<?xml version="1.0" encoding="UTF-8"?>

<record version="1" id="1329">
 <title>Cylindrical and Spherical Particle Kinematics</title>
 <name>CylindricalAndSphericalParticleKinematics</name>
 <created>2026-09-28 01:41:10</created>
 <modified>2026-09-28 01:41:10</modified>
 <type>Topic</type>
 <creator id="1" name="bloftin"/>
 <modifier id="1" name="bloftin"/>
 <comment>finally found the tdot issue so reformulated with \frac</comment>
 <author id="1" name="bloftin"/>
 <classification>
	<category scheme="pacs" code="45.05.+x"/>
 </classification>
 <synonyms>
	<synonym concept="Cylindrical and Spherical Particle Kinematics" alias="M01-11"/>
 </synonyms>
 <keywords>
	<term>cylindrical coordinates</term>
	<term>spherical coordinates</term>
	<term>particle kinematics</term>
	<term>curvilinear coordinates</term>
	<term>moving basis vectors</term>
	<term>velocity</term>
	<term>acceleration</term>
 </keywords>
 <preamble>% this is the default PhysicsLibrary preamble.  as your knowledge
% of TeX increases, you will probably want to edit this, but
% it should be fine as is for beginners.

% almost certainly you want these
\usepackage{amssymb}
\usepackage{amsmath}
\usepackage{amsfonts}

% used for TeXing text within eps files
%\usepackage{psfrag}
% need this for including graphics (\includegraphics).
% If you use \includegraphics{figure} with no file extension, PhysicsLibrary
% will prefer figure.png before falling back to figure.eps.
\usepackage{graphicx}
\DeclareGraphicsExtensions{.png,.jpg,.jpeg,.pdf,.eps}
% for neatly defining theorems and propositions
%\usepackage{amsthm}
% making logically defined graphics
%\usepackage{xypic}

% there are many more packages, add them here as you need them

% define commands here</preamble>
 <content>\section*{Cylindrical and Spherical Particle Kinematics}

Cartesian coordinates are often the simplest coordinates for mechanics because the basis vectors are fixed.  Many three-dimensional problems, however, possess axial or spherical geometry.  In those cases cylindrical or spherical coordinates can describe the motion much more naturally.

The complication is that the curvilinear basis vectors generally move with the particle.  Therefore velocity and acceleration are not obtained by differentiating only the scalar coordinates.  One must also differentiate the basis vectors themselves.

This article derives the complete particle velocity and acceleration formulas for cylindrical coordinates

\begin{equation}
(\rho,\phi,z),
\end{equation}

and spherical coordinates

\begin{equation}
(r,\theta,\phi),
\end{equation}

using the convention that $\theta$ is the polar angle measured from the positive $z$ axis and $\phi$ is the azimuth measured in the $xy$ plane.

\section{Why curvilinear basis vectors matter}

In a Cartesian frame,

\begin{equation}
\mathbf r=x\mathbf e_x+y\mathbf e_y+z\mathbf e_z,
\end{equation}

and the unit vectors satisfy

\begin{equation}
\left(d\mathbf e_x\right)/\left(dt\right)=\left(d\mathbf e_y\right)/\left(dt\right)=\left(d\mathbf e_z\right)/\left(dt\right)=0.
\end{equation}

In cylindrical and spherical coordinates, some unit vectors point in directions defined by the particle's current angular coordinates.  When those angles change, the basis rotates.  Thus a derivative such as

\begin{equation}
\left(d\right)/\left(dt\right)\left(\rho\mathbf e_\rho\right)
\end{equation}

must be evaluated with the product rule:

\begin{equation}
\left(d\right)/\left(dt\right)\left(\rho\mathbf e_\rho\right)
=
\left(d\rho\right)/\left(dt\right)\mathbf e_\rho+\rho\left(d\mathbf e_\rho\right)/\left(dt\right).
\end{equation}

The second term is the source of the angular pieces that distinguish curvilinear kinematics from Cartesian kinematics.

