You are on page 1of 11

J. Peraire, S. Widnall 16.07 Dynamics Fall 2011 Version 1.

2

Lecture L15 - Central Force Motion: Kepler’s Laws

When the only force acting on a particle is always directed towards a fixed point, the motion is called central force motion. This type of motion is particularly relevant when studying the orbital movement of planets and satellites. The laws which govern this motion were first postulated by Kepler and deduced from observation. In this lecture, we will see that these laws are a consequence of Newton’s second law. An understanding of central force motion is necessary for the design of satellites and space vehicles.

Kepler’s Problem
We consider the motion of a particle of mass m, in an inertial reference frame, under the influence of a fixed gravitational force, F , directed towards the origin.

We will be particularly interested in the case when the force is inversely proportional to the square of the distance between the particle and the origin, such as the gravitational force. In this case, F =− µ mer , r2 (1)

where µ is the gravitational parameter, r is the modulus of the position vector, r, and er = r/r. From Newton’s law F = ma, we obtain the governing equation for a body attracted by a fixed gravitational force; the is often referred to as Kepler’s problem. m¨ = − r 1 µ mer r2 (2)

in this case. move solely due to the influence of their mutual gravitational attraction. This equivalence is obvious when M m. this is the governing equation. Therefore. Note Equivalence between the two-body problem and Kepler’s problem Here we consider the problem of two isolated bodies of masses M and m which interact though gravitational attraction. since. the center of mass of the system can be taken to be at M . from momentum conservation. and. Kepler’s problem is equivalent to the two-body problem. we will have. many problems of practical interest can be accurately solved by just looking at two bodies at a time.. even in the more general case when the two masses are of similar size. the problem is considerably more complicated. G. His interest was in describing the motion of planets around the sun. A more complex situation occurs for mutually attracting bodies such as the earth and moon whose joint influence cannot be ignored.The orbits of the planets are ellipses with the Sun at one focus 2. of the two bodies will be rG = M r M + mr m . M +m (3) ˙ Since the two bodies are isolated. The two body problem was studied by Kepler (1571-1630) who lived before Newton was born.. Let r M and r m denote the position vectors of the two bodies relative to a fixed origin O. He postulated the following laws: 1. the problem can be reduced to a Kepler problem (see note below). that r G =constant. When more than two bodies are involved. M and m.The square of the period of a planet is proportional to the cube of the major axis of its elliptical orbit In this lecture. Thus. in which two masses.The line joining a planet to the Sun sweeps out equal areas in equal intervals of time 3. 2 . in general. such as the massive sun which can be considered at rest in calculations of planetary motion. However. in this case.. Although most problems in celestial mechanics involve more than two bodies.For a fixed gravitational attractor. at all times. the position of the center of gravity. It can be shown that. we will start from Newton’s laws and verify that the above three laws can indeed be derived from Newtonian mechanics. and ¨ r G = 0. can be found trivially from the initial conditions. no general solutions are known. the position of the center of gravity.

