Dynamical Masses from Kepler's Third Law: A DerivationDynamicalMassesFromKeplersThirdLawADerivation2026-10-04 15:22:162026-10-04 20:51:00DerivationBinary Stars as Physical Laboratoriesdynamical massNewtonian form of Kepler's third lawrelative semimajor axisbarycentric semimajor axis% 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\section*{Dynamical Masses from Kepler's Third Law: A Derivation}
One of the most important reasons binary stars are astrophysical laboratories is that their masses can be measured dynamically. test edit.
The central relation is
\begin{equation}
\boxed{
M_1+M_2
=
\frac{4\pi^2a^3}{GP^2},
}
\tag{1}
\end{equation}
where
\begin{itemize}
\item $M_1$ and $M_2$ are the stellar masses,
\item $P$ is the common orbital period,
\item $a$ is the semimajor axis of the \emph{relative orbit},
\item $G$ is the Newtonian gravitational constant.
\end{itemize}
Equation (1) is the Newtonian two-body form of Kepler's third law.
It is the basis of the dynamical-mass discussion in BIN01.
The most important geometric point is
\begin{equation}
\boxed{
a=a_1+a_2,
}
\tag{2}
\end{equation}
where $a_1$ and $a_2$ are the semimajor axes of the two stars about the center of mass.
The $a$ appearing in Equation (1) is \emph{not} the orbit size of either star individually.
\begin{center}
\includegraphics[width=0.90\textwidth,keepaspectratio]{BIN01D1_fig01_two_star_barycenter.png}
{\small \PMlinkescapetext{Figure 1. Both stars orbit their common center of mass. The relative separation is the sum of the two barycentric orbital radii in the circular case and the sum of their semimajor axes in the general Keplerian case.}}
\end{center}
\section{Why this relation measures mass}
For an isolated Newtonian binary, orbital motion is produced by the gravitational attraction of the two stars.
The orbital size and orbital period therefore encode the strength of the gravitational field.
A larger total mass produces stronger gravitational acceleration.
For a fixed orbital size, a larger total mass implies a shorter orbital period.
For a fixed period, a larger orbital size requires a larger total mass.
Equation (1) converts this statement into a quantitative measurement.
The mass is called \emph{dynamical} because it is obtained from the motion caused by gravity rather than inferred only from luminosity, temperature, or a stellar-evolution model.
\section{First derivation: circular orbit intuition}
The clearest introductory derivation begins with circular orbits.
Let the two stars have masses
\begin{equation}
M_1
\end{equation}
and
\begin{equation}
M_2.
\end{equation}
Let their distances from the center of mass be
\begin{equation}
a_1
\end{equation}
and
\begin{equation}
a_2.
\end{equation}
Their separation is
\begin{equation}
\boxed{
a=a_1+a_2.
}
\tag{3}
\end{equation}
Because both stars complete one revolution in the same orbital period $P$, they have the same angular speed
\begin{equation}
\boxed{
\omega
=
\frac{2\pi}{P}.
}
\tag{4}
\end{equation}
\section{Force balance for star 1}
The gravitational force magnitude is
\begin{equation}
\boxed{
F_g
=
\frac{GM_1M_2}{a^2}.
}
\tag{5}
\end{equation}
For circular motion, star 1 requires centripetal force
\begin{equation}
F_{c,1}
=
M_1\omega^2a_1.
\end{equation}
Gravity supplies that force:
\begin{equation}
\frac{GM_1M_2}{a^2}
=
M_1\omega^2a_1.
\end{equation}
Cancel $M_1$:
\begin{equation}
\boxed{
\omega^2a_1
=
\frac{GM_2}{a^2}.
}
\tag{6}
\end{equation}
\section{Force balance for star 2}
The same gravitational force acts on star 2.
Its circular-motion equation is
\begin{equation}
\frac{GM_1M_2}{a^2}
=
M_2\omega^2a_2.
\end{equation}
Cancel $M_2$:
\begin{equation}
\boxed{
\omega^2a_2
=
\frac{GM_1}{a^2}.
}
\tag{7}
\end{equation}
\begin{center}
\includegraphics[width=0.90\textwidth,keepaspectratio]{BIN01D1_fig02_circular_force_balance.png}
{\small \PMlinkescapetext{Figure 2. In the circular-orbit derivation, the same mutual gravitational force supplies the centripetal acceleration of each star about the common center of mass.}}
\end{center}
\section{Add the two equations}
Add Equations (6) and (7):
\begin{equation}
\omega^2a_1
+
\omega^2a_2
=
\frac{G(M_1+M_2)}{a^2}.
