Short answer is that you should use the cumulative rainfall in inches (to calculate runoff for each time step) and not the intensity in inch per hour for that time.
Using runoff curve number, we need to account for initial abstraction.
$$\begin{align}Q &= 0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ for~~ P \leq I_a\\
Q &= \frac{(P-I_a)^2}{P-I_a+(\frac{1000}{CN}-10)} ~~~~~~~~~~~~~~~~ for~~ P > I_a\\
\text{where:}\\
Q &= \text{runoff (in)} \\
P &= \text{rainfall (in)} \\
I_a &= \text{initial abstraction (in)} \\
CN &= \text{runoff curve number}\end{align}$$
There is an empirical equation which relates initial abstraction to potential maximum retention (which is also related to curve number).
$$\begin{align}I_a &= 0.2S\\
S &= \frac{1000}{CN}-10\\
\text{where:}\\
S &= \text{potential maximum retention (in)} \\\end{align}$$
Using these equations to replace initial abstraction in the runoff calculation, we get:
$$\begin{align}Q &= 0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ for~~ P \leq I_a\\
Q &= \frac{(P-\frac{200}{CN}+2)^2}{P+\frac{800}{CN}-8} ~~~~~~~~~~~~~~~~~~~~~~~~~ for~~ P > I_a\\\end{align}$$
You can read more here: HEC-HMS Technical Reference Manual