Low pass RC filter

To see what is happening, it sometimes helps to overlap the signals on the display, then put a volt meter across the resistor. Try this experiment with a SINE wave, why is the result different?

If you sit on a swing and then give yourself a kick start, you will oscillate back and forth across the mid-point starting position. Likewise if you hold a plastic ruler over the edge of a table and jab the end, it will oscillate in a similar way. Here in your schematic, the voltage is likewise oscillating either side of the mid-point of your 10v ... ie it averages at the centre point as the seen 5v. If you now turn your pulse width time down , to say 1ms, (and so increase the frequency of its occurance), you will see greater attenuation of the volts, (highlight only your output), but it will still oscillate either side of the 5v mid-point. (The triangular shape seen is due to the input square wave being integrated, and the attenuation occurs because the cap charging/discharging does not have time to fulfil its journey before the next pulse occurs, and so the head and tail get more and more chopped off as the time width gets narrower ... ie the frequency gets higher). That shows itself as the higher frequencies disappearing in to smallness, thus leaving only the longer wave pulses (lower frequencies) being unaffected .... hence a LPF ... The place where that drastic difference occurs is where the cap/resistor RC output volts fulfils its journey ... which is about 6 x RC, or more precisely: 6.284 x RC ... ie: 2π x RC. Time is the inverse of Frequency, and so we arrive at the familiar F=1/2πRC for that cutoff frequency.

The greater attenuation of that output volts with increasing frequency (ie: less V), can also be thought of as a decrease of virtual resistance at the capacitor (ie: less R), and that is where we obtain the usage of Capacitive reactance Xc

Observe the attenuation: http://everycircuit.com/circuit/6396256113655808