Beruflich Dokumente
Kultur Dokumente
Abstract
Lagrangian or referential equilibrium equations for materials undergoing large
deformations are of interest in the developing fields of mechanics of soft biomaterials and
nanomechanics. The main feature of these equations is the necessity to deal with the First
Piola-Kirchhoff, or nominal, stress tensor which is a two-point tensor referring simultaneously
to the reference and current configurations. This two-point nature of the First Piola-Kirchhoff
tensor is not always appreciated by the researchers and the total covariant derivative
necessary for the formulation of the equilibrium equations in curvilinear coordinates is
sometimes inaccurately confused with the regular covariant derivative. Surprisingly, the
traditional continuum mechanics literature does not discuss this issue properly, except for
some brief notions on the two-point nature of the Piola-Kirchhoff tensor. We aim at partially
filling this gap by giving a full yet simple derivation of the Lagrangian equilibrium equations
in cylindrical and spherical coordinates.
1. Introduction
Lagrangian scalar equilibrium equations in cylindrical and spherical coordinates for
materials undergoing large deformations are rarely discussed in the literature. The most
influential monographs on nonlinear elasticity and continuum mechanics, including Antman
(1995); Chadwick (1976); Ciarlet (1988); Eringen (1962); Green and Adkins (1970); Green
and Zerna (1968); Gurtin (1981); Haupt (2000); Jaunzemis (1967); Liu (2002); Lure (1990);
Malvern (1969); Marsden and Hughes (1983); Ogden (1984); Truesdell and Toupin (1961);
1
Truesdell and Noll (1965); Wang and Truesdell (1973); Wilmanski (1998), do not address this
issue. However, the Lagrangian equilibrium equations in cylindrical and spherical coordinates
can be very useful in solving nonlinear problems analytically or semi-analytically.
Sometimes, it is possible to assume incompressibility of the material what allows for using a
simpler Eulerian description for obtaining some elementary analytical solutions. This is not
the general case, however, where we need the Lagrangian equilibrium equations of the form
DivP = 0
(1)
in cylindrical and spherical coordinates. These equations can be derived from the total
covariant derivative of the 1st Piola-Kirchhoff stress tensor P. Though this way may be
elegant we prefer a more straightforward pedestrian way, which, however, does not require
any knowledge of the general tensor calculus from the reader.
2. Cylindrical coordinates
We introduce orthonormal basis in cylindrical coordinates (Malvern, 1969) for the
reference configuration
K R = (cos , sin ,0)T ; K = ( sin , cos ,0)T ; K Z = (0,0,1)T .
(2)
K
= K R .
(3)
k r
= k ;
k
= k r .
(4)
(5)
P
P
P
KR +
K +
KZ .
R
R
Z
(6)
The plan is to compute the right-hand side of this equation term by term.
We start with the first term on the right hand side of Eq. (6)
P
k
K R
P
K R = rR k r K R + PrR r K R + PrRk r
K +
R
R
R R
R
K
k
Pr
k r K + Pr r K + Prk r
K +
R
R R
R
k r
K Z
PrZ
K Z + PrZ k r
k r K Z + PrZ
K R +
R
R
R
k
K R
PR
k K R + PR K R + PRk
K R +
R
R
R
k
K
P
k K + P K + Pk
K R +
R
R
R
k
K Z
PZ
k K Z + PZ K Z + PZ k
K R +
R
R
R
k
K R
PzR
k z K R + PzR z K R + PzRk z
K R +
R
R
R
k
K
Pz
k z K + Pz z K + Pzk z
K R +
R
R
R
k z
K Z
PzZ
K Z + PzZ k z
k z K Z + PzZ
R
R
R
K R
(7)
where k m K N = k m K TN .
With account of orthonormality of the base vectors we have
P
k
P
P
k
P
k
K R = rR k r + PrR r + R k + PR + zR k z + PzR z .
R
R
R
R
R
R
R
(8)
=
=
+
+
kr ,
R
r R R z R
R
(9)
k z k z r k z k z z
=
+
= 0.