\section{Cylindrical coordinates}

Cylindrical coordinates are related to Cartesian coordinates by

\begin{equation}
x=\rho\cos\phi,
\end{equation}

\begin{equation}
y=\rho\sin\phi,
\end{equation}

and

\begin{equation}
z=z.
\end{equation}

Here $\rho$ is the perpendicular distance from the $z$ axis, $\phi$ is the azimuthal angle in the $xy$ plane, and $z$ is the ordinary vertical Cartesian coordinate.

\begin{center}
\includegraphics{M01_11_fig01_cylindrical_geometry.png}

{\small \PMlinkescapetext{Figure 1. Cylindrical coordinates describe a point by distance $\rho$ from the $z$ axis, azimuth $\phi$, and height $z$.  The local basis vectors $\mathbf e_\rho$ and $\mathbf e_\phi$ rotate as $\phi$ changes, while $\mathbf e_z$ remains fixed.}}
\end{center}

The cylindrical unit vectors can be written in the Cartesian basis as

\begin{equation}
\mathbf e_\rho
=
\cos\phi\,\mathbf e_x
+
\sin\phi\,\mathbf e_y,
\end{equation}

\begin{equation}
\mathbf e_\phi
=
-\sin\phi\,\mathbf e_x
+
\cos\phi\,\mathbf e_y,
\end{equation}

and

\begin{equation}
\mathbf e_z=\mathbf e_z.
\end{equation}

The position vector is

\begin{equation}
\boxed{\mathbf r=\rho\mathbf e_\rho+z\mathbf e_z.}
\end{equation}

\section{Cylindrical basis-vector derivatives}

Differentiate $\mathbf e_\rho$ with respect to $\phi$:

\begin{equation}
\left(\partial\mathbf e_\rho\right)/\left(\partial\phi\right)
=
-\sin\phi\,\mathbf e_x
+
\cos\phi\,\mathbf e_y
=
\mathbf e_\phi.
\end{equation}

Likewise,

\begin{equation}
\left(\partial\mathbf e_\phi\right)/\left(\partial\phi\right)
=
-\cos\phi\,\mathbf e_x
-
\sin\phi\,\mathbf e_y
=
-\mathbf e_\rho.
\end{equation}

Since the basis vectors depend on time only through $\phi(t)$,

\begin{equation}
\boxed{\left(d\mathbf e_\rho\right)/\left(dt\right)=\left(d\phi\right)/\left(dt\right)\,\mathbf e_\phi,}
\end{equation}

\begin{equation}
\boxed{\left(d\mathbf e_\phi\right)/\left(dt\right)=-\left(d\phi\right)/\left(dt\right)\,\mathbf e_\rho,}
\end{equation}

and

\begin{equation}
\boxed{\left(d\mathbf e_z\right)/\left(dt\right)=0.}
\end{equation}

\begin{center}
\includegraphics{M01_11_fig02_cylindrical_basis_rotation.png}

{\small \PMlinkescapetext{Figure 2. The cylindrical radial and azimuthal unit vectors rotate through the same small angle as the particle's azimuth. This rotating basis produces the additional terms that appear in cylindrical velocity and acceleration.}}
\end{center}

\section{Cylindrical velocity}

Differentiate

\begin{equation}
\mathbf r=\rho\mathbf e_\rho+z\mathbf e_z.
\end{equation}

Using the product rule,

\begin{equation}
\mathbf v
=
\left(d\rho\right)/\left(dt\right)\mathbf e_\rho
+
\rho\left(d\mathbf e_\rho\right)/\left(dt\right)
+
\left(dz\right)/\left(dt\right)\mathbf e_z.
\end{equation}

Substitute

\begin{equation}
\left(d\mathbf e_\rho\right)/\left(dt\right)=\left(d\phi\right)/\left(dt\right)\mathbf e_\phi
\end{equation}

to obtain

\begin{equation}
\boxed{
\mathbf v
=
\left(d\rho\right)/\left(dt\right)\,\mathbf e_\rho
+
\rho\left(d\phi\right)/\left(dt\right)\,\mathbf e_\phi
+
\left(dz\right)/\left(dt\right)\,\mathbf e_z.
}
\end{equation}

Thus the physical velocity components are

\begin{equation}
v_\rho=\left(d\rho\right)/\left(dt\right),
\qquad
v_\phi=\rho\left(d\phi\right)/\left(dt\right),
\qquad
v_z=\left(dz\right)/\left(dt\right).
\end{equation}

Notice that $\left(d\phi\right)/\left(dt\right)$ itself is an angular rate, not a linear speed.  The azimuthal linear speed is $\rho\left(d\phi\right)/\left(dt\right)$.

