There is a derivation for kinetic energy using calculus:

\begin{align} \Delta E_k&=\int_{x_0}^{x_1} F \ {\rm d}x \\ &= \int_{x_0}^{x_1} ma \ {\rm d}x \\ &= m \int_{x_0}^{x_1} \frac{{\rm d}v}{{\rm d}t} \ {\rm d}x \\ &= m \int_{x_0}^{x_1} \frac{{\rm d}v}{{\rm d}x}\frac{{\rm d}x}{{\rm d}t} \ {\rm d}x \\ &= m \int_{v_0}^{v_1} \frac{{\rm d}x}{{\rm d}t} \ {\rm d}v \\ &= m \frac{v^2}{2}\bigg\vert_{v_0}^{v_1} \end{align}

How to justify change of limits in this case: $t_1 \rightarrow v_1$ and $t_0 \rightarrow v_0$? I suppose it's a simple change of variables, but I'm not sure how to do it.

  • 1
    $\begingroup$ This is a question about mathematics, not physics. (Why were the limits originally $t_0, t_1$ when the integral was wrt $x$?) $\endgroup$ Commented Dec 8, 2017 at 19:51
  • $\begingroup$ You changed the integrand, so you need to change the boundaries. Nothing to see here! The derivation is 100 % correct. $\endgroup$
    – Gert
    Commented Dec 8, 2017 at 20:41

2 Answers 2


I think it would be more helpful if you would go back to "integration by parts". When you change the $dx$ to $dv$, you're changing what you have "respect to" in terms of the integral. Going back to basic calculus, remember how when you let $u$ equal some variables and input it into an integral, you had to change the bounds. This is the same situation.

Let $T = \frac{1}{2}mv^2$,

Therefore, $$\frac{dT}{dt} = \frac{1}{2}m\frac{d}{dt}(v^2)$$ $$=\frac{1}{2}m[\dot{v} \cdot v + v \cdot \dot{v}]$$ $$=\frac{1}{2}m[2(\dot{v} \cdot v)] = m\dot{v} \cdot v = F \cdot v$$

Now for the integration by parts, bring the dt to the right and integrate,

$$dT = F \cdot v \ dt$$ $$T = \int\limits_{t_0}^{t_1} F \cdot v \ dt$$

Recall $v \ dt = dr$,

$$T = \int\limits_{t_0}^{t_1} F \cdot dr$$ $$=\int\limits_{t_0}^{t_1} m \dot{v} \cdot dr$$ $$=m\int\limits_{t_0}^{t_1} \frac{dv}{dt} dr$$ $$=m\int\limits_{t_0}^{t_1} \frac{dv}{dr}\frac{dr}{dt} dr$$ $$=m\int\limits_{v_0}^{v_1}dv \ \frac{dr}{dt} $$ It's at this point where you must change the bound variables because $dv$ is what you're integrating with respect to. Thinking about it in the $u$ and $du$ calculus, here, I used $v$ instead of $u$.

Recall $\frac{dr}{dt} = \dot{v}$,

$$T = m \int\limits_{v_0}^{v_1} \dot{v} \ dv$$ $$= m[\frac{v_1^2}{2} - \frac{v_0^2}{2}]$$


B.C.'s answer is interesting. I never would have thought of using integration by parts to answer this question.

Here's an alternative way to derive it just using the chain rule.

$$T = \frac{1}{2}m{v^2}$$

Assuming the mass is constant: $$\frac{{dT}}{{dx}} = \frac{m}{2}\frac{d}{{dx}}\left( {{v^2}} \right)$$

$$\frac{d}{{dx}}\left( {{v^2}} \right) = \frac{{d\left( {{v^2}} \right)}}{{dv}}\frac{{dv}}{{dx}} = 2\underbrace v_{\frac{{dx}}{{dt}}}\frac{{dv}}{{dx}} = 2\frac{{dx}}{{dt}}\frac{{dv}}{{dx}} = 2\frac{{dv}}{{dt}} = 2a$$

So from the previous two equations we get

$$\frac{{dT}}{{dx}} = \frac{m}{2}\frac{d}{{dx}}\left( {{v^2}} \right) = \frac{m}{2} \cdot 2a = ma = F$$

And finally: $$\int_{{x_0}}^{{x_1}} {Fdx} = \int_{{x_0}}^{{x_1}} {\frac{{dT}}{{dx}}dx = T\left( {v\left( {{x_1}} \right)} \right)} - T\left( {v\left( {{x_0}} \right)} \right) = \frac{1}{2}mv_1^2 - \frac{1}{2}mv_0^2$$

So basically... the reason we are able to substitute $v_0$ and $v_1$ for $x_0$ and $x_1$ as our limits of integration is because $$F = \frac{{dT\left( {v\left( x \right)} \right)}}{{dx}} = \frac{{dT}}{{dv}}\frac{{dv}}{{dx}}$$

For more info on integration by substitution, plus some practice problems, check out these notes


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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