A Fokker 100, operated by Alpi Eagles on behalf of Air Dolomiti, suffered a runway excursion accident on landing at Barcelona Airport (BCN), Spain. Flight DLA2708 had departed Venice, Italy with 39 passengers and a crew of five. The flight was uneventful and the weather at Barcelona was fine. Wind was calm, there was no cloud ceiling and visibility was more than 10 kilometres. The co-pilot was acting as the flying pilot, maintaining a speed of some 140-145 kt, some 9-14 knots above the reference speed. The landing configuration was set with 25ø of flap. In the approach to runway 25 the flight was conducted manually. The aircraft’s wheels touched down at a point some 300 m from the threshold of runway 25 at a ground speed of 128 kt. After nosewheel touchdown, with the aircraft now rolling along the runway, it experienced an increasing vibration which was felt both by the pilots and by the flight attendants and passengers. The captain took control of the aircraft when he felt the vibrations. The lift dumpers opened in the normal manner. The aircraft was zigzagging, with a tendency to pull to the right, and the pilot used the brakes and applied idling reverse to attempt to brake and keep the aircraft on the runway; he increased the power asymmetrically, up to 1.18 EPR in the left-hand engine and 1.13 in the right. The right-hand main landing gear then collapsed. The two wheels, together with the sliding member of the right-hand main landing gear leg, sheared off and struck the fairing of the right-hand engine. The wing leant on that side, scraping the right-hand flap, wingtip and aileron along the ground. In the wing, the MLG Bracket split. However, the fuel tanks remained intact. The aircraft finally came to a halt at the side of the runway, outside the tarmac paved surface, in a grassy area 1,730 m from the threshold of the runway and 100 m to the right of the centre line or axis of the runway. The right wing was resting on the ground. There was no fire. Causes: 1. The aircraft suffered the accident due to the breakage of the main fitting of the RH-MLG. 2. The cause of the failure of the main fitting was exposure to mechanical loads, greater than the loads this part was designed for. The loads were due to a shimmy vibration 3. The vibration became unstable due to a functional failure of the shimmy damper. 4. The low damping capacity of the shimmy damper was due to the excessive thickness of a pack of shims between the housing and the cap which form the sleeve of the damper. 5. The defective assembly was not detected in a mandatory flow test on a hydraulic bench. 6. The existence of the shims themselves could have contributed to the assembly defect; it has been demonstrated that they have no functionality in the design. 7. Another possible contribution to the deficient operation of the damper could have been low tightening torque in the bolts connecting the two sleeves forming the damper housing. 8. A possible contributing factor to the shimmy phenomenon initiation could have been the gentle touchdown made by the aircraft combined with a high coefficient of friction corresponding to the dry runway.