On July 20, 1998, at 1900 central daylight time, a Piper PA-23 twin engine airplane, N3154P, was substantially damaged during an undershoot following a loss of power from the right engine near Bridgeport, Texas. The three occupants, a flight instructor, a private pilot receiving instruction, and one passenger, were not injured. The airplane was registered to and operated by Bridgeport Flight Training Center. Visual meteorological conditions prevailed, and a flight plan was not filed for the Title 14 CFR Part 91 local instructional flight that originated about 5 minutes before the accident. With the pilot receiving instruction manipulating the controls, the airplane took off from runway 17 at the Bridgeport Municipal Airport and climbed to a traffic pattern altitude of 1,500 feet msl. While on downwind, the pilot confirmed that "everything was alright" and lowered the landing gear. Abeam the numbers to runway 17, the pilot extended half flaps and then began to turn left base. At this point, the airplane "suddenly" began loosing altitude. The instructor took control of the airplane and applied full throttle to maintain altitude. Upon turning final, the instructor observed that the airplane would not maintain altitude and that the right engine had failed. He feathered the right engine and retracted the flaps. The airplane impacted trees short of runway 17. Examination of the wreckage by the FAA inspector revealed that both wings were separated outboard of the engines, and the inboard portions of the wings were pulled away from the fuselage. The left engine separated from the wing, and the left propeller separated from the engine crankshaft. Fuel was observed in the fuel tanks. According to the operator, as of January 8, 1999, the right engine had not been disassembled or test run. In the section of the Pilot/Operator Aircraft Accident Report (NTSB Form 6120.1/2) entitled "Recommendation (How Could This Accident Have Been Prevented)," the operator wrote the following: Power was reduced for landing and apparent R/H engine loss was not recognized until unusual descent rate noticed. Possibly - monitoring engine gauges would have alerted them to a problem at an earlier stage.