On August 23, 2024, about 1130 eastern standard time, a Bell Textron Canada 505 helicopter, N945TC, was substantially damaged when it was involved in an accident near Marianna, Florida. The student pilot and the flight instructor were uninjured. The helicopter was operated as a Title 14 Code of Federal Regulation Part 91 instructional flight. ADS-B data showed that the helicopter departed Dothan Regional Airport (DHN), Dothan, Alabama, flew to the Marianna Municipal Airport (MAI), Marianna, Florida, then performed several circuits in the traffic pattern at MAI. According to the flight instructor and the student pilot, the student pilot then performed a practice autorotation to touchdown on the runway. The flight instructor reported that the student pilot initiated the landing flare about 80 ft agl with a “good nose-up pitch to bleed airspeed without any aggressive rotor buildup.” The flight instructor stated that, as the helicopter began to settle to about 10 to 15 ft agl, the student pilot applied increased collective with some forward nose cyclic for a level landing at a touchdown speed of about 20 to 25 kts. Both the flight instructor and the student pilot reported that the landing on the runway was soft with no bounce or drift and good heading control as it entered the slide. As the helicopter continued to slide down the runway with decreasing groundspeed, it began to bounce laterally and shake “violently” with increasing severity before it came to a complete stop. A postaccident examination of the helicopter revealed that the tailboom doublers at the attachment point to the truss assembly were deformed. The helicopter was equipped with an avionics suite that recorded various helicopter position, engine, flight control, attitude, and acceleration parameters. The system automatically adjusted the vertical acceleration recorded values to remove the standard 1G of gravity. The data revealed that, just before the helicopter’s IDLE/FLY switch was switched to the idle position for the simulated autorotation, the helicopter was at an altitude of about 925 ft agl in a slightly nose-down attitude with a slight left bank and an airspeed of about 75 kts. The collective was initially positioned at 33%, the rotor rpm and power turbine speed (NP) were both at 104%, and the gas generator speed (NG) was at approximately 91%. The collective was then lowered to 10%, resulting in a decrease in NG to 82%, while the rotor rpm and NP both remained constant at 104%. During this transition, helicopter’s pitch attitude varied between 2° and 5° nose down, and its airspeed increased to 80 kts. The IDLE/FLY switch was then put into the IDLE position for the simulated autorotation, and the engine responded by going to idle. The helicopter’s pitch was increased, and its airspeed slowed to about 70 kts, which was consistent with the Bell 505 Rotorcraft Flight Manual’s recommended speed for best glide. The helicopter continued to operate near 70 kts as it began a descent toward the runway. As the helicopter descended through about 125 ft agl, its pitch angle leveled off. When the helicopter was about 60 ft agl, its pitch angle increased to about 20° nose up, and the helicopter’s vertical descent rate decreased from about 1,600 ft per minute (fpm) to about 800 fpm. During the flare, the rotor rpm was about 105% and remained stable above 100%, and the collective remained relatively unchanged between 12% and 16% during most of the descent toward the ground. When the helicopter was approximately 30 ft agl, the collective began to increase sharply to 38%, and the rotor rpm began to drop below 100%. The data showed that, as the helicopter touched down on the runway, the collective continued to be increased, and the rotor rpm decreased to about 67%. Shortly after touchdown, rapid oscillations in helicopter pitch, roll, and yaw developed, and several divergent oscillations in the vertical, lateral, and longitudinal accelerations occurred. These oscillations continued for several seconds, with a peak vertical acceleration of 3.631 Gs, a peak lateral acceleration of 2.718 Gs, and a peak longitudinal acceleration of -2.165 Gs. At the time the peak vertical and lateral accelerations were recorded, the rotor rpm was about 45%, and the peak longitudinal acceleration occurred when the rotor rpm was about 41%. According to the helicopter manufacturer’s review of the data, the helicopter’s horizontal speed, rate of descent, attitude, and vertical acceleration at touchdown were consistent with the parameters for a power-off landing. The manufacturer determined that the subsequent divergent oscillations of the vertical, lateral, and longitudinal accelerations, as well as the oscillation frequencies of the helicopter’s pitch and roll angles, closely matched the helicopter’s rotor vibration frequency at the reduced rotor rpm. A review of the Bell 505 Rotorcraft Flight Manual revealed the procedure for an in-flight engine failure (requiring an autorotation) prescribed maintaining heading and attitude while lowering the collective to maintain the required 90% to 111% rotor rpm. The procedure also stated that, after landing, the collective should be lowered smoothly. A review of the operator’s flight training syllabus revealed that, for a standard autorotation to a touchdown, the standard was to maintain rotor speed within the rotorcraft flight manual-specified limits throughout the descent and flare.