# Why is $\Delta (R_i^2)$ equal to $2R_i \Delta (R_i)$

So when deriving the equation for charge, I reached a point where

$\frac{z\eta}{4\epsilon} \sum_{i=1}^N \frac{\Delta (R_i^2)}{(z^2+R_i^2)^\frac{3}{2}}$

But my teacher explains that

$\Delta (R_i^2)$ can be rearranged to $2R_i \Delta (R_i)$

but he cannot explain why. He talks about taking a differential but then it becomes $dr$ and the equation turns into an integral in terms of R, so doesn't that just counter his claim?

• I think it comes from the fact that $(y^2)'=2yy'$. Aug 5, 2018 at 18:49

Unpacking the meaning of $\Delta$ as ... $$\Delta f(t)\equiv f(t+\Delta t)-f(t)$$ We can write ... $$\begin{array} \\ \Delta x^2(t) \\=x^2(t+\Delta t) - x^2(t) \\=(x(t+\Delta t)+x(t))(x(t+\Delta t)-x(t)) \\=\left [\Delta x(t) +2x(t) \right ] \Delta x(t) \end{array}$$

So provided that $\Delta x(t) << 2x(t)$ we have ...

$$\Delta x^2(t) \approx 2x(t) \Delta x(t)$$

• I think it's a bit better to say something like "If we neglect the higher order $\Delta$ terms, i.e. $(\Delta x)^2$, because $(\Delta x)^2 << \Delta x$." Aug 5, 2018 at 19:14
• I guess the truth is that the statement is not so much a rearrangement as it is an approximation
– WW1
Aug 5, 2018 at 19:18
• What do you mean? Aug 5, 2018 at 19:19
• The OP used the phrase "can be rearranged to" , but we realize that there is an approximation being made for finite $\Delta t$
– WW1
Aug 5, 2018 at 19:23
• Yes, it's an approximation, but it will be true when $N \to \infty$. Aug 5, 2018 at 19:24

When taking differentials, $$x = f(t) \implies \mathrm dx = \frac{\mathrm df}{\mathrm dt} \, \mathrm dt$$

Let's apply the above statement. Let $u_i = R_i^2$. As $\Delta u_i$ approaches $0$, it becomes a differential $\mathrm du_i$. So we have

$$u_i = f(R_i) = R_i^2 \implies \mathrm du_i = \frac{\mathrm df}{\mathrm dR_i} \, \mathrm dR_i$$

Note that $\frac{\mathrm df}{\mathrm dR_i} = \frac{\mathrm d}{\mathrm dR_i}[R_i^2] = 2R_i$. Hence

$$\mathrm du_i = (2R_i) \, \mathrm dR_i$$

And, remembering that $u_i = R_i^2$: $$\mathrm d(R_i^2) = 2 R_i \, \mathrm dR_i$$

Which means that for small changes in $R_i^2$ (i.e. as we take the limit approaching $0$), $$\Delta(R_i^2) \approx 2 R_i \Delta R_i^2$$