版權(quán)說明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請進行舉報或認領(lǐng)
文檔簡介
1/1
張量分析翻譯英文原文
Tensor
Tensorsaregeometricobjectsthatdescribelinear
relationsbetweenvectors,scalars,andothertensors.
Elementaryexamplesofsuchrelationsincludethe
dotproduct,thecrossproduct,andlinear
maps.Vectorsandscalarsthemselvesarealsotensors.
Atensorcanberepresentedasamulti-dimensional
arrayofnumericalvalues.Theorder(alsodegreeor
rank)ofatensoristhedimensionalityofthearray
neededtorepresentit,orequivalently,thenumberof
indicesneededtolabelacomponentofthatarray.Forexample,alinearmapcanberepresentedbyamatrix,a2-dimensionalarray,andthereforeisa2nd-ordertensor.Avectorcanberepresentedasa1-dimensionalarrayandisa
1st-ordertensor.Scalarsaresinglenumbersand
arethus0th-ordertensors.
Tensorsareusedtorepresentcorrespondencesbetweensetsofgeometricvectors.Forexample,theCauchystresstensorTtakesadirectionvasinputandproducesthestressT(v)onthesurface
normaltothisvectorforoutputthusexpressing
arelationshipbetweenthesetwovectors,showninthefigure(right).
Becausetheyexpressarelationshipbetweenvectors,tensorsthemselvesmustbe
independentofaparticularchoiceofcoordinatesystem.Takingacoordinatebasisorframeofreferenceandapplyingthetensortoitresultsinanorganizedmultidimensionalarrayrepresentingthetensorinthatbasis,orframeofreference.Thecoordinateindependenceofatensorthentakestheformofa"covariant"transformationlawthatrelatesthearraycomputedinonecoordinatesystemtothatcomputedinanotherone.Thistransformationlawisconsideredtobebuiltintothenotionofatensorinageometricorphysicalsetting,andthepreciseformofthetransformationlawdeterminesthetype(orvalence)ofthetensor.
Tensorsareimportantinphysicsbecausetheyprovideaconcisemathematicalframeworkforformulatingandsolvingphysicsproblemsinareassuchaselasticity,fluidmechanics,andgeneralrelativity.TensorswerefirstconceivedbyTullioLevi-CivitaandGregorioRicci-Curbastro,whocontinuedtheearlierworkofBernhardRiemannandElwinBrunoChristoffelandothers,aspartoftheabsolutedifferentialcalculus.TheconceptenabledanalternativeformulationoftheintrinsicdifferentialgeometryofamanifoldintheformoftheRiemanncurvaturetensor.[1]Cauchystresstensor,asecond-ordertensor.Thetensor'scomponents,inathree-dimensionalCartesiancoordinatesystem,formthematrixwhosecolumnsarethestresses(forcesperunitarea)actingonthee1,e2,ande3facesofthecube.
History
TheconceptsoflatertensoranalysisarosefromtheworkofCarlFriedrichGaussindifferentialgeometry,andtheformulationwasmuchinfluencedbythetheoryofalgebraicformsandinvariantsdevelopedduringthemiddleofthenineteenthcentury.[2]Theword"tensor"itselfwasintroducedin1846byWilliamRowanHamilton[3]todescribesomethingdifferentfromwhatisnowmeantbyatensor.[Note1]ThecontemporaryusagewasbroughtinbyWoldemarVoigtin1898.[4]
Tensorcalculuswasdevelopedaround1890byGregorioRicci-Curbastrounderthetitleabsolutedifferentialcalculus,andoriginallypresentedbyRicciin1892.[5]ItwasmadeaccessibletomanymathematiciansbythepublicationofRicciandTullioLevi-Civita's1900classictextMéthodesdecalculdifférentielabsoluetleursapplications(Methodsofabsolutedifferentialcalculusandtheirapplications).[6]
Inthe20thcentury,thesubjectcametobeknownastensoranalysis,andachievedbroaderacceptancewiththeintroductionofEinstein'stheoryofgeneralrelativity,around1915.Generalrelativityisformulatedcompletelyinthelanguageoftensors.Einsteinhadlearnedaboutthem,withgreatdifficulty,fromthegeometerMarcelGrossmann.[7]Levi-CivitatheninitiatedacorrespondencewithEinsteintocorrectmistakesEinsteinhadmadeinhisuseoftensoranalysis.Thecorrespondencelasted1915–17,andwascharacterizedbymutualrespect:
Iadmiretheeleganceofyourmethodofcomputation;itmustbenicetoridethroughthesefieldsuponthehorseoftruemathematicswhilethelikeofushavetomakeourwaylaboriouslyonfoot.
