On September 14, 2004, approximately 1620 mountain daylight time, a Robinson R22, N777UH, piloted by a private pilot, sustained substantial damage during an autorotation to an open field following a loss of engine power near Spanish Fork, Utah. Visual meteorological conditions prevailed at the time of the accident. The local instructional flight was being conducted under the provisions of Title 14 CFR Part 91 without a flight plan. The flight instructor sustained minor injuries, and the pilot receiving instruction was not injured. The flight originated at Mt. Pleasant (43U), Utah, approximately 1615. According to the flight instructor's written statement and telephone interview, the helicopter was at 6,000 feet, flying at 70 knots, and exiting Spanish Fork Canyon. The helicopter was about 5 miles east of the Spanish Fork-Springville Airport (U77) on a long final approach to runway 30. After performing pre-landing checks that included warning lights extinguished, rotor and engine tachometers and other pertinent instruments in the green arc, and carburetor heat outside the yellow arc, the instructor "pulled the carb[uretor heat] up a little, and then...pushed the trim in to prepare for landing." Shortly thereafter, the engine lost power. The collective control was lowered and the instructor took control of the helicopter. The LOW RPM light illuminated and the warning horn sounded. The instructor noticed the tachometer was "in the green" and manifold pressure (MAP) was in the red (25 inches) arc. He attempted an autorotation into an open field. He flared and the helicopter landed hard "with a slight nose right yaw." The helicopter bounced once, and then rolled over on its right side. The tail boom and tail rotor were severed. According to the pilot's written statement, the helicopter yawed "hard to the left" when the engine lost power. He also noticed MAP was in the red arc, approximately 25-26 inches, and "there was no engine sound whatsoever." When the instructor flared for landing, the helicopter yawed slightly to the left. Forward speed and ground contact angle, however, caused the helicopter to roll over on its right side. The next day, the helicopter was trucked to the Robinson factory in Torrance, California, where, on October 1, 2004, the engine was functionally tested to full power. According to Robinson Helicopter and Textron Lycoming, nothing was found that would have been causal to the power loss. In an electronic mailing between NTSB's investigator-in-charge and the FAA Long Beach office, the question as to whether carburetor ice was raised. Robinson Helicopter's accident investigator didn't think so, citing the large temperature-dew point spread. The following are excerpts from the Robinson R22 "Pilot's Operating Handbook: "CAUTION. The R22 has a light, low-inertia rotor system. Most of the energy required for an autorotation is stored in the forward momentum of the aircraft, not in the rotor. Therefore, a well-timed cyclic flare is required and rotor RPM must be kept in the green until just before ground contact. "CAUTION. During simulated engine failures, a rapid decrease in rotor RPM will occur, requiring immediate lowering of collective control to avoid dangerously low rotor RPM. Catastrophic rotor stall could occur if the rotor RPM ever drops below 80 percent plus 1 percent per 1000 feet of altitude."