+
R
r R R z R
P
P
PR
P
+
K R = rR PR
k + zR k z .
k r + PrR
R
R
R R
R
R
(10)
Analogously to Eqs. (7)-(10) we calculate the last two terms on the right-hand side of Eq. (6)
P
k
K R
1 P
K = rR k r K R + PrR r K R + PrR k r
K +
R
R
k
K
1 Pr
k r K + Pr r K + Pr k r
K +
R
k
K Z
1 PrZ
k r K Z + PrZ r K Z + PrZ k r
K +
R
k
K R
1 PR
k K R + PR K R + PR k
K +
R
k
K
1 P
k K + P K + P k
K +
R
k
K Z
1 PZ
k K Z + PZ K Z + PZ k
K +
k
K R
1 PzR
k z K R + PzR z K R + PzR k z
K +
R
k
K
1 Pz
k z K + Pz z K + Pz k z
K +
R
k z
K Z
1 PzZ
k z K Z + PzZ
K Z + PzZ k z
R
P
k
P
PrRk r + r k r + Pr r + PR k + k
1
P
K =
k z
Pz
k
R
R
+ PzRk z +
k z + Pz
+ P
(11)
(12)
k r k r r k r k r z
=
k ,
+
+
=
r z
k k r k k z
kr ,
=
+
+
=
r z
(13)
k z k z r k z k z z
= 0,
+
+
=
r z
P
P
P
P
K = rR + r
k r
R
R R R
,
Pr PR P
PzR Pz
+
+
+
+
k +
k z
R R R
R R
(14)
P
k
K R
P
K Z = rR k r K R + PrR r K R + PrRk r
K Z +
Z
Z
Z
Z
K
k
Pr
k r K + Pr r K + Prk r
K Z +
Z
Z
Z
k r
K Z
PrZ
k r K Z + PrZ
K Z + PrZ k r
K Z +
Z
Z
Z
k
K R
PR
k K R + PR K R + PRk
K Z +
Z
Z
Z
k
K
P
k K + P K + Pk
K Z +
Z
Z
Z
k
K Z
PZ
k K Z + PZ K Z + PZ k
K Z +
Z
Z
Z
k
K R
PzR
k z K R + PzR z K R + PzRk z
K Z +
Z
Z
Z
k
K
Pz
k z K + Pz z K + Pzk z
K Z +
Z
Z
Z
k z
K Z
PzZ
K Z + PzZ k z
k z K Z + PzZ
Z
Z
Z
K Z
k
P
k z
k r PZ
P
P
,
+
K Z = rZ k r + PrZ
k + PZ + zZ k z + PzZ
Z
Z
Z
Z
Z
Z
Z
(15)
(16)
k r k r r k r k r z
+
=
k ,
+
=
r Z Z z Z Z
Z
k k r k k z
kr ,
=
+
+
=
Z
r Z Z z Z
Z
(17)
k z k z r k z k z z
+
= 0,
+
=
r Z Z z Z
Z
P
P
P
P
K Z = rZ PZ
k + zZ k z .
k r + Z + PrZ
Z
Z
Z
Z
Z
Z
(18)
Finally, substituting Eqs. (10), (14), and (18) in Eq. (6) we have
PrR Pr P PrZ
P
DivP = rR PR
+
+
+
PZ
k r +
R R R R Z
Z
R
PR Pr PR P PZ
+
+
+
+
+
+ PrZ
PrR
k +
R R R Z
R R
Z
(19)
k z
R R Z
R
3. Spherical coordinates
We introduce orthonormal basis in spherical coordinates (Malvern, 1969) for the reference
configuration
K R = (sin cos , sin sin , cos )T
K = (cos cos , cos sin , sin )T .
K = ( sin , cos ,0)
(20)
By direct calculation we have the following nonzero derivatives of the base vectors
K R
K
K R
= K;
= K R ;
= sin K ;
.
K
K
= cos K ;
= sin K R cos K
(21)
(22)
k r
= k ;
k
k r
= k r ;
= sin k ;
.
k
k
= cos k ;
= sin k r cos k
(23)
(24)
P
P
P
KR +
K +
K .
R
R
RS
(25)
The plan is again to compute the right-hand side of this equation term by term.
