Normal distribution in Hourly Arrivals (PatientArrival)

I want to use a normal distribution for each hour in the hourly arrivals table.
When I fill the table the way I think it should be, the model gives the same amount of patient throughput every time I run the model.
Is it not possible to use a normal distribution in the hourly arrivals table or am I doing something wrong?

I attached the hourly arrival table

In the help file of version 5 it clearly states that it is possible what you want but in the helpfile of version 4 (and it looks like you are using that) it is not stated so I guess there it is not possible. Maybe somebody from Flexsim can confirm this?

So maybe you can update to 5?

The normal distribution is a continuous (non discrete) distribution that generates positive and negative decimal numbers. It's of course impossible to create a fraction of a patient or even worse, a negative patient, so the Hourly Arrivals Table forces the value into a positive integer which in effect truncates the decimal number and replaces negative values with 0.

Typically you will want to use a discrete distribution like poisson() to define variance in the Hourly Arrivals Table. Another popular option is to use P() to define a discrete empirical distribution based on percentages. If you do decide to use the normal distribution, then you may want to use an expression similar to the following so that the value generated by the normal distribution is rounded to the nearest integer rather than truncate down to the nearest integer. Be aware that the resultant distribution may not fit a normal distribution at all due to capping and rounding!

maxof(0, round(normal(3.512,1.604)))

Thanks for the different solutions to my problem.
It has been solved.

PS. I've unchecked "Repeat Random Streams" to fix the problem with the same amount of patients every run.