\section{Cylindrical acceleration}

Differentiate the cylindrical velocity:

\begin{equation}
\mathbf v
=
\left(d\rho\right)/\left(dt\right)\mathbf e_\rho
+
\rho\left(d\phi\right)/\left(dt\right)\mathbf e_\phi
+
\left(dz\right)/\left(dt\right)\mathbf e_z.
\end{equation}

The derivative of the first term is

\begin{equation}
\left(d\right)/\left(dt\right)(\left(d\rho\right)/\left(dt\right)\mathbf e_\rho)
=
\left(d^2\rho\right)/\left(dt^2\right)\mathbf e_\rho
+
\left(d\rho\right)/\left(dt\right)\left(d\phi\right)/\left(dt\right)\mathbf e_\phi.
\end{equation}

The derivative of the azimuthal term is

\begin{equation}
\left(d\right)/\left(dt\right)(\rho\left(d\phi\right)/\left(dt\right)\mathbf e_\phi)
=
(\left(d\rho\right)/\left(dt\right)\left(d\phi\right)/\left(dt\right)+\rho\left(d^2\phi\right)/\left(dt^2\right))\mathbf e_\phi
+
\rho\left(d\phi\right)/\left(dt\right)\left(d\mathbf e_\phi\right)/\left(dt\right).
\end{equation}

Using

\begin{equation}
\left(d\mathbf e_\phi\right)/\left(dt\right)=-\left(d\phi\right)/\left(dt\right)\mathbf e_\rho,
\end{equation}

gives

\begin{equation}
\left(d\right)/\left(dt\right)(\rho\left(d\phi\right)/\left(dt\right)\mathbf e_\phi)
=
(\left(d\rho\right)/\left(dt\right)\left(d\phi\right)/\left(dt\right)+\rho\left(d^2\phi\right)/\left(dt^2\right))\mathbf e_\phi
-
\rho\left(\left(d\phi\right)/\left(dt\right)\right)^2\mathbf e_\rho.
\end{equation}

Combining terms,

\begin{equation}
\boxed{
\mathbf a
=
(\left(d^2\rho\right)/\left(dt^2\right)-\rho\left(\left(d\phi\right)/\left(dt\right)\right)^2)\mathbf e_\rho
+
(\rho\left(d^2\phi\right)/\left(dt^2\right)+2\left(d\rho\right)/\left(dt\right)\left(d\phi\right)/\left(dt\right))\mathbf e_\phi
+
\left(d^2 z\right)/\left(dt^2\right)\mathbf e_z.
}
\end{equation}

The radial component is

\begin{equation}
\boxed{a_\rho=\left(d^2\rho\right)/\left(dt^2\right)-\rho\left(\left(d\phi\right)/\left(dt\right)\right)^2,}
\end{equation}

and the azimuthal component is

\begin{equation}
\boxed{a_\phi=\rho\left(d^2\phi\right)/\left(dt^2\right)+2\left(d\rho\right)/\left(dt\right)\left(d\phi\right)/\left(dt\right).}
\end{equation}

The term $-\rho\left(\left(d\phi\right)/\left(dt\right)\right)^2$ is the inward centripetal contribution.  The term $2\left(d\rho\right)/\left(dt\right)\left(d\phi\right)/\left(dt\right)$ appears because the radial distance and the rotating basis change simultaneously.

\section{Polar-coordinate motion as a cylindrical special case}

If the motion remains in a plane of constant $z$, then

\begin{equation}
\left(dz\right)/\left(dt\right)=\left(d^2 z\right)/\left(dt^2\right)=0,
\end{equation}

and cylindrical kinematics reduces directly to planar polar kinematics:

\begin{equation}
\mathbf v
=
\left(d\rho\right)/\left(dt\right)\mathbf e_\rho
+
\rho\left(d\phi\right)/\left(dt\right)\mathbf e_\phi,
\end{equation}

\begin{equation}
\mathbf a
=
(\left(d^2\rho\right)/\left(dt^2\right)-\rho\left(\left(d\phi\right)/\left(dt\right)\right)^2)\mathbf e_\rho
+
(\rho\left(d^2\phi\right)/\left(dt^2\right)+2\left(d\rho\right)/\left(dt\right)\left(d\phi\right)/\left(dt\right))\mathbf e_\phi.
\end{equation}

This is the direct three-dimensional extension of M01-10.