\end{equation}
Factor the left side:
\begin{equation}
\omega^2(a_1+a_2)
=
\frac{G(M_1+M_2)}{a^2}.
\end{equation}
Using
\begin{equation}
a_1+a_2=a,
\end{equation}
we obtain
\begin{equation}
\omega^2a
=
\frac{G(M_1+M_2)}{a^2}.
\end{equation}
Multiply by $a^2$:
\begin{equation}
\boxed{
\omega^2a^3
=
G(M_1+M_2).
}
\tag{8}
\end{equation}
Solve for total mass:
\begin{equation}
M_1+M_2
=
\frac{\omega^2a^3}{G}.
\end{equation}
Now insert
\begin{equation}
\omega
=
\frac{2\pi}{P}.
\end{equation}
Then
\begin{align}
M_1+M_2
&=
\frac{a^3}{G}
\left(
\frac{2\pi}{P}
\right)^2
\\
&=
\boxed{
\frac{4\pi^2a^3}{GP^2}.
}
\end{align}
Thus
\begin{equation}
\boxed{
M_1+M_2
=
\frac{4\pi^2a^3}{GP^2}.
}
\tag{9}
\end{equation}
This is the desired relation.
\section{What the circular derivation teaches}
The circular derivation is especially useful because every step has a simple physical meaning.
It shows that:
\begin{enumerate}
\item gravity provides the centripetal acceleration,
\item both stars have the same angular speed,
\item both stars contribute to the orbital separation,
\item the total mass, not only one stellar mass, controls the relative orbital period.
\end{enumerate}
However, Equation (9) is not restricted to circular orbits.
The same result holds for elliptical Keplerian binaries.
\section{Second derivation: the general two-body problem}
Let the inertial-frame positions be
\begin{equation}
\mathbf r_1
\end{equation}
and
\begin{equation}
\mathbf r_2.
\end{equation}
Define the relative vector
\begin{equation}
\boxed{
\mathbf r
=
\mathbf r_1-\mathbf r_2.
}
\tag{10}
\end{equation}
Newton's equations are
\begin{align}
M_1\ddot{\mathbf r}_1
&=
-\,G
\frac{M_1M_2}{r^3}
\mathbf r,
\\
M_2\ddot{\mathbf r}_2
&=
+\,G
\frac{M_1M_2}{r^3}
\mathbf r.
\end{align}
Divide the first by $M_1$ and the second by $M_2$, then subtract:
\begin{align}
\ddot{\mathbf r}
&=
\ddot{\mathbf r}_1-\ddot{\mathbf r}_2
\\
&=
-\,G
\frac{M_2}{r^3}\mathbf r
-
G
\frac{M_1}{r^3}\mathbf r.
\end{align}
Therefore
\begin{equation}
\boxed{
\ddot{\mathbf r}
=
-\frac{G(M_1+M_2)}{r^3}\mathbf r.
}
\tag{11}
\end{equation}
The two-star problem has become one effective Kepler problem for the relative coordinate.
\begin{center}
\includegraphics[width=0.90\textwidth,keepaspectratio]{BIN01D1_fig03_relative_orbit.png}
{\small \PMlinkescapetext{Figure 3. Subtracting the two Newton equations reduces the binary to one relative orbit governed by the total mass.}}
\end{center}
\section{The gravitational parameter of the relative orbit}
Equation (11) has the same form as the standard Kepler equation
\begin{equation}
\ddot{\mathbf r}
=
-\frac{\mu_g}{r^3}\mathbf r,
\end{equation}
with
\begin{equation}
\boxed{
\mu_g
=
G(M_1+M_2).
}
\tag{12}
\end{equation}
The total stellar mass therefore plays the role that a central mass would play in a test-particle Kepler problem.
This is why the period of a binary depends on
\begin{equation}
M_1+M_2.
\end{equation}
\section{Angular momentum and areal velocity}
Because the force in Equation (11) is a central force,
\begin{equation}
\mathbf r\times\ddot{\mathbf r}=0.
\end{equation}
Define specific angular momentum
\begin{equation}
\boxed{
\mathbf h
=
\mathbf r\times\dot{\mathbf r}.
}
\tag{13}
\end{equation}
Then
\begin{equation}
\boxed{
\frac{d\mathbf h}{dt}=0.
}
\tag{14}
\end{equation}
The motion is therefore planar.
The areal velocity is
\begin{equation}
\boxed{
\frac{dA}{dt}
=
\frac{h}{2}.
