,* = CoytY* = Dg** (a new tensor). ‘The new tensor Dg” is the inner product of two tensors Ag!” and Bik (0) As another example consider two tensors Al and By, each of rank 1. The outer product of AM and By is AY By= CM (say) Applying contraction process by setting v =p, we get AX B, = C,H (scalar or a tensor of rank zero) Thus the inner product of two tensors each of rank 1 is a tensor of rank zero (i.e., invariant), (vi) Quotient Law. In tensor analysis it is often necessary to ascertain whether a given tay isa tensor or not. The direct method requires us to find out if the given entity obeys the tetsor transformation law or not. In practice this is troublesome and a similar test is provided byalaw known as quotient law which states : Any entity whose inner product with an arbitrary tensor (contravariant or covariant) is a ‘ewor, is itself a tensor. Example. Let A (j1, v, 0) be the given entity to be tested whether it is a tensor or not. ‘pply quotient law let us consider an arbitrary tensor B,,* whose inner product with A (11, ".9), is a tensor, i.e, el : + (14) pa have to show that A (1, v, 0) is a tensor. In the other system of variables X#, we must A (U,V, 6) Byg* = A (Uv, 6) Bygt= Cy He Ox axB axt , _Ox* axe A (tv, 0) axe azY axe Bp; = 32 axe Using (14), we get Orb. Oz* axB Oxt ox* axe a ox 3.0 *Byd (hv, 0) Se Sev gz0 POP ay azn 4 (> B 1) * Boy * ox ax6 axt — axe ae [a vos ge 4 8.) eo =0 Mic “y Me “a athematica, Pr ‘On inner multiplication with 22. we obtain icy inner multiplication x Fife xP xt - ax* be a [Kan Sev az 2E 4 cB. | Bp} 0 in — ax® ax? — x oe _ a [a uv.) S Sea age 4 (Bs » | B51 0 ie 7 ax® dat - axt ] . [a WY.) a5y ag0 | agF A (a, B.Y) | Bp,t = 0 Since By, is an arbitrary tensor, we can arrange that only one of its com, bia’ zero. Now each component Bpy* may be chosen in turn as that one which does a 5 on. ‘Therefore, the expression in brackets is identically zero, i.e., adish, yh Oxt — axt A WY 0) ey age agH A (a, B, y) =0 Again on inner muliplying with 25 Sere we get ae aF 5 ax Qe’ xP Ax Axe axb ari A HY) 355 arom ab axtazn 4 (OBL) =0 Regrouping, we get Oe? axb ax¢ axt 7 ax? Ot axe ab agv at aze AV O- 35 ar age 4 (OB. =0 . axe axe or BP St A uv, 0)= 98 a = A(@,B,y) z axe a¥° axe or AH P.8)= 3-0 oe dur 4 (B.A) * ey ag Equivalently A (u,v, «= 28 ae A(@, B, 2) wld) ‘Tis equation represents the transformation law for the entity A(j1, v, 6) which is a tens transformation law. Hence A (11, V, 6) is a tensor. Equation (ts) pees the transformation tensor A Gi, 0) justifies Up socgyam index and vc as contrvariant indices. Hen® “ ity Thus quotient law not only tells that the ent! AK, v, 6) + its § : aeeiat 4 (u,v, 6) is a tensor; but also tells its ty (vii) Extension of a rank — The -ferentiating i each componant with respect to variables of a tensor can be extended by differenti ample ‘ a S example, consider a simple case in which the original tensor of rank zer0: # scalar 5 (2) whose derivatives Telative to Variables x# are #, the seater is $G*) such that Sas aya : ant ‘Pe ie. number of contravariant and cova ables as. In the other system of vari asi ~ 6) a7” ox ara a ‘This show thar 2%. transi : ne forms i us itor of tence of rank zero, ike the components of « tensor of rank one- ™ atthe V6 «tensor of rank one. In general, we may s#Y 245 a on of a tensor with respect 10 variables 3¥ yields a new tensor of rank one great er