\section{Spherical coordinates and convention}

Spherical-coordinate notation is not universal, so the angular convention must be stated explicitly.  In this article,

\begin{itemize}
\item $r$ is radial distance from the origin;
\item $\theta$ is the polar angle measured downward from the positive $z$ axis;
\item $\phi$ is the azimuth measured in the $xy$ plane from the positive $x$ axis.
\end{itemize}

The Cartesian relations are

\begin{equation}
x=r\sin\theta\cos\phi,
\end{equation}

\begin{equation}
y=r\sin\theta\sin\phi,
\end{equation}

and

\begin{equation}
z=r\cos\theta.
\end{equation}

\begin{center}
\includegraphics{M01_11_fig03_spherical_geometry.png}

{\small \PMlinkescapetext{Figure 3. Spherical-coordinate convention used throughout this article.  The polar angle $\theta$ is measured from $+z$ and the azimuth $\phi$ is measured in the $xy$ plane.  Stating this convention prevents the common interchange of $\theta$ and $\phi$.}}
\end{center}

The spherical basis vectors are

\begin{equation}
\mathbf e_r
=
\sin\theta\cos\phi\,\mathbf e_x
+
\sin\theta\sin\phi\,\mathbf e_y
+
\cos\theta\,\mathbf e_z,
\end{equation}

\begin{equation}
\mathbf e_\theta
=
\cos\theta\cos\phi\,\mathbf e_x
+
\cos\theta\sin\phi\,\mathbf e_y
-
\sin\theta\,\mathbf e_z,
\end{equation}

and

\begin{equation}
\mathbf e_\phi
=
-\sin\phi\,\mathbf e_x
+
\cos\phi\,\mathbf e_y.
\end{equation}

The position vector is especially simple:

\begin{equation}
\boxed{\mathbf r=r\mathbf e_r.}
\end{equation}

\section{Spherical basis-vector derivatives}

Because the spherical basis depends on both $\theta$ and $\phi$, its derivatives are richer than in cylindrical coordinates.

The needed partial derivatives are

\begin{equation}
\left(\partial\mathbf e_r\right)/\left(\partial\theta\right)=\mathbf e_\theta,
\end{equation}

\begin{equation}
\left(\partial\mathbf e_r\right)/\left(\partial\phi\right)=\sin\theta\,\mathbf e_\phi,
\end{equation}

\begin{equation}
\left(\partial\mathbf e_\theta\right)/\left(\partial\theta\right)=-\mathbf e_r,
\end{equation}

\begin{equation}
\left(\partial\mathbf e_\theta\right)/\left(\partial\phi\right)=\cos\theta\,\mathbf e_\phi,
\end{equation}

\begin{equation}
\left(\partial\mathbf e_\phi\right)/\left(\partial\theta\right)=0,
\end{equation}

and

\begin{equation}
\left(\partial\mathbf e_\phi\right)/\left(\partial\phi\right)
=
-\sin\theta\,\mathbf e_r
-
\cos\theta\,\mathbf e_\theta.
\end{equation}

Applying the chain rule gives the time derivatives

\begin{equation}
\boxed{
\left(d\mathbf e_r\right)/\left(dt\right)
=
\left(d\theta\right)/\left(dt\right)\mathbf e_\theta
+
\left(d\phi\right)/\left(dt\right)\sin\theta\mathbf e_\phi,
}
\end{equation}

\begin{equation}
\boxed{
\left(d\mathbf e_\theta\right)/\left(dt\right)
=
-\left(d\theta\right)/\left(dt\right)\mathbf e_r
+
\left(d\phi\right)/\left(dt\right)\cos\theta\mathbf e_\phi,
}
\end{equation}

and

\begin{equation}
\boxed{
\left(d\mathbf e_\phi\right)/\left(dt\right)
=
-\left(d\phi\right)/\left(dt\right)\sin\theta\mathbf e_r
-
\left(d\phi\right)/\left(dt\right)\cos\theta\mathbf e_\theta.
}
\end{equation}

\section{Spherical velocity}

Differentiate

\begin{equation}
\mathbf r=r\mathbf e_r.
\end{equation}

Then

\begin{equation}
\mathbf v
=
\left(dr\right)/\left(dt\right)\mathbf e_r
+
r\left(d\mathbf e_r\right)/\left(dt\right).
\end{equation}

Substitute the derivative of $\mathbf e_r$:

\begin{equation}
\boxed{
\mathbf v
=
\left(dr\right)/\left(dt\right)\,\mathbf e_r
+
r\left(d\theta\right)/\left(dt\right)\,\mathbf e_\theta
+
r\sin\theta\left(d\phi\right)/\left(dt\right)\,\mathbf e_\phi.
}
\end{equation}

Thus the physical velocity components are

\begin{equation}
\boxed{
v_r=\left(dr\right)/\left(dt\right),
\qquad
v_\theta=r\left(d\theta\right)/\left(dt\right),
\qquad
v_\phi=r\sin\theta\left(d\phi\right)/\left(dt\right).
}
\end{equation}

