Home

Search

Collections

Journals

About

Contact us

My IOPscience

The work-energy theorem: the effect of varying mass

This article has been downloaded from IOPscience. Please scroll down to see the full text article. 2001 Phys. Educ. 36 61 (http://iopscience.iop.org/0031-9120/36/1/311) View the table of contents for this issue, or go to the journal homepage for more

Download details: IP Address: 129.174.55.245 The article was downloaded on 12/09/2011 at 08:10

Please note that terms and conditions apply.

FEATURES

The workenergy theorem: the effect of varying mass

Ronald Newburgh
Extension School, Harvard University, Cambridge, MA 02138, USA

Abstract This paper examines the workenergy theorem of classical physics and applies it to two situations with non-constant mass. The rst is that of a rocket burning fuel. The second is that of a proton in a particle accelerator reaching relativistic speeds. The analysis leads to a reformulation of the workenergy relation in terms of momentum and velocity rather than force and displacement. This allows graphical determinations of kinetic energy that are both simple to make and clear even to beginning students. The use of graphical integration obviates the necessity of using calculus for the non-constant mass situations. An unexpected result is the invariance of this reformulation even for the relativistic case.

Introduction
A student recently asked a question that prompted me to re-examine the workenergy theorem and its relation to kinetic energy. His question was quite simple. What is the meaning of a plot of velocity versus momentum for a body since an area in momentumvelocity space is an energy? The question is one with several far-reaching implications. The rst of these leads to a reexamination of the workenergy theorem and allows a graphical interpretation of the kinetic energy. The graphical interpretation offers a fresh look at kinetic energy that has considerable pedagogic merit, as judged from student response. This result alone would justify the analysis. However, the graphical technique has shown itself to be more general and powerful, allowing, as it does, treatment of time-varying mass. Graphical integration allows one to determine net work or change in kinetic energy quite simply. The paper examines both the rocket problem with nonconstant mass and particles moving at relativistic velocities. This approach makes these two topics far more accessible to beginning students than do more conventional techniques.
0031-9120/01/010061+04$30.00

Perhaps even more important is the result that a simple equation in elementary mechanics, namely the one relating work, momentum and velocity, is equally valid in relativistic mechanics. Let me stress that this does not mean that there are two equations that are covariant but rather that there is one equation that is the same in both relativistic and classical mechanics. This was a completely unexpected consequence of the analysis.

Net work and momentum

Let us begin with the denition of the net work done on a body undergoing a displacement. For simplicity we shall consider a net force acting parallel to the displacement. The innitesimal net work, dWnet , is given as dWnet = Fnet dx (1)

where Fnet is the net force and dx the innitesimal displacement. The total net work for a body going from x1 to x2 is obtained from integrating
PHYSICS EDUCATION

2001 IOP Publishing Ltd

61

R Newburgh

equation (1) or Wnet =

x2 x1

Fnet dx.

(2)

We can plot this force versus displacement and produce gure 1. The net work done on the object is the area under the curve. Note that an area in the forcedisplacement plane has units of newtons metres, or joules. Rewriting equation (1) leads to another view of the net work. Using Newtons second law, we write dW = (dp/dt)dx = dp(dx/dt) = (dp)v or dW/dp = v. Integrating equation (3) gives the total work as W = v dp (3a) (3)

Figure 1. Plot of an arbitrary force versus displacement. The area under the curve from x1 to x2 represents the change in kinetic energy over that interval.

in which p is momentum, t time and v velocity. This is a less common but equally valid form of the work equation. We shall now apply this form of the equation to several quite different situations.

Work and constant mass

For a body whose mass is constant the momentum is always linear with velocity, since p = mv. (4)

Figure 2. Plot of velocity versus momentum for a body of constant mass. Since the mass is constant, the momentum is a linear function of velocity. The area under the graph represents the net work or the change in kinetic energy as the body accelerates from zero velocity to a maximum value.

In other words, the momentum has the same functional dependence on time as does the velocity. Therefore plotting velocity versus momentum gives the linear graph of gure 2. The plot applies no matter how the force varies with time, as long as the mass is constant. An area in the momentumvelocity plane has the units (newtons seconds)(metres/seconds) or joules. Evaluating the area for the range from zero velocity to the maximum velocity, we nd the net work to be 1 1 Wnet = 2 vmax pmax = 2 m(vmax )2 . (5) This last term is the well-known expression for the kinetic energy. More generally we can write Wnet = KE. (5a)

The rocket with varying mass

Consider a rocket accelerated vertically. Its engine has a continuous exhaust red straight out from the rocket tail. The exhaust velocity of the gas with respect to the rocket is u. The fuel burns at the constant rate such that dm/dt = . (6)

The gravitational force is mg . For simplicity we shall take g to be constant and neglect air resistance. This is the only external force acting on the system. Therefore we may write the net force as Fnet = dp/dt = m dv/dt + u dm/dt = m dv/dt u = mg. (7)

The net work therefore equals the change in the kinetic energy of the body during its displacement.
62
PHYSICS EDUCATION

The workenergy theorem

The mass m is a function of time and may be written as m = m0 t = m0 [1 (/m0 )t ] in which m0 is the initial mass. Rewriting equation (7), we obtain dv = g dt + (/m0 )u dt. 1 t/m0 (9) (8)

Integrating, we nd for the velocity v = gt u ln[1 (/m0 )t ]. (10)

Combining this with equation (8) yields for the momentum p = mv = m0 [1 (/m0 )t ] {gt u ln[1 (/m0 )t ]}.
Figure 3. Plot of velocity versus momentum for the rocket whose specications are given in table 1. The area under the graph represents the net work done on the rocket. The graph is concave upwards because it does not include the work done on the ejected fuel.