cgi 07 tensor can also be extended whi Meath oth respect tothe tensor is Fert AG ae ana tensor ane ty depending upon another tensor Ayy then nn one 8 tensor S * (i pane 25_ 2 By =a tensor of rank 2 (17) Aw poste rank of the tensor of rank zero has been extended by 2. SOLVED EXAMPLES “aH and By are the components of a contravariant and covariant tensors of rank Cy'= ABB, are the components of a mixed tensor of rank two. pelt (Meerut 1995, Rohilkhand 1999) sw hat sautin. By tensor transformation law : == _ OxH , axh Ox" ax8 ax* ax8 Gus AuB, = S At SS By= Se A, At Bp= 52 550 OO" ott See A” sev B8= Bue az 4° 88> 3, Sp sichis a transformation equation for a mixed tensor of rank two. | Hence Cyt = ABB, are the components of a mixed tensor of rank 2. | Ex [Link] and B, are any two vectors, one contravariant and the other covariant, then oe that APB, is invariant. Solution. We have Zu 22H go Ate axe’ a8 B= 25,8 hela ga Foe EY BD ge py 222 A By = BPAY Bp =A Ba ee i je Bean formation of coordinate system; hence AM By, is ‘the Value of AH By is unaltered by the trans! ee ee and covariant vector aay This result expresses a useful relation ugh 7 Prove that Ayy BE CY is an invariant ‘covariant tensor. _ Solution. By transformation of tensors = _ ax ax Aw = azn ax¥ between contravariant if BH and CY are contravariant vectors and Ayy (Meerut 1986) Mathemation 7 “Weg 246 ane ak" oes _ aad, oF OF op _ ax OF" |B ar, Ay BP © = en gqv AP ax ax "aR x ORV BF Ante 2 5 By? AapB2O? = AgpB%CP 3 Tie BH C= Aye BECY : js an invariant tensor. Hence AyyB#C’ is an invariant 3 te Ex. 8 rr ‘an arbitrary contravariant vector and CyyA¥AY is an invariant they es that (Cy + Cry) 8.4 covariant tensor of second order. (Rohilkhand 1998; Banaras 190" . i Vi « Solution. Given AW is an arbitrary contravariant vector and Cy,A¥AYis an invaran, Eqy AH AY = Cyy AB AY Using tensor transformation law, we have ~() = = = 5 OF A OM yp_G OeH ORY yy AB AY = Cw 5 A® Sa AP= Cw Gye OxP A&AB Now interchanging the dummy indices } and v, we get oat Ok ax® axb Again, interchanging the dummy indices o and B Gy At Av 2G, 2 OE ‘Adding equations (2) and (3), we get Cy AY AR= Cy At AB Aan 8) xt oxY Gu + Gy) AP AY = Gyy +6,)@ = i But from equation (1), we have Am AB ) Cuy AB AY = Civ AHAY = Cup A AB and yy AY AM = Cy AYAH = Cpa 8 Ae 0 that Cy + Cy) AY A ; (Guy + Cyy) AH AY = (Cup + Cpq) ARAB we Using (5), equation (4) may be written as (Con * Cua) A AB = Ey Fy) oF gas i ie, [G. a, am aR : ww + Cy) i (Cap + Cy] A248 = 0 Since A and AB te And AP are arbitrary vectors, therefore the expressi ust vans! ression within brackets ™' ox ARV axe x8 ~ (Cap + Cpa) = 0 v m7 on with 22 22% ¢f ng above ‘equation with pen agv' We Bet eat" age axe axe OY OxP Cap axe axh Ew) xe OF oF aw ~¢ an or = 2 a ax ax Gent 5" (Coa a) ation law for the covariant tensor of second rank. Hence (Cyy + Cy) is a re easor of the second rank. es ta scalar, i the tensor Aj. pees pt Gren tensor i Alp The key idea is thatthe contraction process reduces the rank The given tensors of rank 4, Therefor if we aply the conta wat ofthe given tensor willbe reduced to zero; soit will become a ain process oem aw of given tensor is 7 = Be Det gap ax Ox? age ag” A) ox! 2 (a 2) = Oa aga \Ox8 azz) Ar" _ OF OXY oe ap = axe = Aye = OH ORT of 3 ae +2) ‘Ss easor transformation law of rank 2. ‘gan we take inner product and set k = i in equation (2) y= Se Ah a Ay =A => A=A simples that AGB = Ais a scalar. 