I apologize if I’ll appear pedant, but I think there are two concepts that need to be clarified.
1-No matter how many primary one simulates, Fluka results will always be given per single primary (this is not exactly true when using biasing or considering induced activation, but let’s not be too much pedant).
No matter whether you simulate 100 primaries or 100 billions primaries, Fluka will tell you how much energy is deposited in your target by one single primary. The only difference that follows from using a given number of primaries is the statistical uncertainty that comes with your estimate.
2-Fluka results then need to be normalized to a quantity that make sense for the specific study case.
If you want to know how much energy is deposited by a single bunch accelerated in your machine, then you have to multiply the Fluka result per primary for the number of primaries that are in your bunch which could be whatever 10^9, 10^10, or even, provided that you perform a proper conversion 1nA.
Getting to your specific case, if you want to know how much energy (or DPA, or dose, or whatever quantity) is deposited by one bunch that contains 10^20 particles, then you will have to multiply the Fluka results (given per single primary) time 10^20, regardless of the number of primaries you have simulated.
I hope it is more clear now and I apologize if I sounded patronizing.