I am given $\omega\left ( k \right )=\left ( \frac{C+H}{m}\pm \frac{1}{m}\sqrt{C^{2}+H^{2}+2 CH Cos\left ( k a \right )} \right )^{1/2}$.

It is mentioned that this reduces to

$\omega\left ( k \right )=\sqrt{\frac{CH}{2m\left ( C+H \right )}}a\left | k \right |$ when $\lambda \rightarrow \infty$ for $\lambda = \frac{2 \pi}{k}$.

After 5 hours of frustrating attempt, I am unable to progress further.


Note that in the event that $\lambda \rightarrow \infty, k\rightarrow 0$.

Thence, it can be expected that $ka \approx 0$ and so $Cos\left ( ka \right )\approx 1$.

It is useful to recall that for small angle $\left ( ka \right ), Cos\left ( ka \right )\approx 1-\frac{\left ( ka \right )^{2}}{2}$.

Expanding this we obtain at

$\left ( \frac{C+H}{m}-\frac{1}{m}\sqrt{C^{2}+H^{2}+2CH} \right ) =0$

which isn't what I am looking for.

Any help is appreciated.



As you wrote, you need to use $$\cos(ka)=1-\frac{a^2 k^2}{2}+O\left(a^4\right)$$ which makes $$C^{2}+H^{2}+2 CH \cos\left ( k a \right )=(C+H)^2-CH a^2k^2+O\left(a^4\right)$$ $$\sqrt{C^{2}+H^{2}+2 CH \cos\left ( k a \right )}\sim(C+H)\sqrt{1-\frac{CHa^2k ^2}{(C+H)^2}}\sim(C+H)\left(1-\frac{CHa^2k ^2}{2(C+H)^2} \right)$$

Just continue.

  • $\begingroup$ But if k is small, why isn't it valid to neglect $(ka)^{2}$? $\endgroup$ – Mathematicing Oct 31 '17 at 6:42
  • $\begingroup$ @Mathematicing. Just because you want an approximation ! Otherwise, as you did, you would get $0$ just as you did. Just try with a calculator using, say, $C=H=1$ and $ka=\frac \pi {24}$ for which $\cos(ka)=\frac{1}{2} \sqrt{2+\sqrt{2+\sqrt{3}}}$ $\endgroup$ – Claude Leibovici Oct 31 '17 at 6:49
  • $\begingroup$ I'm also having a bit of trouble understanding to second approximation equality. Is this a result of purely algebraic manipulation of approximation? $\endgroup$ – Mathematicing Oct 31 '17 at 6:58
  • $\begingroup$ @Mathematicing. You need to continue the process $$\frac{C+H}{m}-\frac{1}{m}\sqrt{C^{2}+H^{2}+2 CH \cos\left ( k a \right )}=\frac {CHa^2k^2}{2m(C+H)}$$ and you need the square root of that. All of that is based on manipulations using Taylor series or binomial expansion. $\endgroup$ – Claude Leibovici Oct 31 '17 at 7:06
  • $\begingroup$ I'm afraid I am unable to arrive at the final approximation in your last line. $\endgroup$ – Mathematicing Oct 31 '17 at 7:22

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.