Mm Mm ¨ r = −G 2 er . In order to derive an equation for r. Once the solution r to this somewhat ”fictitious” problem has been obtain. from conservation of angular momentum. this orbit will be planar. but the mass of the orbiting body (m in this case). Solution of this ”problem” will provide an orbit. the problem of determining r M and r m is equivalent to that of determining r. Translating the equations of motion for two gravitationally-interacting bodies into their joint motion about their center of mass. r2 which is again a Kepler problem for an orbiting body of mass m. r2 (7) M +m er . the above expression can be written as m¨ = −G r (M + m)m er . M m/(M + m). Note that when M >> m. the above expression is general and applies to general masses M and m. is known. we have ¨ M rM = G Mm er . has been replaced by the reduced mass. However. Using this solution 3 . which travels with constant velocity. This is a very powerful result. r2 m¨ m = −G r Mm er . If we add these two equations. r = r m − r M . Subtracting these two equations. results in boundary conditions on position and velocity for a single ”fictitious” body. then the position vectors of M and m could be computed as rM = rG − rm m r M +m M r. the solution for the motion of the two bodies can be obtain directly for Equations (4) and (5) giving the position of rm and rm in terms of the center of mass rG and the solution to Equation (6) r. er = r/r. the equal and opposite gravitational attraction between them drops out and the result is ¨ M r M + m¨ m = 0 = (M + m)¨G r r (6) This states that the acceleration of the center of mass rG is zero: the center of mass can at most move at ¨ constant velocity. Alternatively. where r is the distance between the two bodies. r2 where r = |r|. and G is the gravitational constant. the reduced mass becomes m.If the position vector of m as observed by M . Thus. we first consider the equations of motion for m and M independently. we obtain the equation for the relative motion of the two masses. in which the gravitational parameter µ is given by µef f = G(M + m). M +m r The above expression shows that the motion of m relative to M is in fact a Kepler problem in which the force is given by −GM mer /r2 (this is indeed the real force). ¨ ¨ ¨ r = r m − r M = −G or. = rG + M +m (4) (5) Therefore. As will be seen.

4 . as a consequence. is µm er = m¨ . In a time dt. and. ¨ 0 = r θ + 2r θ .to describe the motion of the actual two bodies results in two planar orbits. in the same plane.e. dA. − µ ˙ ¨ er = (¨ − rθ2 )er + (rθ + 2rθ)eθ . h = |r × v|. the angular momentum of m with respect ˙ to the origin will be constant. in which the bodies orbit about their common center of mass. ˙˙ Using the following identity. Equations of Motion for the Kepler Problem The equation of motion (F = ma). 1 r the above equation implies that ˙ r2 θ = h ≡ constant. (8) d 2˙ (r θ) dt ¨ = rθ + 2rθ. the position and velocity vectors. that will be determined by the initial conditions. the motion will be planar. the area. i. dA 1 ˙ h = r2 θ = . r ˙˙ r2 Now. Therefore. we consider the radial and circumferential components of this equation separately. we have. Using cylindrical coordinates. with ez being parallel to the angular momentum vector. will be in a plane orthogonal to the angular momentum vector. dt 2 2 which proves Kepler’s second law:The line joining a planet to the Sun sweeps out equal areas in equal intervals of time. Therefore. ˙˙ We note that the constant of integration. r and r. swept by r will be dA = r rdθ/2. is precisely the magnitude of the specific angular momentum vector. h. Circumferential component We have. r r2 − Since the only force in the system is directed towards point O.

and ψ = 0. and r2 = h/θ from equation 8. or focus O.Radial component The radial component of the equation of motion reads. P . then e will be positive. Inserting this expression into equation 9. and a point. r = −h ¨ d2 dθ2 1 r h2 d2 ˙ θ=− 2 2 r dθ 1 r . P O. e = P O/P A. d2 dθ2 1 r + 1 µ = 2. is a constant e. we obtain the following differential equation for 1/r as a function of θ. P A. If we choose θ to be zero when r is minimum. That is. 1 µ = 2 (1 + e cos(θ + ψ)) . Differentiating with respect to time. a conic section is the locus of points. 5 . to the distance between the point and the directrix. − d Since −r2 dt 1 r µ ˙ = r − rθ2 . ¨ r2 (9) ˙ = r. we can write ˙ r=− ˙ h d ˙ θ dt 1 r = −h d dθ 1 r . such that the ratio of the distance between the point and the focus. 1 + e cos θ (10) We shall see below that this is the equation of a conic section in polar coordinates. r h This is a linear second order ordinary differential equation which has a general solution of the form. Conic Sections Conic sections are planar curves that are defined as follows: given a line. The equation describing the trajectory will be r= h2 /µ . or directrix. r h where e and ψ are two constants of integration. and using equation 8.