1. Show that re(Si? we second order tensor in two dimensions “a xP xy) 84 mit, meg 8 = (EP m2? ) nora tenor 1998) A (Delhi 2007, Rohilkhand 2008, its jon. If ris® tensor then # Solution: - pie-a en mr Les sa obey tensor transformation enh If we cons in Cartes! coordinates counterions intonshipbetwee® Come 1s re original unprimed) system and those resolved new rotated cried) system gyreny cos 0 +478i0 0 wert sin @ +x) 008 @ ax’ an _ 4} ‘These equations yields, = cos @, sin an os 4} 0) ag OL Frans oe In the primed coordinate system the (T"})! component must be —x) x2'3 then from tensor transformation law yg aOR ay (ri)! say =D Se gr ay Bey i" yy, 1 Bt yy BH! BL par ax! Ox! 99 By dy Pt ax Om)” Oxy Ox + Oxy Ox . Using equation (2), we get (THY = cos? 7"! + c0s 8 sin 8 T!? + sin @ cos @ T®! + sin?@ 7”? Now expanding the left hand side in primed coordinates from (1) and substituting the values of components 7"!, T!2, 72! and 7%? in right hand side, we ite ~ (x; cos @ +2 sin 8) (~x; sin @ + x2 cos @) 2 - cos? i cos? @ x12 ~ cos @ sin @ x)? + sin @ cos @ x)? + sin? @ x12 op 217 sin @ cos @ ~x1% c0s?@ + x12) sin? @ ~ x2? sin 8 cos @ 2 == cos? 8 xx) — i i 142 ~c0s @ sin @ x)? + sin @ cos @ + sin? @ x12 This equation is an i of other tee ccapana Take! re that tensor transformation law is obeyed- Ret oD Hence oe oe tfes that all ransform according to tensor-transformation ‘Hence le that matrix 7 is a tensor of rank 2. Now we consider “t,x? ( ° “tin having components SU ayn, $1? = on, tition Faw ol 8 +821 = 21? und $2 = rx) al 249 we again have, BH 56, Ms sin a on : ax = sin @ and oo =cos 0 n she component Sin primed coordinates must be , Ox)" dx,’ (sy! =o’ = ae ar OY dayne guy 5 BRL Bey. gp , Bay, oy, n'a - —L Say Bey an) By Gag dt Bee axe Set Bay Dae = cos?@ S!! + cos 6 sin @. S!2 + sin 6 cos @ $2! + sin?@ . $22 Now expressing the left hand side in unprimed coordinates from (1) and substituting th caves of components S'!, S12, $2! and $2 in right hand side, we get (1) and substituting the = (x 608 @ +.x9 sin 8) (— x; sin 8 + x2 cos 8) 522 cos? @ x1x2 — cos @ sin @ x)? + sin 8 cos @ x,? — sin?@ x)x7 = x;2 cos @ sin © — x,x2 cos? @ + x,x2 sin? @ — x,? sin 8 cos @ 5 sin 6 cos @ (x4? = x9?) — xyx7 = sin @ cos (x,2~ x32) + x,x2 (sin? @ ~ cos? @) in © cos @ (x)? — x92) — x47 => Xpx (sin? @ - cos? 8) X 1X2 This equation is not satisfied since both sides are not equal. In other words the components of S do not satisfy the tensor-transformation law, hence, matrix S is not a tensor, 411. Symmetric and Antisymmetric Tensors (a) Symmetric Tensors : If two contravariant or covariant indices can be interchanged without altering the tensor, then the tensor is said to be symmetric with respect to these two indices. For example, if AMY = AYE . Seon 0 fea the contravariant tensor of second rank AB or covariant tensor of second tank Ayy is said ‘be symmetric. For a tensor of higher rank A,Y*? if Ayito= Avs then the tensor A,H¥9 is said to be symmetric with respect to indices Hand v. | The symmetry property of a tensor is independent of coordinate system used. So ifa tensor Symmetric with respect to two indices in any coordinate system, it remains symmetric with 10 these two indices in any other coordinate system. This can be seen a follows 1€ tensor 4;¥¥® js symmetric with respect to first two indices pt and V, we have is | Ayiv9 = A,vHo y Mathemat; 7 "01 Phi ‘ion law coordi to tensor transformation i eee age BEY OEE OF 4 apy 2EE OF OFC Oy nay aye oe SS Ser gee one BF Oa aA AP Msing 4) Now interchanging the dummy indices o and B, we get c cu aR” OF Ab apy OF" AEH OFT Oy Oy - aye ey Sar age A= aus an et aan AN *T= Apne : is agai ic with respect to first two indices in the new coory; ie. given tensor is again symmetric wil ties ordinate syste. ‘This result can also be proved for covariant indices. Thus the symmetry Property of, tensor is independent of coordinate system. : - ‘We can not usually define symmetry property with respect to two indices, of which one is contavariant and the other covariant, because this symmetry property can not be preserved aftr ‘a. coordinate transformation. It can be confirmed as follow: Let Ay#"® be symmetric with respect to two indices, one contravariant 1 and the other covariant A, then we have Aye = Ayr +125) By tensor transformation law, we have ORM OEY OES aS | og ARH IXY AF? AS. yy Ai = Fa DB qt age A8'= Qe xB ax? gga 4a PY (using 25) Now interchanging dummy indices a and 8, we have Ayre a 22" OEY OFF ave a= ax OxB xT axe According to tensor transformation law z Ox* AXY Axe ad awa = OEP ORY IEE OS 4 gy A= aa OB axt ggu oD) Cornparing equations (26) and (27), we see that Ayo eve i.e. symmetry property is not preserved after a change of coordinate system if we consider the symmetry with respect to one contravariant and one covariant indices, But Kronecker delta which is a mixed tensor is symmetric with respect to its indices. Ager 08) win all the indices of either a contravariant or a covariant tensor can be interchanged altering the tensor, the tensor is said to the symmetric. A symmetric tensor of rank 2 in n-dimensional space aint) Pace has at most “+1 independent components. This may be proved # follows : The total number of components i terms will be different and the rest (x2 ay’ n?-n mn) ‘Therefore the total number array will be n?, out of which all the 1 diago™! ~n) will be equal in pairs. The number of pairs W! of independent components =n +e =A(n+ 1) 2 (b) Antis i magnitude, whet nee cota A tensor, whose each component alters in sign but NOt symmetric or antisymmetrle with sce indices are interchanged, is sid 10 be SHO" Wh respect to these two indices. For example, if 0 251 AW = — Avi Ay =-A vi, (28) i y ; + giaariant tensor APY or covariant tensor Ay of second rank is antisymmeti, vaca ensorof higher rank Ay! if Ashi = Ayo seer ATs asymmetric with respect 0 indices v and o, ‘The skew-symmetric property of a tensor is also independent of the choice of coordinate “gst. SO+ if a tensor is skew-symmetric with respect to two indices in any coordinate sa it remains skew-systematic with respect to these two indices in Sem This can be seen a8 follows : any other coordinate sfatensot A;!* is antisymmetric with respect to first two indices j1 and v, then we have Ayn? = — A," ‘ py tensor transformation law, we have ae" az axe as Ag =— Sa BF pay gu AvPmusing (29) Now interchanging dummy indices 0 and , we get azn OxY Ox* oud Avo, —— Se. AIS =~ O58 Ox® xt Oe* Agi = — Aye ie given tensor is again antisymmetric with respect to first two indices in new coordinate stem. Thus, antisymmetry property is retained under coordinate transformation. Antisymmetry property, like symmetry property, can not be defined with respect to two