I was first shown VR in a Whirlwind 10, as part of the RAF's basic rotary course.
This was on the same sortie as the retreating blade stall demo (which I'd in fact already seen, courtesy of my instructor who got to Vne running in for a quickstop then pulled too hard in a turn, at about 150 feet agl. Thankfully, the turn was a right handed one, so the Whirlybird flicked upright, rather than inverted, or I might not be here today.
Anyway, VR! We were required to climb to 10,000 feet and wear parachutes for this instructional sortie. My QHI must have chickened out at the incipient stage because I don't remember anything particularly exciting happening. He wouldn't do the demo again!
Next time I saw VR was about twelve years later in a Puma HC1, during a night operational sortie. We were doing a certain job where we were required to hover as high as we could to avoid detection from "unfriendlies" on the ground. Gaining the hover (more or less on instruments) involved flying a slightly climbing, into-wind quickstop (we found that we made less blade slap that way) at around 10,000 feet. If the aircraft fell out of the hover attempt due to lack of power, we were too high for the ambient conditions so we would fly a racetrack, flying forwards and down by a thousand feet or so, then try again. When we subsequently achieved a hover, we then checked max continuous engine power wasn't being exceeded (normally the limit was N1s, the Turmos ran quite cool). If it was, again we move fly forwards and down until we were in limits. If we had some spare power, and our own folks on the ground advised they could hear us, we would pull to the N1 limit and climb vertically until the aircraft settled at it's preferred altitude.
I was acting as co-pilot on this occasion and we had been flying for about four or five hours, so we were getting tired (we wore NVGs, too, which didn't help).
One of my ex-students from some years before was flying the aircraft and operating as captain. There was little or no wind (we would obviously try and find the wind if possible, to make the job easier and to get higher). We had a "cross hairs hovermeter" but as it was Doppler based it often was inaccurate at the altitudes we operated at.
As the aircraft partly "settled" in the first attempt at a free air hover, I felt through the seat of my pants that we were beginning to move backwards slightly, although we had no visual references. The aircraft then suddenly pitched slightly nose up (I think this was the rearwards airflow hitting the top of the stabiliser).
Suddenly, the instruments came alive. Unfortunately, not the ASI, but the altimeter began rapidly unwinding and the VSI needle went very rapidly to the bottom stop (2500 FPM plus). The aircraft then exhibited classic signs of VR, randomly pitching, rolling and yawing. I called "Airspeed!" twice, no apparent recovery action was taken. I then called it a third time, and gave a very strong hint by pushing forwards on the cyclic with the palm of my hand. The handling pilot then woke up to what was required and as the ASI was showing some action, he recovered as for an IMC "U.P."
We lost a lot more than 3,000 feet, in a very short time. The debrief was short as we were very tired after flying all night.