—AlbertEinstein,TheItalianMathematiciansofRelativity[8]
Tensorswerealsofoundtobeusefulinotherfieldssuchascontinuummechanics.Somewell-knownexamplesoftensorsindifferentialgeometryarequadraticformssuchasmetrictensors,andtheRiemanncurvaturetensor.TheexterioralgebraofHermannGrassmann,fromthemiddleofthenineteenthcentury,isitselfatensortheory,andhighlygeometric,butitwassometimebeforeitwasseen,withthetheoryofdifferentialforms,asnaturallyunifiedwithtensorcalculus.TheworkofélieCartanmadedifferentialformsoneofthebasickindsoftensorsusedinmathematics.Fromaboutthe1920sonwards,itwasrealisedthattensorsplayabasicroleinalgebraictopology(forexampleintheKünneththeorem).[citationneeded]Correspondinglytherearetypesoftensorsatworkinmanybranchesofabstractalgebra,particularlyinhomologicalalgebraandrepresentationtheory.Multilinearalgebracanbedevelopedingreatergeneralitythanforscalarscomingfromafield,butthetheoryisthencertainlylessgeometric,andcomputationsmoretechnicalandlessalgorithmic.[clarificationneeded]Tensorsaregeneralizedwithincategorytheoryby
meansoftheconceptofmonoidalcategory,fromthe1960s.
Definition
Thereareseveralapproachestodefiningtensors.Althoughseeminglydifferent,theapproachesjustdescribethesamegeometricconceptusingdifferentlanguagesandatdifferentlevelsofabstraction.
Asmultidimensionalarrays
Justasascalarisdescribedbyasinglenumber,andavectorwithrespecttoagivenbasisisdescribedbyanarrayofonedimension,anytensorwithrespecttoabasisisdescribedbyamultidimensionalarray.Thenumbersinthearrayareknownasthescalarcomponentsofthetensororsimplyitscomponents.Theyaredenotedbyindicesgivingtheirpositioninthearray,insubscriptandsuperscript,afterthesymbolicnameofthetensor.Thetotalnumberofindicesrequiredtouniquelyselecteachcomponentisequaltothedimensionofthearray,andiscalledtheorderortherankofthetensor.[Note2]Forexample,theentriesofanorder2tensorTwouldbedenotedTij,whereiandjareindicesrunningfrom1tothedimensionoftherelatedvectorspace.[Note3]
Justasthecomponentsofavectorchangewhenwechangethebasisofthevectorspace,theentriesofatensoralsochangeundersuchatransformation.Eachtensorcomesequippedwithatransformationlawthatdetailshowthecomponentsofthetensorrespondtoachangeofbasis.Thecomponentsofavectorcanrespondintwodistinctwaystoachangeofbasis(seecovarianceandcontravarianceofvectors),
wherethenewbasisvectorsareexpressedintermsoftheoldbasisvectorsas,
whereRijisamatrixandinthesecondexpressionthesummationsignwassuppressed(anotationalconvenienceintroducedbyEinsteinthatwillbeusedthroughoutthisarticle).Thecomponents,vi,ofaregular(orcolumn)vector,v,transformwiththeinverseofthematrixR,
wherethehatdenotesthecomponentsinthenewbasis.Whilethecomponents,wi,ofacovector(orrowvector),wtransformwiththematrixRitself,
Thecomponentsofatensortransforminasimilarmannerwithatransformationmatrixforeachindex.Ifanindextransformslikeavectorwiththeinverseofthebasistransformation,itiscalledcontravariantandistraditionallydenotedwithanupperindex,whileanindexthattransformswiththebasistransformationitselfiscalledcovariantandisdenotedwithalowerindex.Thetransformationlawforanorder-mtensorwithncontravariantindicesandm?ncovariantindicesisthusgivenas,