}
\tag{15}
\end{equation}
This is Kepler's second law.
\begin{center}
\includegraphics[width=0.90\textwidth,keepaspectratio]{BIN01D1_fig04_equal_areas.png}
{\small \PMlinkescapetext{Figure 4. Conservation of orbital angular momentum gives constant areal velocity, which allows the orbital period to be related to the total area of the ellipse.}}
\end{center}
\section{Specific angular momentum of an ellipse}
For a Kepler ellipse with semimajor axis $a$ and eccentricity $e$,
\begin{equation}
\boxed{
h^2
=
G(M_1+M_2)a(1-e^2).
}
\tag{16}
\end{equation}
Thus
\begin{equation}
\boxed{
h
=
\sqrt{
G(M_1+M_2)a(1-e^2)
}.
}
\tag{17}
\end{equation}
The semiminor axis of the ellipse is
\begin{equation}
\boxed{
b
=
a\sqrt{1-e^2}.
}
\tag{18}
\end{equation}
The area of the ellipse is therefore
\begin{align}
A
&=
\pi ab
\\
&=
\boxed{
\pi a^2\sqrt{1-e^2}.
}
\tag{19}
\end{align}
\section{Use the time required to sweep the whole ellipse}
The orbital period is the time required for the radius vector to sweep the entire orbital area.
Since
\begin{equation}
\frac{dA}{dt}
=
\frac{h}{2},
\end{equation}
we have
\begin{equation}
P
=
\frac{A}{dA/dt}.
\end{equation}
Substitute Equations (15) and (19):
\begin{align}
P
&=
\frac{
\pi a^2\sqrt{1-e^2}
}{
h/2
}
\\
&=
\frac{
2\pi a^2\sqrt{1-e^2}
}{h}.
\end{align}
Now insert Equation (17):
\begin{equation}
P
=
\frac{
2\pi a^2\sqrt{1-e^2}
}{
\sqrt{
G(M_1+M_2)a(1-e^2)
}
}.
\end{equation}
The eccentricity factors cancel:
\begin{equation}
P
=
\frac{
2\pi a^2
}{
\sqrt{
G(M_1+M_2)a
}
}.
\end{equation}
Simplify the powers of $a$:
\begin{equation}
\boxed{
P
=
2\pi
\sqrt{
\frac{a^3}{G(M_1+M_2)}
}.
}
\tag{20}
\end{equation}
Square:
\begin{equation}
P^2
=
\frac{
4\pi^2a^3
}{
G(M_1+M_2)
}.
\end{equation}
Rearrange:
\begin{equation}
\boxed{
M_1+M_2
=
\frac{4\pi^2a^3}{GP^2}.
}
\tag{21}
\end{equation}
This is the same result obtained from the circular derivation.
\section{Why eccentricity disappears}
A striking feature of Equation (21) is that it contains no eccentricity.
At first this can seem surprising.
An eccentric binary moves much faster near periastron than near apastron.
However, the ellipse area contains the factor
\begin{equation}
\sqrt{1-e^2},
\end{equation}
and the angular momentum contains the same factor.
They cancel when the full orbital period is calculated.
Therefore
\begin{equation}
\boxed{
\text{for fixed total mass and semimajor axis, the Keplerian period is independent of eccentricity.}
}
\end{equation}
This does not mean that the instantaneous speed or separation is independent of eccentricity.
Only the period for a given $a$ and total mass has this property.
\section{Relative orbit versus barycentric orbits}
The individual stellar semimajor axes satisfy
\begin{equation}
\boxed{
a_1
=
\frac{M_2}{M_1+M_2}a
}
\tag{22}
\end{equation}
and
\begin{equation}
\boxed{
a_2
=
\frac{M_1}{M_1+M_2}a.
}
\tag{23}
\end{equation}
Adding them gives
\begin{equation}
a_1+a_2
=
\frac{M_1+M_2}{M_1+M_2}a,
\end{equation}
so
\begin{equation}
\boxed{
a=a_1+a_2.
}
\tag{24}
\end{equation}
This is the orbital length scale required in Equation (21).
Using only $a_1$ or $a_2$ in the total-mass formula would give the wrong result.
\section{Astronomical-unit form}
Equation (21) becomes especially simple in standard astronomical units.
For Earth's orbit around the Sun,
\begin{equation}
1\,{\rm yr}^2
=
\frac{
4\pi^2(1\,{\rm AU})^3
}{
GM_\odot
}
\end{equation}
to the usual Newtonian approximation.
Therefore, when
\begin{itemize}
\item $a$ is in AU,
\item $P$ is in years,
\item mass is in solar masses,
\end{itemize}
Equation (21) becomes
\begin{equation}
\boxed{
\frac{M_1+M_2}{M_\odot}
\approx
\frac{
a^3({\rm AU})
}{
P^2({\rm yr})
}.
