I am reading the first few pages of Nakahara and refreshing my memory on physics I learned a while ago as a physics math undergrad. Nakahara defines a field $F$ to be conservative if it's the gradient of some potential $V(x)$. (standard) Then he defines the total energy as $(1/2)mv^2+V(x)$. He then states that $E$ is "often" the kinetic+potential energy, hence deserving the name "total energy." I was taught that kinetic is defined as $(1/2)mv^2$. So what does he mean here?

Addum: The exact quote of Ref. 1 is:

The function $E$, which is often the sum of the kinetic energy and the potential energy, is called the energy.


  1. M. Nakahara, Geometry, Topology & Physics, 2nd edition, 2003, Section 1.1.1.
  • 5
    $\begingroup$ Because the correct free value is $(\gamma - 1)mc^2$ (from relativity). In the limit of $v \ll c$ this is well approximated by the Newtonian form. $\endgroup$ – dmckee --- ex-moderator kitten Sep 21 '13 at 0:18
  • $\begingroup$ but the chapter is talking about classical mechanics there I think. Not to mention that if we're speaking relativistically, then the kinetic energy is then never exactly equal to $1/2mv^2$ except when $v=0$. Are you saying that in classical, it is true by definition that $1/2mv^2$ is the kinetic energy? $\endgroup$ – Jeff Sep 21 '13 at 0:23
  • 3
    $\begingroup$ It's almost a matter of definition. You can define work and energy arbitrarily, so that $work = \kappa \int_\Gamma \vec{F}\cdot\mathrm{d}\vec{r}$, which, through Newton's second law, works out to $1/2 \kappa m v^2$ when that force works on an otherwise free body, where $\kappa$ defines your work units. But, as @dmkee says, relativity and the resultant mass-energy equivalence means your unit system must also think of work on a free body begetting increased inertia, so your units will become very complicated and tangled unless you set $\kappa = 1$. Relativity unifies phenomena such that .... $\endgroup$ – Selene Routley Sep 21 '13 at 0:32
  • 2
    $\begingroup$ You haven't defined what is 'E'. Please state the question properly and provide the exact quote from the text. $\endgroup$ – guru Sep 21 '13 at 13:30
  • 1
    $\begingroup$ Related: physics.stackexchange.com/q/11905/2451 and links therein. $\endgroup$ – Qmechanic Sep 23 '13 at 11:02

The definition for point particles is:

$T = \frac{p^2}{2m}$, where $T$ denotes my kinetic energy.

  1. If you are dealing with classical physics, the momentum for a point-particle equals $\vec{p} = m\vec{v}$, which would give you $T = \frac{mv^2}{2}$.
  2. In the theory of special relativity we notice that objects become heavier when they move, the momentum becomes: $\vec{p} = {m\vec{v}\gamma}$, where $\gamma = \frac{1}{\sqrt{1-(v/c)^2}}$. Hence the kinetic energy becomes: $T={mc^2\gamma}$, this is derived below.
  3. If you try to make a non-relativistic electromagnetic Lagrangian, it's not possible to incorporate the magnetic field in terms of a potential energy. So the solution is to define the momentum equal to $\vec{p} = m\vec{v}-\frac{q\vec{A}}{c}$, where A is the vector potential so that $\vec{B} = \nabla\times\vec{A}$. This would give an kinetic energy (or kinetic term in your Hamiltonian) equal to: $T = \frac{(m\vec{v}-\frac{q\vec{A}}{c})^2}{2m}$. One often calls $m\vec{v}$ the kinematic momentum and $\vec{p}$ the canonical momentum (since this is the momentum that follows from a Lagrangian). It are the canonical momenta that are called "thé momenta".
  4. And last but not least, in quantum mechanics you work on a different kind of mathematical structure (in Hilbert-space in stead of the classical phase-space). By equivalence with wave-theory the momentum is derived as $\hat{p} = \frac{\hbar}{i}\frac{\partial}{\partial x}$, where we have an operator $\hat{p}$ which works on that Hilbertspace.

Note: from dj_mummy

Note that the given relations for kinetic energy all assume the particle was originally at rest at a given time $t=t_0$. The kinetic energy is derived by using the work-energy theorem (difference in kinetic energy = work done on the point particle). Using the laws of classical physics (Newton's second law) we get:

$T = \int\limits_{x_0}^{x_1}\vec{F}\cdot d\vec{x} = \int\limits_{t_0}^{t_1}\vec{F}\cdot \vec{v}\,dt = \int\limits_{t_0}^{t_1}\left(m\frac{d\vec{v}}{dt}\right)\cdot \vec{v}\,dt= m\int\limits_{t_0}^{t_1} \vec{v}\cdot d\vec{v} = \frac{m}{2}v^2$. Where $\vec{v}$ is the velocity at time $t_1$.

This gives us the kinetic energy in the case of classical physics! For the relativistic particle we define a 4-momentum $p^\alpha = m\frac{du^\alpha}{d\tau}$. Where $u^\alpha$ is the four-velocity and $\tau$ is the proper time. The derivation is actually done here and we get the formula above !

| cite | improve this answer | |
  • $\begingroup$ Huh? Your factors of $\gamma$ in #2 are wrong. If you fix the gamma in the equation for p, the gamma in your equation for T is still wrong. $\endgroup$ – user4552 Sep 23 '13 at 15:12
  • $\begingroup$ @BenCrowell It should be fixed by now ... $\endgroup$ – Nick Sep 23 '13 at 17:42
  • $\begingroup$ @Nick: But now it's not consistent with the assertion at the beginning that $T=p^2/2m$ in all cases, which I think is false. $\endgroup$ – user4552 Sep 23 '13 at 19:20
  • $\begingroup$ @BenCrowell that's what I get for doing to many things at the same time, I god rid of the additive constant! $\endgroup$ – Nick Sep 23 '13 at 21:40
  • 1
    $\begingroup$ @Nick: Huh? I don't understand your most recent comment. There is no additive constant. The issue is that $T=p^2/2m$ is false in the relativistic case. $\endgroup$ – user4552 Sep 23 '13 at 21:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.