I am working on the following physics problem and have run into some trouble
The figure above shows particles $1$ and $2$, each of mass $m$, attached to the ends of a rigid massless rod of length $L_1 + L_2$, with $L_1 = 20cm$ and $L_2 = 80cm$. The rod is held horizontally on the fulcrum and then released. What are the magnitudes of the initial accelerations of (a) particle $1$ and (b) particle $2 \space ?$
My Approach:
So I first considered the net torque when the system is at rest so that I could get to the angular acceleration using the equation $\tau_{net} = I \alpha$. Given that particle $2$ would induce clockwise motion, it is given a negative sign while particle $1$ is given a positive sign because it induces counter-clockwise motion so $\tau_{net} = F_{t2}r_2 – F_{t1}r_1 = I \alpha$. (Where $F_{ti}$ represents the tangent force acting on particle $i$)
Solving for alpha we have that $\alpha = \frac{F_{t2}r_2 – F_{t1}r_1}{I} = \frac{mgL_2 - mgL_1}{I}$. The next step is then to find the rotational inertia. Now after consulting with my solutions manual I see that this can be found by simply treating the fulcrum as the axis of rotation, but I didn't see this approach when solving the problem. Instead I used the parallel axis theorem $I = I_{com} + Mh^2$. Now even though this approach is a waste I'm trying to figure out why I didn't arrive at the same answer anyway, so I've included my work computing the rotational inertia in this way.
Computing Rotational Inertia Using Parallel Axis Theorem
First I computed the center of mass of the rod as follows: $x_{com} = \frac{m_w * 0 + m_w* 0.80 m}{m_w + m_w}= 0.4m$ (Note: I use $m_w$ to denote mass while I use $m$ to denote distance). Next I computed $I_{com}$ as follows: $I_{com} = \Sigma \space m_{wi} \cdot \space r_i^2 = 2\space m_w \cdot (0.4 m)^2$.
And by the parallel axis theorem $I = I_{com} + Mh^2 = 2\space m_w \cdot (0.4 m)^2 + 2\space m_w (0.2m)^2$. After inputting this value into the original equation for angular acceleration I arrive at an invalid value. What have I done wrong? Also, how do we know definitively when to use the parallel axis theorem?
Any help understanding my problem would be appreciated greatly
Note: When either rotational inertia value is inputted into the angular acceleration formula the masses will cancel so using the correct approach and canceling the mass $\frac{I}{m_w} = (0.2m)^2 + (0.8m)^2 = 0.68 m^2$ while using my original approach I have $\frac{I}{m_w} = 2((0.4 m)^2 + (0.2m)^2) = 0.4m^2$