In particular. thus proving Kepler’s first law:The orbits of the planets are ellipses with the Sun at one focus. Since for a hyperbolic or parabolic orbit. and the velocity at the point of closest approach vπ is perpendicular to the radius rπ . The constant e ≥ 0 is called the eccentricity. these points are called the perihelion and aphelion. we have that when e=0 e<1 e=1 e>1 the curve is a circle the curve is an ellipse the curve is a parabola the curve is a hyperbola. When considering orbits around the earth. and equation(10) which provides the solution of the motion of a point mass in a gravitational field. whereas for orbits around the sun. the conic surface will be either an open or closed curve. we have r= p . periapsis and is denoted by rπ . (12) When e < 1. therefore rα − > inf. depending on its value. (14) 6 . these points are called the perigee and apogee. the point in the trajectory which is farthest away from the focus is called the apoapsis and is denoted by α. we can obtain an expression for the velocity at rπ from the angular momentum h = rπ vπ resulting in vπ = µ(1 + e) rπ (13) The point in the trajectory which is closest to the focus is called the For elliptical orbits. 1 + e cos θ (11) Here. p is the parameter of the conic and is equal to r when θ = ±90o . Elliptical Trajectories If a is the semi-major axis of the ellipse. and. we see that the trajectory of a mass under the influence of a central force will be a conic curve with parameter p = h2 /µ. r− > inf. Since the angular momentum h is constant for an orbit. In particular.Since P O = r and P A = p/e − r cos θ. we can identify the properties of the conic section orbits in terms of the physical parameters of the Kepler problem. at or before θ− > π. the radius goes to infinity. the trajectory is an ellipse. respectively. there is no second crossing of the x axis at finite r. Comparing equation(11) which deals solely with the property of a conic section. then 2a = rπ + rα .

the distance between O and the center of the ellipse will be a − rπ = a e. 7 (24) (23) (22) .Using equation (10) to evaluate rπ and rα . plus the distance between the focus and the directrix. Also. (21) (20) (18) (19) (17) (15) (16) Other geometric properties of the ellipse are that the distance between point D and the directrix will be equal to DO/e. since dA/dt = h/2 is a constant. That is. we have A = hT /2. DO = a e2 + p = a. the semi-minor axis of the ellipse √ will be b = a 1 − e2 . (for θ = 0 and θ = π) we obtain p 1+e p rα = 1−e rπ = resulting in a = p/(1 − e2 ). b2 + (a e)2 = a2 . and rπ = a(1 − e) rα = a(1 + e) Since p = h2 /µ. we can also write h2 = aµ(1 − e2 ) Also. Therefore. DO/e = ae + p/e. which in turn will be equal to the sum of the distance between the focus and the center of the ellipse. using Pythagoras’ theorem. Hence. The area of the ellipse is given by A = πab.

the periapsis. such as θ(t). which states that the motion of the planet sweeps out area at a constant rate. T πab (28) To find the area AP we construct a circle of radius a with origin at the center of the ellipse. tP .where T is the period of the orbit. This result is significant: it states that the period of an orbit is a function only of its semimajor axis. and an arbitrary point P . 8 . tP AP = . the time required to travel from π to P . However. we use Kepler’s second law. (27) This result proves Kepler’s third law:The square of the period of a planet is proportional to the cube of the major axis of its elliptical orbit. Referring to the figure. since the total period of the orbit is T and the total area of an ellipse is πab. on the ellipse intersecting the point O” on the axis. to the period T equals the fraction that the area AP represents of the total area of the ellipse. we have T2 = This is often written as µ= 2π T (2π)2 a3 µ 2 (25) (26) a3 . relative to the total period T is the ratio of the curved area denoted by AP (AP is the sector defined by O. this solution gives us no direct information about the time behavior of the motions. we see that the time required to travel between the point π. More specifically. We identify a point P on the circle to be in a vertical line with the point of interest P . the prediction of the shape of the orbit. In order to do that. and is independent of its eccentricity e. In many situations we will need to determine the time of flight between two arbitrary points along the ellipse. P ) to the total area of the ellipse. Time of Flight (TOF) in Elliptical Trajectories We have found r(θ). Equating these two expressions and expressing h in terms of the semi-major axis as h2 = µp = µa(1 − e2 ). π. the ratio of tP .