Suchatensorissaidtobeoforderortype(n,m?n).[Note4]Thisdiscussionmotivatesthefollowingformaldefinition:[9]
Definition.Atensoroftype(n,m?n)isanassignmentofamultidimensionalarray
toeachbasisf=(e1,...,eN)suchthat,ifweapplythechangeofbasis
thenthemultidimensionalarrayobeysthetransformationlaw
ThedefinitionofatensorasamultidimensionalarraysatisfyingatransformationlawtracesbacktotheworkofRicci.[1]Nowadays,thisdefinitionisstillusedinsomephysicsandengineeringtextbooks.[10][11]
Tensorfields
Mainarticle:Tensorfield
Inmanyapplications,especiallyindifferentialgeometryandphysics,itisnaturaltoconsideratensorwithcomponentswhicharefunctions.Thiswas,infact,thesettingofRicci'soriginalwork.Inmodernmathematicalterminologysuchanobjectiscalledatensorfield,buttheyareoftensimplyreferredtoastensorsthemselves.[1]
Inthiscontextthedefiningtransformationlawtakesadifferentform.The"basis"forthetensorfieldisdeterminedbythecoordinatesoftheunderlyingspace,andthe
definingtransformationlawisexpressedintermsofpartialderivativesofthe
coordinatefunctions,,definingacoordinatetransformation,[1]
Asmultilinearmaps
Adownsidetothedefinitionofatensorusingthemultidimensionalarrayapproachisthatitisnotapparentfromthedefinitionthatthedefinedobjectisindeedbasisindependent,asisexpectedfromanintrinsicallygeometricobject.Althoughitispossibletoshowthattransformationlawsindeedensureindependencefromthebasis,sometimesamoreintrinsicdefinitionispreferred.Oneapproachistodefineatensorasamultilinearmap.Inthatapproachatype(n,m)tensorTisdefinedasamap,
whereVisavectorspaceandV*isthecorrespondingdualspaceofcovectors,whichislinearineachofitsarguments.
ByapplyingamultilinearmapToftype(n,m)toabasis{ej}forVandacanonicalcobasis{εi}forV*,
ann+mdimensionalarrayofcomponentscanbeobtained.Adifferentchoiceofbasiswillyielddifferentcomponents.But,becauseTislinearinallofitsarguments,thecomponentssatisfythetensortransformationlawusedinthemultilineararraydefinition.ThemultidimensionalarrayofcomponentsofTthusformatensoraccordingtothatdefinition.Moreover,suchanarraycanberealisedasthecomponentsofsomemultilinearmapT.Thismotivatesviewingmultilinearmapsastheintrinsicobjectsunderlyingtensors.
Usingtensorproducts
Mainarticle:Tensor(intrinsicdefinition)
Forsomemathematicalapplications,amoreabstractapproachissometimesuseful.Thiscanbeachievedbydefiningtensorsintermsofelementsoftensorproductsofvectorspaces,whichinturnaredefinedthroughauniversalproperty.Atype(n,m)tensorisdefinedinthiscontextasanelementofthetensorproductofvector
spaces,[12]
IfviisabasisofVandwjisabasisofW,thenthetensorproducthasa
naturalbasis.ThecomponentsofatensorTarethecoefficientsofthetensorwithrespecttothebasisobtainedfromabasis{ei}forVanditsdual{εj},i.e.
Usingthepropertiesofthetensorproduct,itcanbeshownthatthesecomponentssatisfythetransformationlawforatype(m,n)tensor.Moreover,theuniversalpropertyofthetensorproductgivesa1-to-1correspondencebetweentensorsdefinedinthiswayandtensorsdefinedasmultilinearmaps.
