Skip to content

Energy and Momentum Are One Vector

Irodov 1.381DifficultyDifficulty 3 of 5UniquenessUniqueness 4 of 5

Problem Irodov 1.381

A particle moves along the xx axis of a frame KK, carrying momentum pp (directed along that axis, with component pxp_x) and total energy EE. A second frame K′K' moves at a constant velocity VV relative to KK, along the positive direction of the shared xx axis. Writing β=V/c\beta = V/c, show that the particle’s momentum component px′p_x' and total energy E′E', as measured in K′K', are given by

px′=px−EV/c21−β2,E′=E−pxV1−β2.p_x' = \frac{p_x - EV/c^2}{\sqrt{1-\beta^2}}, \qquad E' = \frac{E - p_xV}{\sqrt{1-\beta^2}} .
The standard configuration: K' slides along the shared x axis at V, and the particle P runs along that same axis at u as measured in K.

The standard configuration: K′K' slides along the shared xx axis at VV, and the particle PP runs along that same axis at uu as measured in KK.

Given
  • particle speed uu along xx in KK
  • px=m0u/1−u2/c2p_x = m_0u/\sqrt{1-u^2/c^2}
  • E=m0c2/1−u2/c2E = m_0c^2/\sqrt{1-u^2/c^2}
  • boost speed VV, β=V/c\beta = V/c
Find
Demonstrate the stated transformation of pxp_x and EE

Solution

Step 1Write both quantities in terms of the one variable that transforms

Energy and momentum are not independent unknowns here: both are built from the particle’s velocity uu and its rest mass, and only uu changes under the boost. With γu=1/1−u2/c2\gamma_u = 1/\sqrt{1-u^2/c^2},

px=γum0u,E=γum0c2.p_x = \gamma_u m_0 u, \qquad E = \gamma_u m_0 c^2 .

Dividing one by the other removes both m0m_0 and γu\gamma_u and leaves a relation we shall use twice:

px=Euc2p_x = \frac{Eu}{c^2}
(1)
Step 2Transform the velocity, which is the only thing that moves

For motion along the boost direction the velocity transformation is

u′=u−V1−uVc2.u' = \frac{u-V}{1-\dfrac{uV}{c^2}} .

The primed energy and momentum are built from u′u' by exactly the same two definitions, with the same invariant m0m_0, so the whole demonstration reduces to evaluating γu′\gamma_{u'}.

Step 3Evaluate the primed Lorentz factor

This is the one piece of algebra in the problem. Form 1−u′2/c21-u'^2/c^2 over a common denominator:

1−u′2c2=(1−uVc2)2−(u−V)2c2(1−uVc2)2.\begin{aligned} 1-\frac{u'^2}{c^2} &= \frac{\left(1-\dfrac{uV}{c^2}\right)^{2} -\dfrac{\left(u-V\right)^{2}}{c^2}} {\left(1-\dfrac{uV}{c^2}\right)^{2}} . \end{aligned}

Expand the numerator on its own. The two cross terms in uV/c2uV/c^2 cancel:

1−2uVc2+u2V2c4−u2c2+2uVc2−V2c2=1+u2V2c4−u2c2−V2c2=(1−u2c2)(1−V2c2).\begin{aligned} 1-\frac{2uV}{c^2}+\frac{u^2V^2}{c^4} -\frac{u^2}{c^2}+\frac{2uV}{c^2}-\frac{V^2}{c^2} &= 1+\frac{u^2V^2}{c^4}-\frac{u^2}{c^2}-\frac{V^2}{c^2} \\[2pt] &= \left(1-\frac{u^2}{c^2}\right)\left(1-\frac{V^2}{c^2}\right) . \end{aligned}

Putting the factored numerator back over its denominator, with V2/c2=β2V^2/c^2 = \beta^2,

1−u′2c2=(1−u2c2)(1−β2)(1−uVc2)2,1-\frac{u'^2}{c^2} = \frac{\left(1-\dfrac{u^2}{c^2}\right)\left(1-\beta^2\right)} {\left(1-\dfrac{uV}{c^2}\right)^{2}} ,

and taking the reciprocal square root of the whole quotient gives the primed factor as a product of the two unprimed ones:

γu′=γu 11−β2(1−uVc2)\gamma_{u'} = \gamma_u\,\frac{1}{\sqrt{1-\beta^2}} \left(1-\frac{uV}{c^2}\right)
(2)
Step 4Read off the transformed energy

Multiply (2) by m0c2m_0c^2, which is the same number in both frames:

E′=γu′m0c2=γum0c21−β2(1−uVc2)=E−Euc2V1−β2=E−pxV1−β2.\begin{aligned} E' = \gamma_{u'}m_0c^2 &= \frac{\gamma_u m_0c^2}{\sqrt{1-\beta^2}}\left(1-\frac{uV}{c^2}\right)\\[2pt] &= \frac{E-\dfrac{Eu}{c^2}V}{\sqrt{1-\beta^2}} \\[2pt] &= \frac{E-p_xV}{\sqrt{1-\beta^2}} . \end{aligned}

The last line used Eu/c2=pxEu/c^2 = p_x from (1). That is the first of the two formulas asked for; it is collected with the second in the box below.

