On April 25, 2002, about 1745 eastern daylight time, a Cessna 421C, N237SM, operated and registered to MBA Graphics of Tampa, struck the ground after landing with the left main gear retracted at Vandenberg Airport, Tampa, Florida. Visual meteorological conditions prevailed at the time. No flight plan was filed for the 14 CFR Part 91 personal flight. The airplane was substantially damaged. The commercial-rated pilot reported no injuries. The flight had departed from the same airport about 1445, en route to Plant City, Florida. According to the pilot's statement, after a landing and upon touchdown at Plant City, he said he "…observed an abnormal touchdown...and knew there was a mechanical problem." After noticing the mechanical problem, he aborted the landing, cycled the gear twice, and noted that the gear unsafe light remained illuminated. He then proceeded to his departure airport, for further assessment of the landing gear situation by the tower personnel. The tower personal confirmed there was damage to the left main gear, he then proceeded to burn off about 1 1/2 hours of fuel before attempting a landing. The pilot landed on the grass next to the paved runway, touching down on the nose and right main landing gears. The airplane settled on the left wing, turned left onto the paved runway and came to rest 180 degrees opposite the direction of touchdown. Examination of the landing gear after the accident revealed that the upper barrel (p/n 5141002-11) fractured during the landing attempt at Plant City. The fractured upper barrel assembly was examined at Cessna's Material and Process Engineering's Laboratory. The purpose of the examination was to determine the cause of the main landing gear (MLG) fracture. Based on the observations made during the examination of the fractured MLG barrel, and according to Cessna's factual report, the following conclusions were reached: 1.) The upper barrel broke at a location coinciding with the position of the collar, "approximately 3-9/16 inches below the trunnion." 2.) "Pre-existing cracking was present on the trunnion barrel at two separate locations approximately 180 degrees apart," along the outer diameter of the barrel. "Fatigue cracks had propagated through approximately 75 to 80 percent of the wall thickness of the barrel." 3.) The crack origins coincided with "...a transition in barrel wall thickness. Machining marks were present on the radius of the transition in wall thickness." 4.) The dimensions of the barrel "were according" to the engineering drawing with the exception of the location of "...the brazed-on collar, where the outer barrel diameter was smaller than specified." Sectioning of the barrel through the fracture surface showed the presence of a "groove machined" into the barrel immediately above the collar. 5.) Smaller than specified barrel diameter and the presence of machining marks on the machined radius "…could have contributed to the cause of cracking." The surface roughness at the radius was 140-160 µ-in (micro-inches), which "exceeded" the engineering drawing requirement of 125 µ-in. 6.) The barrel chemical composition and hardness "were according" to the engineering drawing. The thickness of the braze metal between the barrel and the collar "met the requirements" for a Class I or Class II copper braze joint per CSMP009 as well as the requirements of CES2044 which is listed on the engineering drawing. A review of the Cessna report by the NTSB's Materials Laboratory revealed that the Cessna factual report contained an adequate factual description of the fracture, dimensional measurements, and other features associated with the fracture of the upper barrel.