where AP is the area swept out up to the point P . which is referred to as the eccentric anomaly. 2π πab 2π∗AP AT where AB − AA is the area swept out between points A and B. MP = 2π. We start by obtaining a relation between θ and u.The various geometric quantities of the elliptical orbit have standard definitions: the position angle θ is often called the true anomaly. tP is the time of flight from the periapsis to the point P . cos u = e + cos θ . Thus when a complete orbit has been traversed. we can use equation (29) and write T OF = tB − tA = T AB − AA (MB − MA ) = T . the time t equals the period T and the mean anomaly Mπ = 2 ∗ π. The radial line of the circle for the origin O to P and the major axis of the ellipse major axis define an angle u. In addition. T πab (29) Here. From simple trigonometry. as MP = 2πtP AP = 2π .) All is needed now is an expression for the mean anomaly M as a function of the orbit parameters. When the area swept out equals the total area of the ellipse AT . 1 + e cos θ → 9 cos θ = cos u − e . 1 − e cos u (31) (30) . the mean anomaly M . The mean anomaly can also be written as MP = . If we want to determine the time of flight between two points A and B on the ellipse. we define a third anomaly. (The subscript π denotes the return to the periapsis π. we have that a cos u − r cos θ = ae or noting that r = a(1 − e2 )/(1 + e cos θ).

The height of the small triangle is b × sin(u). P . and the vertical height of the circle is a. Therefore. with the goal to find the formula for the area AP . A − A4 = (b/a) × (A1 − A2 ). P. (32) The base of the small triangle of area A4 . The curved vertical segment formed from removing the large imbedded triangle A2 from the arc segment of the circle A1 –call it A3 – is geometrically similar to the curved segment formed by removing the small triangle from the area of the swept segment of the ellipse. is r × cos(θ) = a × cos(u) − ae by equation (30). O . P. The final step in identifying the area segment swept out between point π and P is to identify the curved segment from O”. the area of the small triangle is A4 = (1/2) × (a ∗ cos(u) − e) × b × sin(u). Therefore. the final result for the area swept out by the point r moving from point π to point P is A = b/a × (A1 − A2 ) + A4 10 (34) (33) . π. the area of the large triangle formed by the angle u within the circle is A2 = a2 cos(u)sin(u)/2. the area of the large curved segment from O”. which is then added to the triangle section A4 to form the complete swept area. O. This area plus the curved segment O”. Therefore. P . The area A1 is the wedge in the circle occupied by the angle u. π is the total area swept by the point P . A1 = a2 u. Since the vertical height of the ellipse is b. the area swept out by the point r as it travels from the periapsis π to the point P .We now develop relationships between the various areas indicated on the figure. the area of the desired curved segment can by obtained from that of the corresponding segment of the circle by multiplying by b/a. Specifically. π is A1 − A2 = (1/2) × a2 × (u − Cos(u)Sin(u)).

This equation is very easy to use if we want to know the time at which the satellite is at position θ. Meriam and L. we need to solve Kepler’s equation which is non-linear using an iterative numerical algorithm such as Newton’s method. in this case. DYNAMICS. 5th Edition 3/13 (except energy analysis) 11 . if we need to find the position θ of the satellite at a given time t. On the other hand.) u − e sin u = M .And the mean anomaly is MP = 2π ∗ AP 2π × (b/a × (A1 − A2 )) + A4 = AT AT (35) Thus. The only thing required.G.L. then. Engineering Mechanics. ADDITIONAL READING J. we obtain Kepler’s equation (It took a Kepler to work this out. Kraige. (32) and (34). combining equations (29). is the calculation of the eccentric anomaly u using equation (31).