On August 5, 1993, at 0815 central daylight time (CDT), a Taylorcraft L2B, N3113S, registered to Russell R. Johnson of Duluth, Minnesota, and piloted by a private pilot, was substantially damaged when it collided with trees and swampy terrain during a forced landing. Visual meteorological conditions prevailed at the time of the accident. The personal 14 CFR Part 91 flight was not operating under a flight plan. The pilot reported minor injuries. The flight originated from a restricted landing area near Friedenberg, Minnesota, at 0810 CDT. The pilot stated he performed a run-up and began the takeoff shortly thereafter. Once the airplane was approximately five feet above the runway, in level flight, the engine stopped running for a short time according to the pilot. The engine began to run at what the pilot referred to as takeoff power while flying down the runway. He said the burst of power was enough to get the airplane over the trees. As the airplane was passing over the trees, the pilot said the engine stopped running a second time. He said he decided to land in the swamp because there were trees all around him. During the descent into the swamp, the pilot stated the airplane's right wing tip struck a tree causing the airplane to nose over into the swamp. The pilot reported the right wing spar was broken at the fuselage attach fitting, the right rear wing strut was bent, right wing tip bow was bent, and the right landing gear had crushed. The pilot's written statement included with NTSB Form 6120.1/2 revealed the airplane's engine had "...stalled..." after approximately 10 to 15 seconds of flight following takeoff. The pilot stated the engine had "...another burst of full power near the end of runway @ 30' above trees at this point." After the engine power loss, the pilot stated he shut the magnetos off and planned for a forced landing in the swamp. During an interview with the pilot it was revealed he had performed maintenance on the airplane the day before the accident. He stated he had installed insulation under the fuselage header fuel tank. During this activity, he stated his shoulder brushed against the header tank fuel valve. He said he did not believe it was enough to cause any problems since the valve's cockpit control had a "...little play in it."