Operations
Thereareanumberofbasicoperationsthatmaybeconductedontensorsthatagainproduceatensor.Thelinearnatureoftensorimpliesthattwotensorsofthesametypemaybeaddedtogether,andthattensorsmaybemultipliedbyascalarwithresultsanalogoustothescalingofavector.Oncomponents,theseoperationsaresimplyperformedcomponentforcomponent.Theseoperationsdonotchangethetypeofthetensor,howevertherealsoexistoperationsthatchangethetypeofthetensors.
Raisingorloweringanindex
Mainarticle:Raisingandloweringindices
Whenavectorspaceisequippedwithaninnerproduct(ormetricasitisoftencalledinthiscontext),operationscanbedefinedthatconvertacontravariant(upper)indexintoacovariant(lower)indexandviceversa.Ametricitselfisa(symmetric)(0,2)-tensor,itisthuspossibletocontractanupperindexofatensorwithoneoflowerindicesofthemetric.Thisproducesanewtensorwiththesameindexstructureastheprevious,butwithlowerindexinthepositionofthecontractedupperindex.Thisoperationisquitegraphicallyknownasloweringanindex.
Converselythematrixinverseofthemetriccanbedefined,whichbehavesasa(2,0)-tensor.Thisinversemetriccanbecontractedwithalowerindextoproduceanupperindex.Thisoperationiscalledraisinganindex.
Applications
Continuummechanics
Importantexamplesareprovidedbycontinuummechanics.Thestressesinsideasolidbodyorfluidaredescribedbyatensor.Thestresstensorandstraintensorarebothsecondordertensors,andarerelatedinagenerallinearelasticmaterialbyafourth-orderelasticitytensor.Indetail,thetensorquantifyingstressina3-dimensionalsolidobjecthascomponentsthatcanbeconvenientlyrepresentedasa3×3array.Thethreefacesofacube-shapedinfinitesimalvolumesegmentofthesolidareeachsubjecttosomegivenforce.Theforce'svectorcomponentsarealsothreeinnumber.Thus,3×3,or9componentsarerequiredtodescribethestressatthiscube-shapedinfinitesimalsegment.Withintheboundsofthissolidisawholemassofvaryingstressquantities,eachrequiring9quantitiestodescribe.Thus,asecondordertensorisneeded.
Ifaparticularsurfaceelementinsidethematerialissingledout,thematerialononesideofthesurfacewillapplyaforceontheotherside.Ingeneral,thisforcewillnotbeorthogonaltothesurface,butitwilldependontheorientationofthesurfaceinalinearmanner.Thisisdescribedbyatensoroftype(2,0),inlinearelasticity,ormorepreciselybyatensorfieldoftype(2,0),sincethestressesmayvaryfrompointtopoint.
Otherexamplesfromphysics
Commonapplicationsinclude
?Electromagnetictensor(orFaraday'stensor)inelectromagnetism
?Finitedeformationtensorsfordescribingdeformationsandstraintensorforstrainincontinuummechanics
?Permittivityandelectricsusceptibilityaretensorsinanisotropicmedia
?Four-tensorsingeneralrelativity(e.g.stress-energytensor),usedtorepresentmomentumfluxes
?Sphericaltensoroperatorsaretheeigenfunctionsofthequantumangularmomentumoperatorinsphericalcoordinates
?Diffusiontensors,thebasisofDiffusionTensorImaging,representratesofdiffusioninbiologicenvironments
?QuantumMechanicsandQuantumComputingutilisetensorproductsforcombinationofquantumstates
Applicationsoftensorsoforder>2
Theconceptofatensorofordertwoisoftenconflatedwiththatofamatrix.Tensorsofhigherorderdohowevercaptureideasimportantinscienceandengineering,ashasbeenshownsuccessivelyinnumerousareasastheydevelop.Thishappens,forinstance,inthefieldofcomputervision,withthetrifocaltensorgeneralizingthefundamentalmatrix.
