The graphical part is interface. The computation part is what is triggered when you run sim() . Part of running a simulation might be to update inputs to a graphing block, but that does not mean that it is mandatory that the blocks appear on the screen.
Using Simulink Scopes and similar is not like using figure() and so on, in that using figure() always causes a window to pop up. Simulink Scopes record data, and will display it if you open the graphical interface -- but the graphical interface is not necessary for the model to run.
You can use commands such as get_param and related to get the state of blocks.
You can log signals.
Especially if you have configured Single Output, https://www.mathworks.com/help/simulink/gui/single-simulation-output.html then the output of sim() is a Simulink.SimulationOutput object that holds the result of the simulation, and with the way you are running, that is going to get stored in ans . You should assign the value to a variable and examine it to see what went on with your model.
