Professional Documents
Culture Documents
VectorCalculus
Reading:
Kreyszig,AdvancedEngineeringMathematics,10thEd.,2011
Selectionfromchapters9and10
Prerequisites:
Kreyszig,AdvancedEngineeringMathematics,10thEd.,2011
1.Vectorquantities.Dotandcrossproducts:Sections9.19.3
2. Doubleintegral:Section10.3
2.Tripleintegral:Section10.7
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 1
Contents
1. Vector physical quantities.
2. Vector calculus: Motivation and applications
3. Linear vector space. Dot, cross, and triple products
4. Scalar and vector fields and their derivatives
5. Curves
6. Surfaces
7. Line integrals of scalar and vector fields
8. Gradient, divergence, and curl
9. Pathindependent line integrals
10. Double integrals
11. Triple integrals
12. Surface integrals of scalar and vector fields
13. Greens theorem for a plane
14. Stokess theorem
15. Gauss divergence theorem
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 2
6.1Vectorphysicalquantities
Notion of a vector is inspired by the existence of physical quantities that are characterized by
both magnitude and direction, e.g., velocity and force. In physics
The scalar is a quantity that is determined by its magnitude (temperature).
The vector is a quantity that is represented by a directed line segment, and thus, has
magnitude | | (absolute value, length, or norm of the vector) and direction (velocity,
Force, angular momentum).
Two vectors and are equal to each other, = , if they have the same length and
direction even if their initial and terminal points are different. We say that any vector is a free
vector if translation (displacement without rotation) does not change the vector.
Two basic operations on vector physical quantities are the vector addition and
scalar multiplication . Since physical vectors are defined geometrically, we introduce
such operations also geometrically as
: | |
Vector addition Scalar multiplication
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 3
6.1Vectorphysicalquantities
And assume that they posses the following properties
, , , 0
(6.1.1)
, , , 1
Here:
is the zero vector, | | 0.
has the same length as , but opposite direction.
Existence of allows one to determine the vector subtraction: .
The unit vector is a vector of unit length, | | 1.
A whole set of objects (vectors) on which we can perform vector addition and scalar
multiplication with properties given by Eqs. (6.1.1) is called the (linear) vector space.
The idea behind the vector calculus is to utilize vectors and their functions for analytical
calculations.
This is possible if any vector is completely represented it terms of numbers, not directed line
segments.
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 4
6.2Vectorcalculus:Motivationandapplications
Motivation:Calculationofworktypequantities
Work=forcexdistance
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov
6.2Vectorcalculus:Motivationandapplications
Motivation:Calculationoffluxtypequantities
Flux of a physical quantity is the amount of
this quantity transferred through a given area
(surface) per unit time
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov
6.2Vectorcalculus:Motivationandapplications
Applications of the vector calculus: All science and engineering fields where problems are
formulated in terms of PDEs or require analysis of vector fields in multidimensional spaces.
Formulation of physical laws in terms of scalar, vector, and tensor fields.
General mathematical properties of such mathematical models .
Fluid mechanics and gas dynamics, combustion
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 7
6.2Vectorcalculus:Motivationandapplications
Solid mechanics Electromagnetic theory
(6.3.1) cos , =
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 9
6.3Linearvectorspace.Dot,cross,andtripleproducts
Based on observations and our experience we can postulate that the physical space in the
classical mechanics is Euclidean, i.e. we can introduce the Cartesian coordinate system, which is
composed of three mutually perpendicular axis , , and . Next we can fix an arbitrary
linear scale and introduce three unit vectors , , , directed along every axis, which we call the
(Cartesian) basis.
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 10
6.3Linearvectorspace.Dot,cross,andtripleproducts
Based on the Pythagorean theorem we see that the
norm or length of the vector is now defined as
(6.3.3)
(6.3.8)
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 13
6.3Linearvectorspace.Dot,cross,andtripleproducts
The cross (vector) product of vectors and
is the vector that
1. Is equal to zero vector if | | 0 or
| | 0.
2. Is orthogonal to and .
3. Has the length
| sin | | sin |=areaofparallelogram
where is the angle between and . Righthanded Lefthanded
4. Has such direction that , , and Cartesiancoordinates Cartesiancoordinates
is the righthanded triplet, i.e. rotation
from to from the top of occurs in the
counterclockwise direction.
Then
, ,
... ...