}
\tag{25}
\end{equation}
This is the form most often used for visual binaries.
\section{Visual binaries and parallax}
A visual-binary orbit usually yields an angular semimajor axis
\begin{equation}
a^{\prime\prime}.
\end{equation}
Kepler's law requires a physical semimajor axis.
If the annual parallax is
\begin{equation}
\varpi^{\prime\prime},
\end{equation}
then
\begin{equation}
\boxed{
a({\rm AU})
=
\frac{
a^{\prime\prime}
}{
\varpi^{\prime\prime}
}.
}
\tag{26}
\end{equation}
Substitute this into Equation (25):
\begin{equation}
\boxed{
\frac{M_1+M_2}{M_\odot}
\approx
\frac{
\left(
a^{\prime\prime}/\varpi^{\prime\prime}
\right)^3
}{
P^2({\rm yr})
}.
}
\tag{27}
\end{equation}
This is the direct observational route from an angular visual orbit to total dynamical mass.
\begin{center}
\includegraphics[width=0.92\textwidth,keepaspectratio]{BIN01D1_fig05_visual_binary_mass_chain.png}
{\small \PMlinkescapetext{Figure 5. A visual-binary dynamical mass combines angular orbit scale, parallax, and period.}}
\end{center}
\section{Numerical example}
Suppose a visual binary has
\begin{align}
a^{\prime\prime}&=0.50\ {\rm arcsec},\\
\varpi&=25\ {\rm mas},\\
P&=20\ {\rm yr}.
\end{align}
Convert parallax to arcseconds:
\begin{equation}
\varpi
=
0.025\ {\rm arcsec}.
\end{equation}
The physical semimajor axis is
\begin{align}
a
&=
\frac{0.50}{0.025}
\\
&=
20\ {\rm AU}.
\end{align}
Then
\begin{align}
\frac{M_1+M_2}{M_\odot}
&\approx
\frac{20^3}{20^2}
\\
&=
\frac{8000}{400}
\\
&=
20.
\end{align}
Therefore
\begin{equation}
\boxed{
M_1+M_2
\approx
20M_\odot.
}
\end{equation}
This example also illustrates the strong cubic dependence on orbital size.
\section{Sensitivity to measurement errors}
From
\begin{equation}
M
\propto
a^3P^{-2},
\end{equation}
take a logarithmic differential:
\begin{equation}
d\ln M
=
3\,d\ln a
-
2\,d\ln P.
\end{equation}
Thus
\begin{equation}
\boxed{
\frac{dM}{M}
=
3\frac{da}{a}
-
2\frac{dP}{P}.
}
\tag{28}
\end{equation}
For a visual binary,
\begin{equation}
a
=
\frac{a^{\prime\prime}}{\varpi^{\prime\prime}},
\end{equation}
so
\begin{equation}
\frac{da}{a}
=
\frac{da^{\prime\prime}}{a^{\prime\prime}}
-
\frac{d\varpi}{\varpi}.
\end{equation}
Therefore
\begin{equation}
\boxed{
\frac{dM}{M}
=
3\frac{da^{\prime\prime}}{a^{\prime\prime}}
-
3\frac{d\varpi}{\varpi}
-
2\frac{dP}{P}.
}
\tag{29}
\end{equation}
For independent random uncertainties, the approximate variance relation is
\begin{equation}
\boxed{
\left(
\frac{\sigma_M}{M}
\right)^2
\approx
9
\left(
\frac{\sigma_a}{a}
\right)^2
+
4
\left(
\frac{\sigma_P}{P}
\right)^2.
}
\tag{30}
\end{equation}
If the physical semimajor axis is obtained from angular semimajor axis and parallax,
\begin{equation}
\boxed{
\left(
\frac{\sigma_M}{M}
\right)^2
\approx
9
\left(
\frac{\sigma_{a^{\prime\prime}}}{a^{\prime\prime}}
\right)^2
+
9
\left(
\frac{\sigma_\varpi}{\varpi}
\right)^2
+
4
\left(
\frac{\sigma_P}{P}
\right)^2,
}
\tag{31}
\end{equation}
when correlations can be neglected.
This is why accurate parallaxes are so valuable for visual-binary mass measurements.
\section{Total mass versus individual masses}
Equation (21) gives
\begin{equation}
M_1+M_2.
\end{equation}
It does not by itself separate the two component masses.
A second observable is required.
For example, if the barycentric orbit sizes are known,
\begin{equation}
\boxed{
\frac{M_1}{M_2}
=
\frac{a_2}{a_1}.