iadces of which one is contravariant and the other covariant. [fall the indices of a contravariant or covariant tensor can be interchanged so that its components change sign at each interchange of a pair indices, then the tensor is said to be auisymmetric i.e, If A#¥® is antisymmetric tensor, then, AVES = + AVOH | Thus we may state that a contravariant or covariant tensor is antisymmetric if its components change sign under an odd permutation of its indices and do not change sign under % ven permutation of its indices. Auvo = An antisymmetric tensor of rank 2 in n-dimensional space has~~"> independent “mponents. This may be seen as follows : é The total number of components in the array is (n)?, out of which all diagonal terms of ray will be zero since all the quantities AM# (no summation) are zero. The rest (n? — 1) : 2 "be equal in pairs, The number of pairs will be": ‘Therefore the total number of independent components n-n_n(n=1), h ay 2 iq." #eFal an antisymmetric tensor of rank rin n-dimensional space will have at most yen! | Vm yt independent components, Wns nthen"C,="C, = MAOHNeMatical Physio, he number of dimensions of the Space equal 10 has ie rensor of 70" any ponent. f dimensions of th ent one meee Ok ie a numbers © Pace are 7 = tensors Of Antisymmettt’ te nt or two covariant indices can , varia’ - two coe tric and the other antisymmetric, Thus. fs, sum of WO _ Ave ta expressed a @ me an + AY) ham AY) ; | ; (30) oat part and the second is the antisymmetric part, the first teem on the right is sr be wrisymmetric partis a antisyrmmettic tensor gyrametric part is symmetric tens saan of symmetric and. antisymmetric tensors not only The process of writing @ Seis general. For example a tensor Aga?” can be written as rrokds for tensors of rank 2; but 1s ‘ WY + Aga] +2 . i tric part metric antisymmet 1 mn yy) + HAgitY - Aa!) Ae” = Float + Anat] + 21Aea! — “0 antisymmetric part | sat” =1[Ao! (aon ~“oh symmetric part SOLVED EXAMPLES Ex. 11. Show that any tensor of rank 2 can be expressed as a sum of a symmetric and an antisymmetric tensor, both of rank 2. {Rohilkhand 2001, 1990, 88; Meerut 2005, 2000, 96, 93, 84, 83] Solution. Any tensor A” of rank 2, may be expressed as AW = 3 (AMY + AH) +3 (ABY— AYH) Be + OW =} (AMY + AY) and CW =} (ABY — AYH) From addition and subtraction laws of tensor it follows that BY and CY are tensors of rank where Be 2 Interchanging indices in BY and O, we get Be = F(A + AM) = 5 (ANY 4 AM = Baw and mes =5 (A Any) 2b 2 (AMY — Avi) = _ Cv which show the B*” is symmetric, while CHV ay ge - haps is antisymmetric, both being tensors of rank 2 x12 I Ay is antis) ‘Sacha’ vie reat fensor, find the component A ilkhand 1998) " on of antisymmetric ten; W (Rohilkhan Seting i= j= 1 80F Ay =~ Ay Auee Ai = 2Ay, = 0 oF Ans0 Ex. 13. Uf the tensors dy ag independent ¢ ie inj and, per omponen in four dimensiine® antisymmetric in every pair of indices, find tei" Solution, The independent o J ir of indices, dimensional space are ™ components of an antisymmetric tensor of rank 7 in ™ 253 (om independent components of ain four-dimensional space: _asagis 2 12RS0F of rank 2, its independent components in 4-dimensional space are 4c, 24 X38 Tx 275 1x2 psn be seen 25 HlOWS ‘The total number of components of ai (tensor of rank 2) in 4-dimensional space = (4)? elt 6 . 5 From definition of antisymmetric tensor ix = ~

You might also like