We start with
P
k
K R
P
K R = rR k r K R + PrR r K R + PrRk r
K +
R
R
R R
R
k
K
Pr
k r K + Pr r K + Prk r
K +
R
R R
R
k
K
Pr
k r K + Pr r K + Pr k r
K R +
R
R
R
k
K R
PR
k K R + PR K R + PRk
K R +
R
R
R
k
K
P
k K + P K + Pk
K R +
R
R
R
k
K
P
k K + P K + P k
K R +
R
R
R
k
PR
K R
K R +
k K R + PR K R + PRk
R
R
R
k
P
K
K R +
k K + P K + Pk
R
R
R
.
k
P
K
K R
k K + P
K + Pk
R
R
R
(26)
(27)
=
+
+
=
k + s
k
R
r R R R R
R
k k r k k
=
+
+
=
k r + c k ,
R
r R R R
R
R
k
R
(28)
k r k k
+
+
k r c k .
= s
r R R R
R
R
P
sPR
K R = rR PR
k r
R
R
R
R
PR
P
k
+ R + PrR
cPR
+ sPrR
+ cPR
k +
R
R
R
R
R
R
(29)
Analogously to Eqs. (26)-(29) we calculate the last two terms on the right-hand side of Eq.
(25)
1 P
P
k
K R
K = rR k r K R + PrR r K R + PrRk r
K +
R
R
1 Pr
k
K
k r K + Pr r K + Prk r
K +
R
1 Pr
k
K
k r K + Pr r K + Prk r
K +
R
1 PR
k
K R
k K R + PR K R + PR k
K +
R
1 P
K
k
k K + P K + Pk
K +
R
1 P
k
K
k K + P K + Pk
K +
R
k
1 PR
K R
K +
k K R + PR K R + PR k
R
k
1 P
K
K +
k K + P K + Pk
R
k
1 P
K
K
k K + P K + P k
R
P
k
P
PrRk r + r k r + Pr r + PRk + k
P
1
,
K =
P
k
R
R
k
+ PR k + k + P
+ P
(30)
(31)
k r k r r k r k r
=
+
+
=
k + s
k ,
k k r k k
=
+
+
=
kr + c
k ,
(32)
k r k k
+
+
kr c
k ,
= s
r
P
P
P
P
P
k r
K = rR + r
s
R
R
R R R
P
P
P
P
k ,
c
+ R + + r
R
R R R
(33)
P
P
P
P
k
+ c
+ R + + s r
R
R
R R
1 PrR
P
k
K R
K =
k r K R + PrR r K R + PrRk r
K +
RS
RS
1 Pr
k
K
k r K + Pr r K + Prk r
K +
RS
1 Pr
k
K
k r K + Pr r K + Pr k r
K +
RS
1 PR
k
K R
k K R + PR K R + PRk
K +
RS
1 P
k
K
k K + P K + Pk
K +
RS
1 P
k
K
k K + P K + P k
K +
RS
k
1 PR
K R
K +
k K R + PR K R + PR k
RS
k
1 P
K
K +
k K + P K + Pk
RS
k
1 P
K
K
k K + P K + P k
RS
k
P
K =
P
k
RS
RS
P
k
k + P + SPR k + CPk + k + P
+
(34)
(35)
k r k r r k r k r
=
+
+
=
k + s
k ,
k k r k k
=
+
+
=
kr + c
k ,
(36)
k r k k
+
+
kr c
k ,
= s
r
P
P
Pr
CP
P sP
k r
K = rR + r +
RS
RS
RS RS RS
R
P
P
CP
P cP
k .
+ R + + + r
RS
RS RS RS
R
(37)
P
CP
P
sP cP
k
+ R + + + r
+
RS
RS RS RS
R
Finally, substituting Eqs. (29), (33), and (37) in Eq. (25) we have
DivP =
P
P
P
PrR
+ 2 rR + r
sPR
PR
R R R
R
R
R
kr
P CPr Pr
P sP
+
+
R RS
RS RS RS
P
P
P
PR
+ PrR
cPR
+ 2 R + + r
.
R
R
R
R R R
+
k
P CP P
P cP
+ r
+
+
c
R RS
RS RS RS
P
P
PR
+ sPrR
+ cPR
+ 2 R + + s r
R
R
R
R R
R
+
k
P CP P
sPr cP
+
+
+
+
+ c
R RS
RS RS RS
(38)
4. Conclusion
Lagrangian equilibrium equations in cylindrical (Eq. 19) and spherical coordinates (Eq.
38) have been derived in the present work.
5. References
10
11