Michael F. Zimmer
Abstract:In lieu of abstract, first paragraph reads: Six months after the author derived a constant of motion for a 1D damped harmonic oscillator [1], a similar result appeared by Liu, Madhavan, and Tegmark [2, 3], without citing the author. However, their derivation contained six serious errors, causing both their method and result to be incorrect. In this Comment, those errors are reviewed.
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is to correct and criticize the mistakes made by Liu, Madhavan and Tegmark in their study of one - dimensional damped harmonic oscillators. Specifically, Zimmer pointed out six serious mistakes in their article and explained how these mistakes led to the incorrectness of their methods and results. The following is a detailed summary of these mistakes:
1. **Inconsistency between the equation of motion and the solution**:
- The solutions for position \(x(t)\) and momentum \(p(t)\) given by Liu et al. (formula 2) do not satisfy the equation of motion (formula 1). For example:
\[
\frac{d}{dt}\left[e^{-\gamma t}\cos(t + \varphi)\right]\neq e^{-\gamma t}\sin(t + \varphi)
\]
- This is because the derivative operation does not conform to the Leibniz product rule.
2. **Factor error of the damping coefficient \(\gamma\)**:
- In formula 1, the damping coefficient \(\gamma\) should be multiplied by 2, or all subsequent occurrences of \(\gamma\) should be replaced with \(\gamma/2\).
3. **Lack of the pseudo - frequency \(\omega\)**:
- They ignored the pseudo - frequency \(\omega\) introduced by damping and only used the natural frequency \(\omega_0\). The correct form should be:
\[
x(t)=Ae^{-\gamma t}\cos(\omega t + \varphi),\quad\text{where}\quad\omega = \sqrt{\omega_0^2-\gamma^2}
\]
4. **Definition problem of the complex variable \(z\)**:
- Using the corrected \(x(t)\) and \(p(t)\), the complex variable \(z = x+ip\) should be written as:
\[
z = Ae^{-\gamma t}\left[\cos(\omega t+\varphi)-i\sin(\omega t+\varphi+\beta)\right]
\]
- where \(\sin\beta=\gamma\) and \(\cos\beta = \omega\). This makes their method no longer applicable.
5. **Error in the scope of application**:
- Liu et al. only considered the under - damped case (\(\gamma < \omega_0\)), but they also used the critically damped (\(\gamma=\omega_0\)) and over - damped (\(\gamma > \omega_0\)) cases in their numerical tests, which is incorrect.
6. **Improper handling of the logarithmic function**:
- When deriving \(\log H\), they did not correctly handle the complex logarithmic function \(\log(e^{i\theta})\), which requires restricting the range of \(\theta\) and using the Riemann surface.
In addition, Liu et al. claimed to have found a constant \(H_1\), but in fact this so - called "constant" changes over time, as shown in Figure 1. Therefore, their method not only has errors, but even if these errors are corrected, it cannot be continued to be used.
Overall, this review article aims to point out the mistakes made by Liu et al. in their study of one - dimensional damped harmonic oscillators and emphasize the importance of these issues to avoid similar mistakes from occurring again in future research.