Step 5Read off the transformed momentum

Apply relation (1) in the primed frame, where it reads px′=E′u′/c2p_x' = E'u'/c^2, and substitute both the primed energy just obtained and the transformed velocity:

px′=E′u′c2=1c2⋅E(1−uVc2)1−β2⋅u−V1−uVc2=E(u−V)c21−β2=Euc2−EVc21−β2.\begin{aligned} p_x' &= \frac{E'u'}{c^2} = \frac{1}{c^2}\cdot \frac{E\left(1-\dfrac{uV}{c^2}\right)}{\sqrt{1-\beta^2}}\cdot \frac{u-V}{1-\dfrac{uV}{c^2}} \\[4pt] &= \frac{E\left(u-V\right)}{c^2\sqrt{1-\beta^2}} \\[4pt] &= \frac{\dfrac{Eu}{c^2}-\dfrac{EV}{c^2}}{\sqrt{1-\beta^2}} . \end{aligned}

The bracket (1−uV/c2)\left(1-uV/c^2\right) has canceled between the two factors, and Eu/c2=pxEu/c^2 = p_x once more, which completes the demonstration:

px′=px−EVc21−β2,E′=E−pxV1−β2p_x' = \frac{p_x-\dfrac{EV}{c^2}}{\sqrt{1-\beta^2}}, \qquad E' = \frac{E-p_xV}{\sqrt{1-\beta^2}}
(3)
E against p_xc: the particle sits at P on the hyperbola E^2-p_x^2c^2 = (m_0c^2)^2, and the boost tilts the axes to E' and p_x'c by equal angles in opposite senses. The dotted lines are the massless limit E = p_xc.

EE against pxcp_xc: the particle sits at PP on the hyperbola E2−px2c2=(m0c2)2E^2-p_x^2c^2 = \left(m_0c^2\right)^2, and the boost tilts the axes to E′E' and px′cp_x'c by equal angles in opposite senses. The dotted lines are the massless limit E=±pxcE = \pm p_xc.

Check

The transformation must leave the invariant E2−px2c2E^2-p_x^2c^2 alone, and that is a real test because it uses both formulas at once. Substituting,

E′2−px′2c2=(E−pxV)2−(px−EVc2)2c21−β2.E'^2-p_x'^2c^2 = \frac{\left(E-p_xV\right)^2 -\left(p_x-\dfrac{EV}{c^2}\right)^2c^2}{1-\beta^2} .

Expanding the two squares, the cross terms 2EpxV2Ep_xV cancel and the numerator collapses to

E2(1−β2)−px2c2(1−β2),E^2\left(1-\beta^2\right)-p_x^2c^2\left(1-\beta^2\right) ,

so the denominator divides out and E′2−px′2c2=E2−px2c2E'^2-p_x'^2c^2 = E^2-p_x^2c^2, which is the invariant restored. Two limits also check out: V→0V \to 0 returns E′=EE' = E and px′=pxp_x' = p_x, and putting u=Vu = V, so that the particle is at rest in K′K', gives px=EV/c2p_x = EV/c^2 and hence px′=0p_x' = 0, as it must.

Remark

Compare the pair with the coordinate transformation x′=(x−Vt)/1−β2x' = \left(x-Vt\right)/\sqrt{1-\beta^2} and ct′=(ct−βx)/1−β2ct' = \left(ct-\beta x\right)/\sqrt{1-\beta^2}: they are the same equations with E/cE/c playing the part of ctct and pxp_x the part of xx. That is what is meant by calling (E/c, p)\left(E/c,\,\vect p\right) a four-vector, and it is why the figure above can borrow a spacetime diagram wholesale. The derivation divided by m0m_0, yet the result needs no mass: hold EE fixed and let m0→0m_0 \to 0, so that u→cu \to c and (1) becomes the massless relation px=E/cp_x = E/c. Both formulas are linear in EE and pxp_x, so they pass to that limit unchanged, and for a photon the energy one reads E′=E(1−β)/(1+β)E' = E\sqrt{\left(1-\beta\right)/\left(1+\beta\right)}, the relativistic Doppler shift.