The factor $r\sin\theta$ is the perpendicular distance from the particle to the $z$ axis, so the azimuthal speed is exactly the circular-motion form $\rho\left(d\phi\right)/\left(dt\right)$ with

\begin{equation}
\rho=r\sin\theta.
\end{equation}

\section{Spherical acceleration}

Start from

\begin{equation}
\mathbf v
=
\left(dr\right)/\left(dt\right)\mathbf e_r
+
r\left(d\theta\right)/\left(dt\right)\mathbf e_\theta
+
r\sin\theta\left(d\phi\right)/\left(dt\right)\mathbf e_\phi.
\end{equation}

Differentiate each term with the product rule and then substitute the three basis-vector derivatives.  Collecting coefficients of $\mathbf e_r$, $\mathbf e_\theta$, and $\mathbf e_\phi$ gives

\begin{equation}
\boxed{
\mathbf a
=
a_r\mathbf e_r+a_\theta\mathbf e_\theta+a_\phi\mathbf e_\phi,
}
\end{equation}

where

\begin{equation}
\boxed{
a_r
=
\left(d^2 r\right)/\left(dt^2\right)
-r\left(\left(d\theta\right)/\left(dt\right)\right)^2
-r\sin^2\theta\left(\left(d\phi\right)/\left(dt\right)\right)^2,
}
\end{equation}

\begin{equation}
\boxed{
a_\theta
=
r\left(d^2\theta\right)/\left(dt^2\right)
+2\left(dr\right)/\left(dt\right)\left(d\theta\right)/\left(dt\right)
-r\sin\theta\cos\theta\left(\left(d\phi\right)/\left(dt\right)\right)^2,
}
\end{equation}

and

\begin{equation}
\boxed{
a_\phi
=
r\sin\theta\left(d^2\phi\right)/\left(dt^2\right)
+2\left(dr\right)/\left(dt\right)\sin\theta\left(d\phi\right)/\left(dt\right)
+2r\cos\theta\left(d\theta\right)/\left(dt\right)\left(d\phi\right)/\left(dt\right).
}
\end{equation}

\begin{center}
\includegraphics{M01_11_fig04_spherical_basis_map.png}

{\small \PMlinkescapetext{Figure 4. The spherical acceleration formula follows from differentiating the velocity while also differentiating the moving spherical basis.  Each coupling term has a geometric origin in a changing scale factor or rotating unit vector.}}
\end{center}

\section{Interpreting the spherical terms}

The radial acceleration contains three contributions:

\begin{equation}
a_r
=
\left(d^2 r\right)/\left(dt^2\right)
-r\left(\left(d\theta\right)/\left(dt\right)\right)^2
-r\sin^2\theta\left(\left(d\phi\right)/\left(dt\right)\right)^2.
\end{equation}

The first is ordinary radial acceleration.  The other two are inward curvature terms associated with angular motion.

The polar component

\begin{equation}
a_\theta
=
r\left(d^2\theta\right)/\left(dt^2\right)+2\left(dr\right)/\left(dt\right)\left(d\theta\right)/\left(dt\right)-r\sin\theta\cos\theta\left(\left(d\phi\right)/\left(dt\right)\right)^2
\end{equation}

contains polar angular acceleration, radial-polar coupling, and a geometric contribution from azimuthal motion.

The azimuthal component

\begin{equation}
a_\phi
=
r\sin\theta\left(d^2\phi\right)/\left(dt^2\right)
+2\left(dr\right)/\left(dt\right)\sin\theta\left(d\phi\right)/\left(dt\right)
+2r\cos\theta\left(d\theta\right)/\left(dt\right)\left(d\phi\right)/\left(dt\right)
\end{equation}

contains azimuthal angular acceleration plus coupling terms from changing radial distance and changing polar angle.

These are kinematic terms produced by coordinate geometry; they do not by themselves represent new physical forces.

\section{Scale factors}

The physical differential displacement in cylindrical coordinates is

\begin{equation}
d\mathbf r
=
d\rho\,\mathbf e_\rho
+
\rho\,d\phi\,\mathbf e_\phi
+
dz\,\mathbf e_z.
\end{equation}

Therefore the cylindrical scale factors are

\begin{equation}
\boxed{h_\rho=1,\qquad h_\phi=\rho,\qquad h_z=1.}
\end{equation}

In spherical coordinates,

\begin{equation}
d\mathbf r
=
dr\,\mathbf e_r
+
r\,d\theta\,\mathbf e_\theta
+
r\sin\theta\,d\phi\,\mathbf e_\phi,
\end{equation}

so

\begin{equation}
\boxed{h_r=1,\qquad h_\theta=r,\qquad h_\phi=r\sin\theta.}
\end{equation}

The velocity formulas can be read directly from these physical line elements.

