In this model we have set up an initial velocity of 10m/s at 45 degrees an a gravitational force of 9.8N at -90 degrees. Notice that the initial amount of time over which we observe the object is 20s. In that amount of time the y position of the object goes way below the reference frame origin. If the situation was that zero represented ground level, our model gives the nonsense result that the object penetrates the Earth. It is one thing to neglect air resistance (We are still doing that.) but neglecting dirt resistance is a bit much. In modeling you have to know when to quit. Rather that trying to model the ball's behavior on impact with the Earth, we can just ignore anything that happens after y=0.
Since the negative y part of the plot has no meaning, you might want to get rid of it by limiting the observation time. Move the cursor to about the point where y crosses the x axis and enter that value of time in the time limit data entry box. You may repeat this process to sort of zero in on the time when y=0.