(6.3.10)
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 14
6.3Linearvectorspace.Dot,cross,andtripleproducts
Note:Allpropertiesofthevectorproductcanbederivedfromthedefinition.Itisimportantthat
,
The triple product , , of vectors , , and
is the scalar calculated as
, ,
(6.3.11)
cos
where , , arecoordinatesandthescalar(dot)productis
Relative position of point with respect to point is characterized by the relative position
vector and distance
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 16
6.4Scalarandvectorfieldsandtheirderivatives
Letsconsidersomedomain ,whichcanbe
1. Onedimensional(1D):setofpointsatthereal axis
2. Twodimensional(2D):Setofpointsin plane
3. Threedimensional(3D):Setofpointsin space.
WeassumethatweintroduceaCartesiancoordinates withtheglobalbasis , ,
Ineverycasewecancharacterizeapoint of withtheposition(radius)vector
In3D,wewillconsideronlyrighthandcoordinates
1D2D3D
Example:Sphericallysymmetricgravitationalfield
The potential energy of a point mass in the spherically symmetric field of mass with center
in point , , is equal to
, ,
Isosurfaces aresphereswithcenterat , , .
Contourlinesat planearecircleswithcentersat , .
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 18
6.4Scalarandvectorfieldsandtheirderivatives
Now lets assume that in every point of we define a vector . Then is called
the vector field:
(6.4.2)
1D:
2D: , , ,
3D: , , , , , , , ,
The curve at every point of which the vector field is tangent to the curve is called the field line
of the vector field . Field lines are used for visualization of vector fields.
Example 1: Potential flow field around a cylinder in a crossflow.
, are polar coordinates ( cos , sin )
Scalar field: Velocity potential
, cos
2
Vectorfield:Fluidvelocity
1 sin cos cos sin
Whitecurves:Contourlinesofthevelocitypotentials
Blackcurves:Fieldlinesofthevelocityfields.
Influidmechanics,fieldlinesofthevelocityfieldarecalledthestreamlines.
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 19
6.4Scalarandvectorfieldsandtheirderivatives
Example2:ScalartemperaturefieldExample3:Vectorvelocityfieldintheflowover
acircularcylinder
Localvelocityvectors
Fieldlinesofvelocityfield=streamlines
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 20
6.4Scalarandvectorfieldsandtheirderivatives
Example4:Electricfieldlinesofadipole Example5:MagneticfieldlinesovertheSun
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 21
6.4Scalarandvectorfieldsandtheirderivatives
Wecandefinederivativesofscalarandvectorfields,e.g.,withrespectto
, , , ,
lim (6.4.3)
, , , ,
lim
, , , , , , , , , , , ,
lim
(6.4.4)
Many rules valid for ordinary derivatives also hold for derivatives of the dot and cross products,
in particular
Thesepropertiesremainvalidbecause
arerepresentedintheformofsums,
whereeverytermisaproductoftwo
functions,e.g. .
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 22
6.5Curves
Lets consider a 3D space with Cartesian coordinates and a vector in this space, which
every coordinate is a function of some parameter :
, (6.5.1)
is called the vector function of parameter . Functions , , and are called the
coordinate functions of the vector function .
If is the position vector, then Eq. (6.5.1) determines a set of point in 3D space that form a
curve. In this case Eq. (6.5.1) is called the parametric representation of a curve.
Note: In 2D space, e.g., on the plane , we can use the
parametric representation of a plane (2D) curve in the form
Plane curve is a particular case of twisted (3D) curves.
Curve given by Eq. (6.5.1) is called oriented since an
increase of the parameter defines the direction of motion
along , is the initial point, is the terminal point.
Example 1: Parametric representation of a straight line in 3D
,
i.e. differential of the parameter corresponds to the
increment of the arc length called the linear
element of .
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 25
6.5Curves
Since is an increasing function, then can be used as another parameter. For this purpose,
we can resolve equation with respect to ). Then :
, 0
where , etc. For every curve there is a specific parameterization where
parameter is the arc length. For this parameterization
where
1
Thus,
| | 1
i.e. if the parameter is the arc length, the tangent vector is the unit tangent vector .
Theorem: If a vector function has unit length, 1, then the tangent vector is
orthogonal to , i.e. 0.
Proof: Assume 1. Then differentiation of the this equation results in:
2 2 2
0
2
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 26
6.5Curves(Optional)
Hereallderivativesarecalculated
for ,i.e. ,
Consequence:
If the parameter is the arc length then
0
If | | 0, then is called the principal normal
vector, /| | is called the unit principal
normal vector, and is called the unit
binormal vector.