\section{Coordinate singularities}

Curvilinear coordinates may become singular even when physical space is perfectly regular.

In cylindrical coordinates, $\phi$ is undefined on the $z$ axis where

\begin{equation}
\rho=0.
\end{equation}

In spherical coordinates, all angles are undefined at

\begin{equation}
r=0,
\end{equation}

and the azimuth $\phi$ is undefined on the polar axis where

\begin{equation}
\sin\theta=0.
\end{equation}

These are coordinate singularities, not physical singularities.  One should avoid interpreting divergent coordinate rates near such locations without first checking the actual Cartesian motion.

\section{Worked example 1: uniform helical motion in cylindrical coordinates}

A particle follows

\begin{equation}
\rho=2.0\ {\rm m},
\qquad
\phi=3.0t,
\qquad
z=0.50t.
\end{equation}

Then

\begin{equation}
\left(d\rho\right)/\left(dt\right)=0,
\qquad
\left(d\phi\right)/\left(dt\right)=3.0\ {\rm rad/s},
\qquad
\left(dz\right)/\left(dt\right)=0.50\ {\rm m/s}.
\end{equation}

The velocity is

\begin{equation}
\mathbf v
=
(2.0)(3.0)\mathbf e_\phi
+0.50\mathbf e_z,
\end{equation}

so

\begin{equation}
\boxed{\mathbf v=6.0\mathbf e_\phi+0.50\mathbf e_z\ {\rm m/s}.}
\end{equation}

Its speed is

\begin{equation}
|\mathbf v|=\sqrt{6.0^2+0.50^2}=6.02\ {\rm m/s}.
\end{equation}

Because $\left(d^2\rho\right)/\left(dt^2\right)=\left(d^2\phi\right)/\left(dt^2\right)=\left(d^2 z\right)/\left(dt^2\right)=0$,

\begin{equation}
\mathbf a
=
-\rho\left(\left(d\phi\right)/\left(dt\right)\right)^2\mathbf e_\rho,
\end{equation}

so

\begin{equation}
\boxed{\mathbf a=-18.0\mathbf e_\rho\ {\rm m/s^2}.}
\end{equation}

The vertical motion does not contribute to acceleration because $\left(dz\right)/\left(dt\right)$ is constant.

\section{Worked example 2: expanding spiral with vertical motion}

A particle has

\begin{equation}
\rho=1+t^2,
\qquad
\phi=2t,
\qquad
z=3t,
\end{equation}

with SI units understood.  Find $\mathbf v$ and $\mathbf a$ at $t=1$ s.

At $t=1$,

\begin{equation}
\rho=2,
\qquad
\left(d\rho\right)/\left(dt\right)=2,
\qquad
\left(d^2\rho\right)/\left(dt^2\right)=2,
\end{equation}

\begin{equation}
\left(d\phi\right)/\left(dt\right)=2,
\qquad
\left(d^2\phi\right)/\left(dt^2\right)=0,
\end{equation}

and

\begin{equation}
\left(dz\right)/\left(dt\right)=3,
\qquad
\left(d^2 z\right)/\left(dt^2\right)=0.
\end{equation}

Therefore

\begin{equation}
\boxed{\mathbf v=2\mathbf e_\rho+4\mathbf e_\phi+3\mathbf e_z\ {\rm m/s}.}
\end{equation}

For the acceleration,

\begin{equation}
a_\rho
=
2-(2)(2^2)
=-6\ {\rm m/s^2},
\end{equation}

and

\begin{equation}
a_\phi
=
(2)(0)+2(2)(2)
=8\ {\rm m/s^2}.
\end{equation}

Hence

\begin{equation}
\boxed{\mathbf a=-6\mathbf e_\rho+8\mathbf e_\phi\ {\rm m/s^2}.}
\end{equation}

The nonzero azimuthal acceleration occurs even though $\left(d^2\phi\right)/\left(dt^2\right)=0$ because $\rho$ is increasing while the basis rotates.