Thefieldofnonlinearopticsstudiesthechangestomaterialpolarizationdensityunder
extremeelectricfields.Thepolarizationwavesgeneratedarerelatedtothegeneratingelectricfieldsthroughthenonlinearsusceptibilitytensor.IfthepolarizationPisnotlinearlyproportionaltotheelectricfieldE,themediumistermednonlinear.Toagoodapproximation(forsufficientlyweakfields,assumingnopermanentdipolemomentsarepresent),PisgivenbyaTaylorseriesinEwhosecoefficientsarethenonlinearsusceptibilities:
Hereisthelinearsusceptibility,givesthePockelseffectandsecond
harmonicgeneration,andgivestheKerreffect.Thisexpansionshowsthewayhigher-ordertensorsarisenaturallyinthesubjectmatter.
Generalizations[edit]
Tensorsininfinitedimensions
Thenotionofatensorcanbegeneralizedinavarietyofwaystoinfinitedimensions.One,forinstance,isviathetensorproductofHilbertspaces.[15]Anotherwayofgeneralizingtheideaoftensor,commoninnonlinearanalysis,isviathemultilinearmapsdefinitionwhereinsteadofusingfinite-dimensionalvectorspacesandtheiralgebraicduals,oneusesinfinite-dimensionalBanachspacesandtheircontinuousdual.[16]TensorsthuslivenaturallyonBanachmanifolds.[17]
Tensordensities
Mainarticle:Tensordensity
Itisalsopossibleforatensorfieldtohavea"density".Atensorwithdensityrtransformsasanordinarytensorundercoordinatetransformations,exceptthatitisalsomultipliedbythedeterminantoftheJacobiantotherthpower.[18]Invariantly,inthelanguageofmultilinearalgebra,onecanthinkoftensordensitiesasmultilinearmapstakingtheirvaluesinadensitybundlesuchasthe(1-dimensional)spaceofn-forms(wherenisthedimensionoft
溫馨提示
- 1. 本站所有資源如無特殊說明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請下載最新的WinRAR軟件解壓。
- 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
- 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁內(nèi)容里面會有圖紙預覽,若沒有圖紙預覽就沒有圖紙。
- 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
- 5. 人人文庫網(wǎng)僅提供信息存儲空間,僅對用戶上傳內(nèi)容的表現(xiàn)方式做保護處理,對用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對任何下載內(nèi)容負責。
- 6. 下載文件中如有侵權(quán)或不適當內(nèi)容,請與我們聯(lián)系,我們立即糾正。
- 7. 本站不保證下載資源的準確性、安全性和完整性, 同時也不承擔用戶因使用這些下載資源對自己和他人造成任何形式的傷害或損失。
最新文檔
- 2025版物業(yè)公司崗位安全責任書:設(shè)備運行安全協(xié)議(2025年度)3篇
- 二零二五年度綠色環(huán)保型小區(qū)物業(yè)承包服務合同范本3篇
- 2020中國郵政儲蓄銀行招聘試題及答案解析
- 二零二五年度個人股權(quán)退股協(xié)議范本:企業(yè)股權(quán)收購與退股合同3篇
- 二零二五年包裝設(shè)計合同終止及市場適應性協(xié)議2篇
- 區(qū)人大代表述職報告范文4篇
- 小學一年級上冊數(shù)學教學反思5篇
- 贊助合同模板五篇
- 汽車試用買賣合同范本
- 合作拍攝故事片協(xié)議書
- 2025至2030年中國減肥肽數(shù)據(jù)監(jiān)測研究報告
- 2024內(nèi)蒙古公務員省直行測、行政執(zhí)法、省考行測考試真題(5套)
- 2025年安徽馬鞍山市兩山綠色生態(tài)環(huán)境建設(shè)有限公司招聘筆試參考題庫附帶答案詳解
- 山東省濱州市濱城區(qū)2024-2025學年九年級上學期期末考試化學試題
- 貨運企業(yè)2025年度安全檢查計劃
- 北京房地產(chǎn)典當合同
- PHILIPS HeartStart XL+操作培訓課件
- 檔案工作管理情況自查表
- 蘇科版九年級(初三)物理下冊全套課件
- 100個超高難度繞口令大全
- 畢業(yè)論文-基于51單片機的智能LED照明燈的設(shè)計
評論
0/150
提交評論