(11)

To provide a numerical example take the exhaust velocity u to be 2 103 m s1 , the initial mass m0 to be 100 kg, and the rate of change of mass to be (m0 /100) kg s1 . Table 1 gives the values for mass, velocity and momentum for a range of times from 1 to 50 seconds. Figure 3 is a plot of velocity versus momentum for these values. By applying equation (3a ) it is easy enough to calculate the area under the graph of gure 3. As before, it represents the net work done on the system. However, one must remember that the system changes mass continuously by expelling gases. This expelled gas carries energy supplied previously by the motor. As a result work done on the system does not equal the change in kinetic energy of the rocket alone. It must also include the continuous change in kinetic energy of the exhaust gas as well. This is the reason that the graph of gure 3 has a concave upward form.
Table 1. The values of mass, velocity and momentum from 1 to 50 s for the rocket. Time (s) 1 5 10 20 30 40 50 Mass (kg) 99 95 90 80 70 60 50 Velocity (m s1 ) 10.3 53.6 113 250 419 630 896 Momentum (kg m s1 ) 1.02 103 5.09 103 1.02 104 2.00 104 2.94 104 3.78 104 4.48 104

Particle moving at relativistic velocities

An elementary particle in an accelerator such as a cyclotron can reach relativistic speeds. Since mass is velocity dependent at such speeds, the momentum (as with the rocket problem) is no longer linear with velocity. The expression for the relativistic momentum is p = m0 v/(1 v 2 /c2 )1/2 (12)

in which c is the velocity of light. Figure 4 is a plot of velocity versus the momentum divided by the rest mass (p/m0 ). Notice how different equation (12) is from equation (11), as is gure 4 from gure 3. Unlike gure 3, gure 4 is concave downward. The momentum increases more rapidly than it does in the classical case of constant mass. It also differs from the rocket problem in that there is no physical loss of mass during the acceleration process but rather an increase. How do we interpret the area under the graph? The total energy, E , of a relativistic body may be written as E = mc2 = m0 c2 /(1 v 2 /c2 )1/2 . Combining equations (12) and (13) leads to E = c(p2 + m0 c2 )1/2 .
PHYSICS EDUCATION

(13)

(14)
63

R Newburgh

Figure 4. Plot of velocity expressed as a fraction of light velocity versus the momentum divided by the rest mass. The graph represents the change in relativistic kinetic energy as the proton accelerates.

The relativistic kinetic energy, T , equals the total energy minus the rest energy or T = E m0 c 2 . Differentiating equation (15) we nd dT /dp = dE/dp d = [c(p 2 + m0 c2 )1/2 m0 c2 ] dp = pc/(p2 + m0 c2 )1/2 = pc2 /c(p 2 + m0 c2 )1/2 = pc2 /E = v. (16) (15)

kinetic energy, this change does not always equal the net work. We have treated three cases, two classical and one relativistic. The mass is constant for the rst case (gure 2), and the velocity is linear with momentum. Here the net work done on the object equals its change in kinetic energy. The second case (gure 3) treats a rocket with non-constant, continuously varying mass. The change in mass is a consequence of the continuous mass loss from the fuel combustion. The change in kinetic energy does not equal the net work, since some of the net work is done on the lost mass. The kinetic energy at any instant is that of the rocket plus the amount of fuel on board, not that of the rocket plus the original fuel. The relativistic case (gure 4) is like the rst one, in that no mass is lost. However, because of the relativistic mass increase with velocity, the momentum increases more rapidly than if the momentum were linear with velocity. Moreover, all the work done on the particle by the accelerator equals the increase of kinetic energy dened relativistically. Thus, once again, net work equals the change in kinetic energy. Perhaps the most striking fact of this analysis is the universality of equations (3), (16) and (17). The only difference between the relativistic and classical cases is the generalization of the denition of kinetic energy. As I implied in the introduction, this generalization is a consequence of the invariance of the equation relating work, momentum and velocity in both classical and relativistic physics.

Equation (16) is identical with equation (3) except for the change in the denition of kinetic energy. The area under the graph is T = v dp. (17)

Acknowledgments
I wish to thank William Boone, a student in my class at the Harvard Extension School, who rst raised the question of the meaning of the pv plot. This paper is an attempt to answer his question.
Received 17 April 2000, in nal form 15 August 2000 PII: S0031-9120(01)13415-5

Rosser [1] has an especially good treatment of this material.

Discussion
Figures 2, 3 and 4 are plots of velocity versus momentum and, as such, represent the change in kinetic energy according to equations (3a ) and (17). Though they do represent the change in

Reference
[1] Rosser W G V 1964 Introduction to the Theory of Relativity 1st edn (London: Butterworths) pp 1845

64

PHYSICS EDUCATION