Divergence of Vector Functions in Curvilinear Coordinates
📂Mathematical PhysicsDivergence of Vector Functions in Curvilinear Coordinates
Theorem
The divergence of the vector function F=F(q1,q2,q3)=F1q^1+F2q^2+F3q^3 in curvilinear coordinates is as follows.
∇⋅F=h1h2h31[∂q1∂(h2h3F1)+∂q2∂(h1h3F2)+∂q3∂(h1h2F3)]
hi is the scale factor.
Cartesian coordinates:
h1=h2=h3=1
∇⋅F=∂x∂Fx+∂y∂Fy+∂z∂Fz
Cylindrical coordinates:
h1=1,h2=ρ,h3=1
∇⋅F=ρ1(∂ρ∂(ρFρ)+∂ϕ∂(Fϕ)+∂z∂(ρFz))=ρ1∂ρ∂(ρFρ)+ρ1∂ϕ∂Fϕ+∂z∂Fz
h1=1,h2=r,h3=rsinθ
∇⋅F=r2sinθ1(∂r∂(r2sinθFr)+∂θ∂(rsinθFθ)+∂ϕ∂(rFϕ))=r21∂r∂(r2Fr)+rsinθ1∂θ∂(sinθFθ)+rsinθ1∂ϕ∂Fϕ
Derivation
In the 3D Cartesian coordinate system, the divergence ∇⋅F of the vector function F told us how F flows. We can get the divergence in the curvilinear coordinates in the same way. Let’s first compute it only in the direction of q1 axis. The amount passing through da1 and da2 can be calculated by the inner product of the two vectors. The computation is the same as in the Cartesian coordinates, so some parts are omitted.
F(q1+dq1,q2,q3)⋅da1F(q1,q2,q3)⋅da2=F1(q1+dq1,q2,q3)h2h3dq2dq3=−F1(q1,q2,q3)h2h3dq2dq3
And the sum of these two is the influx (outflux).
=≈F1(q1+dq1,q2,q3)h2h3dq2dq3−F1(q1,q2,q3)h2h3dq2dq3dq1F1(q1+dq1,q2,q3)h2h3−F1(q1,q2,q3)h2h3dq1dq2dq3∂q1∂(F1h2h3)dq1dq2dq3
In the same way, if calculated for q2 and q3, it is as follows.
∂q2∂(F2h1h3)dq1dq2dq3and∂q3∂(F3h1h2)dq1dq2dq3
Adding these gives the amount of F entering or leaving, and dividing it by the volume dV=h1h2h3dq1dq2dq3 gives the influx (outflux) per unit volume.
=dV1×⟹∂q1∂(F1h2h3)dq1dq2dq3+∂q2∂(F2h1h3)dq1dq2dq3+∂q3∂(F3h1h2)dq1dq2dq3 (∂q1∂(F1h2h3)+∂q2∂(F2h1h3)+∂q3∂(F3h1h2))dq1dq2dq3h1h2h31(∂q1∂(F1h2h3)+∂q2∂(F2h1h3)+∂q3∂(F3h1h2))
■
See also