Vectors , , and form the local trihedron of
mutually orthogonal vectors in a point of a curve.
Planes , , , , and , are called
osculating, normal, and rectifying planes.
Twomajornumericalpropertiesofatwistedcurvearethecurvature andtorsion :
| | | | (6.5.4)
Note: is parallel to , since , i.e. is orthogonal
to both and .
Meaning of curvature: It shows how fast the curve deviates from its tangent direction in the
point (for a straight line and 0; 1/ the radius of a circle).
Meaning of torsion: It shows how fast the curve deviates from a plane curve in the osculating
plane (for a plane curve and 0).
Note:Withexceptionofthefollowingexample,wewillnotusethecurvatureandtorsion.
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 27
6.5Curves(Optional)
Example: Properties of a helix curve.
Parametric representation of the helix curve around axis of radius and pitch :
cos sin
2
sin cos
2
| |
2
cos sin
2
sin cos
2
cos sin sin cos
2 2
cos sin cos sin
| | 2 2
Curvature and torsion of the helix curve:
1 1 /2
,
1
2 2
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 28
6.6Surfaces
Thesurface in3Dspaceisasetofpointsthatcanbegivenbyanequation
, , 0 (6.6.1)
Example: is the sphere of radius with the center in the
point , , .
Locally, we can resolve Eq. (6.6.1) with respect to one of coordinates , , and obtain, e.g.
, (6.6.2)
For instance, at any point of the sphere, where , we can write
Equation in the form (6.6.2) may be not valid, however, for the whole surface, since given and
may correspond to a few different at the surface (e.g., for a sphere given and correspond
to two points on the sphere surface).
, ,
On the other hand, Eq. (6.6.2) shows that locally
the surface can be represented parametrically as
(6.6.3) , ,
where , denotes a radius vector of a point at ,
the surface. Eq. (6.6.3) shows that a surface is a
twodimensional set of points, since we need two
parameters ( and ) to describe it parametrically. ,
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 29
6.6Surfaces
In general, parameters do not need to coincide with Cartesian coordinates, so one can introduce
the general parametric representation of a surface in the form
(6.6.4) , , , ,
where and vary in some region on plane.
Eq. (6.6.4) describes the parametric mapping of the plane region into
3D surface.
Mapping
In order to define any surface parametrically, we need to
1. Introduce three functions: , , , , and , .
2. Define region on plane where and varies.
Example: Parametric representation of a cylinder and a sphere
0 2 0 2
0
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 33
6.7Lineintegralsofscalarandvectorfields
Lets consider an oriented curve given by the parametric
representation:
, a
We assume that is a smooth curve, so / exists
in a any point of and the linear element is
.
The line integral of a scalar field over the curve is
the integral
, , (6.7.1)
The line integral of a vector field over the curve is the integral
(6.7.2)
The curve is called the path of integration. If , then is the closed path and the
line integral of a vector field is called circulation of the vector field and denoted as .
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 34
6.7Lineintegralsofscalarandvectorfields
Note: Direction that corresponds to increasing parameter is called positive.
Example: Lets consider motion of a point mass in a force field (e.g. gravitational field).
Then the line integral over the trajectory is the work done by
force .
Properties of the line integral:
1. Linearity
2. Additivity
3. Additivity allows one to extent definition of the line integral to any piecewise smooth curve
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 35
6.8Lineintegralsofscalarandvectorfields
Theorem:
For a given vector field and curve , all parametric representations of curve that give the
same positive direction on yield the same value of the line integral.
Proof:
Lets consider another parameterization of curve with some parameter :
,
. Since it is the representation of the same curve, . Then
Example 1: 0 0 . Lets calculate line integral over a straight path.
Parametric representation of the path:
, ,
0 1
, ,
2
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 36
6.7Lineintegralsofscalarandvectorfields
Example 2: , Line integral over the helix curve. Radius 1
Parametric representation of a helix curve around axis : Pitch 6
cos sin 3 , 0 2
sin cos 3
3 cos sin t sin cos 3
3 sin cos 3 sin
(6.8.2)
0
lim lim
The directional derivative indicates the rate of change of the scalar field in the direction of
vector . Lets calculate through partial derivatives / etc. using the chain rule:
(6.8.3)
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 38
6.8Gradient,Divergence,andCurl
Note: Gradient and directional derivative are easily applied to 2D fields:
grad , (6.8.4)
0 0
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 40
6.8Gradient,Divergence,andCurl
Note: In 2D, the contour lines of the scalar field , and field lines of its gradient
constitute two families of curves that are locally normal to each other in every point.