}
\tag{32}
\end{equation}
For a double-lined spectroscopic binary,
\begin{equation}
\boxed{
\frac{M_1}{M_2}
=
\frac{K_2}{K_1},
}
\tag{33}
\end{equation}
where $K_1$ and $K_2$ are the radial-velocity semiamplitudes.
Combining total mass with a mass ratio gives the individual masses.
\section{Physical interpretation}
Equation (21) can be written schematically as
\begin{equation}
\boxed{
M
\sim
\frac{\text{orbital size}^3}
{\text{orbital time}^2}.
}
\end{equation}
The dimensional structure reflects Newtonian gravity.
Since
\begin{equation}
[G]
=
\frac{L^3}{MT^2},
\end{equation}
we have
\begin{equation}
\frac{a^3}{GP^2}
\end{equation}
with dimensions of mass.
The combination is therefore not arbitrary.
It is the natural mass scale formed from the orbital length, orbital time, and gravitational constant.
\section{Common mistakes}
\begin{enumerate}
\item Using the semimajor axis of one star instead of the relative semimajor axis.
\item Using the instantaneous separation instead of the semimajor axis.
\item Treating one star as fixed when both masses are significant.
\item Applying the circular centripetal-force argument directly at every point of an eccentric orbit.
\item Concluding from the circular derivation that Kepler's third law is valid only for circular motion.
\item Forgetting that the general elliptical derivation uses the semimajor axis, not periastron or apastron distance.
\item Using angular semimajor axis directly in Equation (21) without converting it to physical units.
\item Forgetting to convert milliarcseconds to arcseconds before using the simple parallax conversion.
\item Assuming total dynamical mass automatically gives the individual stellar masses.
\item Ignoring the cubic sensitivity of dynamical mass to semimajor-axis uncertainty.
\end{enumerate}
\section{Practice exercises}
\begin{enumerate}
\item Reproduce the circular derivation beginning with the force balance for each star.
\item Show explicitly why adding the two circular equations produces the total mass $M_1+M_2$.
\item Derive the relative equation of motion by subtracting the two Newton equations.
\item Explain why the relative problem depends on $M_1+M_2$ rather than on either mass alone.
\item Starting from constant areal velocity, derive the elliptical form of Kepler's third law.
\item Show exactly where the factor $\sqrt{1-e^2}$ cancels in the elliptical derivation.
\item A binary has $a=10$ AU and $P=25$ yr. Estimate its total mass.
\item A visual binary has $a^{\prime\prime}=0.40$ arcsec, parallax $20$ mas, and period $30$ yr. Find the total mass.
\item If the orbital semimajor axis has a one-percent uncertainty and the period uncertainty is negligible, estimate the resulting fractional mass uncertainty.
\item Explain what additional measurement is required to separate a total dynamical mass into individual stellar masses.
\end{enumerate}
\section{Summary}
For a binary star, the Newtonian two-body problem gives
\begin{equation}
\boxed{
\ddot{\mathbf r}
=
-\frac{G(M_1+M_2)}{r^3}\mathbf r.
}
\end{equation}
The relative orbit therefore behaves as a Kepler orbit controlled by the total mass.
For an ellipse,
\begin{equation}
\boxed{
P
=
2\pi
\sqrt{
\frac{a^3}{G(M_1+M_2)}
}.
}
\end{equation}
Thus
\begin{equation}
\boxed{
M_1+M_2
=
\frac{4\pi^2a^3}{GP^2}.
}
\end{equation}
The same formula is obtained from the simpler circular force-balance derivation.
In astronomical units,
\begin{equation}
\boxed{
\frac{M_1+M_2}{M_\odot}
\approx
\frac{a^3({\rm AU})}{P^2({\rm yr})}.
}
\end{equation}
For a visual binary,
\begin{equation}
\boxed{
a({\rm AU})
=
\frac{a^{\prime\prime}}{\varpi^{\prime\prime}}.
}
\end{equation}
The combination of orbit size, distance, and period therefore provides a direct dynamical measurement of stellar mass.
\section*{References}
\begin{thebibliography}{10}
\bibitem{CarrollOstlie2017}
B. W. Carroll and D. A. Ostlie,
\textit{An Introduction to Modern Astrophysics}, 2nd ed.,
Cambridge University Press, 2017.
\bibitem{Hilditch2001}
R. W. Hilditch,
\textit{An Introduction to Close Binary Stars},
Cambridge University Press, 2001.
\bibitem{MurrayDermott1999}
C. D. Murray and S. F. Dermott,
\textit{Solar System Dynamics},
Cambridge University Press, 1999.
\bibitem{Taylor2005}
J. R. Taylor,
\textit{Classical Mechanics},
University Science Books, 2005.
\end{thebibliography}