\section{Worked example 3: constant-$r$, constant-$\theta$ spherical motion}

A particle moves with

\begin{equation}
r=5.0\ {\rm m},
\qquad
\theta=60^\circ,
\qquad
\phi=0.40t.
\end{equation}

Thus

\begin{equation}
\left(dr\right)/\left(dt\right)=\left(d\theta\right)/\left(dt\right)=0,
\qquad
\left(d\phi\right)/\left(dt\right)=0.40\ {\rm rad/s},
\end{equation}

and all second coordinate derivatives vanish.

The velocity is purely azimuthal:

\begin{equation}
\mathbf v
=
r\sin\theta\left(d\phi\right)/\left(dt\right)\mathbf e_\phi.
\end{equation}

Therefore

\begin{equation}
\boxed{\mathbf v=1.73\mathbf e_\phi\ {\rm m/s}.}
\end{equation}

The acceleration components are

\begin{equation}
a_r
=
-r\sin^2\theta\left(\left(d\phi\right)/\left(dt\right)\right)^2
=-0.600\ {\rm m/s^2},
\end{equation}

\begin{equation}
a_\theta
=
-r\sin\theta\cos\theta\left(\left(d\phi\right)/\left(dt\right)\right)^2
=-0.346\ {\rm m/s^2},
\end{equation}

and

\begin{equation}
a_\phi=0.
\end{equation}

Thus

\begin{equation}
\boxed{\mathbf a=-0.600\mathbf e_r-0.346\mathbf e_\theta\ {\rm m/s^2}.}
\end{equation}

Its magnitude is

\begin{equation}
|\mathbf a|=0.693\ {\rm m/s^2}.
\end{equation}

This is exactly the centripetal acceleration for circular motion about the $z$ axis.  The circle has cylindrical radius

\begin{equation}
\rho=r\sin\theta=4.33\ {\rm m},
\end{equation}

so

\begin{equation}
\omega^2\rho=(0.40)^2(4.33)=0.693\ {\rm m/s^2}.
\end{equation}

\section{Practice problems}

\begin{enumerate}
\item In cylindrical coordinates, a particle has $\rho=3.0$ m, $\left(d\rho\right)/\left(dt\right)=1.0$ m/s, $\left(d\phi\right)/\left(dt\right)=2.0$ rad/s, and $\left(dz\right)/\left(dt\right)=-0.50$ m/s.  Find its velocity vector in the cylindrical basis and its speed.

\item A particle moves on a cylinder of fixed radius $\rho=4.0$ m with $\phi=0.50t^2$ and $z=2.0t$.  Find $\mathbf v$ and $\mathbf a$ at $t=2.0$ s.

\item Show that for $\rho=R=$ constant, $z=$ constant, and $\left(d\phi\right)/\left(dt\right)=\omega=$ constant, the cylindrical acceleration reduces to $\mathbf a=-\omega^2R\mathbf e_\rho$.

\item A particle has $\rho=t^2$, $\phi=t$, and $z=0$.  Evaluate $a_\rho$ and $a_\phi$ at $t=2$ s.

\item In spherical coordinates, $r=4.0$ m, $\theta=45^\circ$, $\left(dr\right)/\left(dt\right)=1.0$ m/s, $\left(d\theta\right)/\left(dt\right)=0.20$ rad/s, and $\left(d\phi\right)/\left(dt\right)=0.50$ rad/s.  Find $v_r$, $v_\theta$, and $v_\phi$.

\item A particle moves radially outward with $r=2+t^2$, while $\theta$ and $\phi$ remain constant.  Find its velocity and acceleration in the spherical basis.

\item A particle moves with $r=3.0$ m, $\theta=90^\circ$, and $\phi=2.0t$.  Use the spherical formulas to find its speed and acceleration, and show that the result matches uniform circular motion in the $xy$ plane.

\item Starting from the Cartesian expression for $\mathbf e_r$, explicitly verify $\partial\mathbf e_r/\partial\theta=\mathbf e_\theta$ and $\partial\mathbf e_r/\partial\phi=\sin\theta\,\mathbf e_\phi$.
\end{enumerate}

\section{Answer check}

\begin{enumerate}
\item $\mathbf v=1.0\mathbf e_\rho+6.0\mathbf e_\phi-0.50\mathbf e_z$ m/s; $|\mathbf v|=6.10$ m/s.

\item At $t=2$ s, $\left(d\phi\right)/\left(dt\right)=2.0$ rad/s and $\left(d^2\phi\right)/\left(dt^2\right)=1.0$ rad/s$^2$.  Thus $\mathbf v=8.0\mathbf e_\phi+2.0\mathbf e_z$ m/s and $\mathbf a=-16.0\mathbf e_\rho+4.0\mathbf e_\phi$ m/s$^2$.