Example 2: Electric field of a dipole ,
Contourlinesoftheelectrostatic
Potential
Fieldlinesoftheelectricalfield
Note: Two families of locally orthogonal curves can be used in order to introduce nonCartesian
orthogonal coordinates in 2D.
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 41
6.8Gradient,Divergence,andCurl
Example 3: Potential fluid flow over a cylinder in the cross flow
cos , sin
, cos
2 Contourlinesoftheflow
potential white
Fieldlinesofthevelocityfield
(streamlines;black)
Thesecurvescanbeusedinordertointroduceorthogonalcoordinates.
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 42
6.8Gradient,Divergence,andCurl
Example 4: Application of the gradient in the optimization problem
Optimization in , space = finding a maximum of a goal function
,
1. Plotcontourlines ,
2. grad isnormaltocontourlines
3. grad correspondstothe
directionofthefastestincreaseof
thegoalfunction
4. Stepsalonggrad leadtoalocal
maximum
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 43
6.8Gradient,Divergence,andCurl
Divergence
Lets consider a 3D vector field , , , , , , , , . The
divergence of the vector field is the scalar field div , which value is equal to
(6.8.5) div
Properties of divergence:
1. Lets consider a small parallelepiped of volume
. For any face we can define the flux of
vector field as , where is the unit vector
normal to the face and directed outwards and is the
face area. Then the flux through the whole surface is
, , , , , , , ,
, , , , div
Now we see that
div lim (6.8.6)
Divergence div defines the ratio of the flux of the vector field through the surface of an
(infinitely) small domain to the volume of this domain.
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 44
6.7Gradient,Divergence,andCurl
2. Divergence is the physical scalar, i.e. its value does not change at any coordinate
transformation. Proof is based on Eq. (6.7.6), but also can be performed directly by
calculating div after a coordinate transformation.
Example 1: Divergence of a central vector field. The potential energy of a point mass in the
spherically symmetric gravitational field of mass with center in point , , is equal to
, , / , .
Then
3 /
2
Note: Divergence of any central vector field is a central scalar field.
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 45
6.8Gradient,Divergence,andCurl
The divergence is important for models based on the conservation laws: If is a flux density of
some physical quantity, then div is related density of sources or drains of this quantity.
curl (6.7.7)
This definition is valid for righthanded Cartesian coordinates. For lefthanded coordinates, the
definitions includes the negative sign: curl .
Properties of curl:
1. The curl is the physical vector, i.e. it does not change at any coordinate transformation
between any righthanded coordinates.
2. Lets assume that is the velocity field of a rigid body. It is proved in kinematics that the
velocity field in the rigid body with respect to pole at point can be represented in the form
where is the angular velocity vector. Lets use righthanded coordinates where .
Then
1
curl 2 2 curl (6.7.8)
2
0
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 47
6.8Gradient,Divergence,andCurl
Deformable body: Relative positions of points vary in the course of motion
One can show that for arbitrary deformable body (solid, liquid or gaseous), the kinematic
formula can be generalized in the form:
1
curl
2
where is the deformation velocity which describes the change of relative positions of
points in the body due to its deformation. Thus, for any body, rigid or deformable, the curl of
the velocity field describes the part of the motion that corresponds to the rigid rotation.
Rigidbodyrotation, 0 Deformablebodyrotation, 0
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 48
6.8Gradient,Divergence,andCurl
3. Circulation of the vector field on the plane
Lets consider a small rectangle in plane that corresponds to increments and . We can
introduce the circulation of the vector field along the contour of this rectangle:
, , , ,
curl
Nowweseethat
(6.8.9) curl lim
i.e. limit of circulation of a vector field over a small contour to
the area bounded by this contour is equal to the component
of the curl normal to the contour. Righthandsidescrewrule
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 49
6.8Gradient,Divergence,andCurl
Laplacian
The Laplacian of a scalar field , , is
Example:Theheatconductionequationinthemediumwithconstantpropertiesis
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 50
6.8Gradient,Divergence,andCurl
Properties:
Any gradient field is irrotational, i.e. curl grad 0.
Curl produces divergencefree field: div curl 0.
Any central field is a gradient field with the potential | | , .
ME501,MechanicalEngineeringAnalysis,AlexeyVolkov 51