\item Direct substitution gives $a_\rho=-R\omega^2$, $a_\phi=0$, and $a_z=0$.

\item At $t=2$: $\rho=4$, $\left(d\rho\right)/\left(dt\right)=4$, $\left(d^2\rho\right)/\left(dt^2\right)=2$, $\left(d\phi\right)/\left(dt\right)=1$, $\left(d^2\phi\right)/\left(dt^2\right)=0$.  Hence $a_\rho=-2$ m/s$^2$ and $a_\phi=8$ m/s$^2$.

\item $v_r=1.0$ m/s, $v_\theta=0.80$ m/s, $v_\phi=1.41$ m/s.

\item $\mathbf v=2t\mathbf e_r$ and $\mathbf a=2\mathbf e_r$.

\item $v=r\left(d\phi\right)/\left(dt\right)=6.0$ m/s and $\mathbf a=-r\left(\left(d\phi\right)/\left(dt\right)\right)^2\mathbf e_r=-12.0\mathbf e_r$ m/s$^2$; at $\theta=90^\circ$, $\mathbf e_r$ lies in the $xy$ plane and points radially outward.

\item Differentiation gives the stated basis identities exactly.
\end{enumerate}

\section{Summary}

Cylindrical and spherical coordinates simplify problems with axial or radial geometry, but their basis vectors move with the particle.  The central cylindrical results are

\begin{equation}
\boxed{\mathbf v=\left(d\rho\right)/\left(dt\right)\mathbf e_\rho+\rho\left(d\phi\right)/\left(dt\right)\mathbf e_\phi+\left(dz\right)/\left(dt\right)\mathbf e_z,}
\end{equation}

\begin{equation}
\boxed{\mathbf a=(\left(d^2\rho\right)/\left(dt^2\right)-\rho\left(\left(d\phi\right)/\left(dt\right)\right)^2)\mathbf e_\rho+(\rho\left(d^2\phi\right)/\left(dt^2\right)+2\left(d\rho\right)/\left(dt\right)\left(d\phi\right)/\left(dt\right))\mathbf e_\phi+\left(d^2 z\right)/\left(dt^2\right)\mathbf e_z.}
\end{equation}

For the stated spherical convention,

\begin{equation}
\boxed{\mathbf v=\left(dr\right)/\left(dt\right)\mathbf e_r+r\left(d\theta\right)/\left(dt\right)\mathbf e_\theta+r\sin\theta\left(d\phi\right)/\left(dt\right)\mathbf e_\phi,}
\end{equation}

with acceleration components

\begin{equation}
\boxed{a_r=\left(d^2 r\right)/\left(dt^2\right)-r\left(\left(d\theta\right)/\left(dt\right)\right)^2-r\sin^2\theta\left(\left(d\phi\right)/\left(dt\right)\right)^2,}
\end{equation}

\begin{equation}
\boxed{a_\theta=r\left(d^2\theta\right)/\left(dt^2\right)+2\left(dr\right)/\left(dt\right)\left(d\theta\right)/\left(dt\right)-r\sin\theta\cos\theta\left(\left(d\phi\right)/\left(dt\right)\right)^2,}
\end{equation}

and

\begin{equation}
\boxed{a_\phi=r\sin\theta\left(d^2\phi\right)/\left(dt^2\right)+2\left(dr\right)/\left(dt\right)\sin\theta\left(d\phi\right)/\left(dt\right)+2r\cos\theta\left(d\theta\right)/\left(dt\right)\left(d\phi\right)/\left(dt\right).}
\end{equation}

The safest derivation strategy is always the same: write the position vector, determine how the local basis vectors change, and then differentiate carefully with the product and chain rules.

\begin{thebibliography}{9}

\bibitem{M0010}
PhysicsLibrary, M01-10, \emph{Polar-Coordinate Particle Kinematics}.

\bibitem{WB}
Wikibooks, \emph{Classical Mechanics}, sections on generalized and curvilinear coordinate descriptions, CC BY-SA.

\bibitem{WIKI}
Wikipedia, articles on cylindrical coordinate systems, spherical coordinate systems, and vector calculus in curvilinear coordinates, CC BY-SA.

\bibitem{Marion}
J. B. Marion and S. T. Thornton, \emph{Classical Dynamics of Particles and Systems}, 5th ed., Brooks/Cole, 2004.  Used as a scope and notation reference.

\end{thebibliography}</content>
</record>
