Accident Details
Probable Cause and Findings
The pilot’s improper banner pickup approach, which resulted in the banner tow hook contacting the airplane, and the subsequent overload separation of the right rudder control horn, resulting in a loss of airplane control.
Aircraft Information
Registered Owner (Current)
Analysis
On May 25, 2023, about 1209 eastern daylight time, a Piper PA-25-260, N454AB, operated by Aerial Banners North Inc. was substantially damaged when it was involved in an accident near Hollywood, Florida. The commercial pilot was seriously injured. The airplane was operated as a Title 14 Code of Federal Regulations Part 91 banner tow flight.
According to the operator’s director of ground operations, who supervised the flight from the banner pickup point and communicated with the operator’s airplanes with a handheld radio, the pilot took off from runway 10L at North Perry Airport (HWO), Hollywood, Florida around 1155 for a scheduled 1200 banner tow flight.
The pilot completed his first pass in the banner box for hook deployment and verification of that hook. He then missed the banner during his first three pickup attempts. All seemed normal (airspeed and altitude) to the director, but he noticed that the pilot was making his crosswind turns earlier than where he was supposed to (over taxiway Delta).
There was then a change of the banner the pilot would tow, and the director relayed that information to the pilot via radio. The director told the pilot to go around and fly his approach on a different lane than the previous one that was assigned to him. All seemed “fine and normal,” regarding the airplane’s airspeed and altitude.
The pilot approached the banner but missed the pickup again. During this attempt, the banner tow hook struck the right cone of the lane during the climb out. The director was a couple of ft from his lane, where he usually stood, and advised the pilot that he missed on his attempt to pick up the banner. The director then took a couple of steps towards the pick lane to fix the cone, and as he turned to watch the airplane, he observed that it was at an approximate altitude of 100 ft above ground level in a “spin” before it then impacted the ground.
According to another witness, who was also a pilot with banner towing experience, she observed the accident from about 1,000 ft away while riding as a passenger in a car. As she approached the banner tow pickup area, she saw the airplane approaching. She also saw that the hook was out behind the airplane for a normal looking pickup heading east. The hook did not catch the rope attached to the banner. The airplane pitched up and climbed to about 50 ft in altitude and then started a steep banked climbing turn to the left. The airplane then rolled wings level onto a westerly course at an altitude of about 200 ft above the ground. Moments later, the left wing suddenly dropped and the airplane pitched nose down, rotating to the left in a spin. It rotated 180 degrees and then impacted the ground, almost horizontal in about a 20-degree nose down attitude.
The pilot was interviewed about 4 months after the accident. In the interview he described a series of events similar to those described by the director of ground operations. After the initial missed pickup attempts and change to a different banner lane, the pilot made another approach, but missed the banner pickup. During the climb he applied a usual amount of rudder but then felt it “snap” or “let go,” after which the airplane immediately turned left. He further described that he had little time to react and could only push the airplane’s nose downward. He recalled making a statement over the radio about the airplane’s rudder but could not recall the details.
Review of preliminary Automatic Dependent Surveillance Beacon data provided by the Federal Aviation Administration (FAA) depicted a series of events consistent with the witness descriptions. It also depicted that the pilot had flown a progressively tighter pattern on each attempted banner pickup prior to the accident.
The pilot held a commercial pilot certificate with ratings for airplane single-engine land, multiengine land, and instrument airplane. His most recent FAA first-class medical certificate was issued on January 10, 2023. On that date, the pilot reported that he had accrued approximately 821 total hours of flight experience. The operator’s records indicated that the pilot was hired on November 14, 2022. He had completed 60 hours of classroom, ground, and flight training, along with several written and practical examinations. The records also indicated that he had been flying on a regular basis with the company since being hired.
According to FAA airworthiness and maintenance records, the airplane was manufactured in 1974 and was powered by a Lycoming O-540-G1A5, 260-horsepower engine. The airplane’s most recent annual inspection was completed on March 6, 2023, at 9,573.1 total hours of operation.
Postaccident examination of the cockpit indicated that both magneto switches were in the “ON” position, the electric fuel boost pump was “ON,” the throttle was full forward, the mixture was full rich, and the carburetor heat control knob was in the “OFF” position. The vertical speed indicator needle indicated a descent of about 300 ft per minute. The tow hook, which was constructed of ¼-inch-thick steel rebar bent into three roughly “J” shapes that were bound together, was also recovered and examined. One of the three prongs was bent from its normal shape outward and was nearly straight (figure 1).
Figure 1. Tow hook with one of its three prongs (top) bent nearly straight
Examination of the propeller revealed that it had separated from the engine. The fracture surface signatures were consistent with torsional overload. One blade of the two blade propeller displayed S-bending, leading-edge gouging, and chordwise scratching. The other blade displayed curling on the outer portion and about 4-inches of the tip had separated from the rest of the blade.
Examination of the engine revealed that there was oil in the rocker boxes, galleries, and sump. Continuity of the powertrain was established from the propeller flange to the rear gears. Both magnetos were functional when tested. Thumb compression and suction was attained on all cylinders. The carburetor jets were clear, the carburetor floats were devoid of fuel, and the float valve assembly was functional. The carburetor bowl contained trace amounts of fuel. Testing of the fuel in the carburetor bowl revealed no evidence of water or other contamination. Examination of the fuel strainer revealed that it was full of fuel, had some debris (in the bottom of the strainer bowl outside the screen), and when the fuel was tested, it also revealed no evidence of water contamination. The screens for the electric boost pump, fuel strainer, and carburetor were clean.
Examination of the pitch control system revealed that continuity could be established from the control stick in the cockpit, through the breaks in the system to the elevator. Examination of the roll control system also revealed that continuity could be established from the control stick in the cockpit, through the breaks in the system to the ailerons. Examination of the yaw control system revealed that continuity could be established from the rudder pedals in the cockpit, to the left rudder control horn, but could not be established to the right rudder control horn. Examination of the right rudder control horn revealed that it was fractured, with the face of the fracture displaying corrosion.
The rudder arm assembly, with two extensions (installed as part of a supplemental type certificate modification) were forwarded to the NTSB Materials Laboratory for detailed examination. The right side of the rudder horn had fractured approximately 1.25 inches from the vertical arm assembly body. The cast link had also fractured in two locations. All fracture surfaces exhibited enough surface oxidation (rust) to conceal the underlying metal surfaces, including the fracture surfaces. Much of the yellow paint in these areas had also begun to spall. The outboard rudder horn fragment was sectioned, and to examine the underlying fracture features, the sectioned horn piece was cleaned. The fracture surfaces of the outer horn fragment exhibited a rough, tortuous surface texture with a dull, fibrous luster. The smaller filament exhibited thinning and necking near the fracture surface. In addition, there was upward deformation and smearing of the necked area near the hole bore to the right. Examination of this surface using a field emission scanning electron microscope (SEM) revealed dimpled rupture. These features were consistent with overstress fracture of ductile material in tension and localized bending.
The larger filament on the right side of the rudder horn fragment demonstrated a fracture surface at a 45° slant angle relative to the part. There was a localized downward deformation of the highly textured grain flow near the bole bore. Examination of this part of the surface using an SEM revealed localized flow deformation. The fracture surface exhibited dimpled rupture, consistent with overstress fracture. Examination of the rightmost (L-shaped) side of the horn fracture showed localized necking near the fracture surface. These features were consistent with tensile and bending overstress fracture of the horn. The rudder horn was examined using energy-dispersive x-ray spectroscopy (EDS) and x-ray fluorescence. The composition was consistent with carbon steel and the hardness of the material was unremarkable.
The fractured link casting adjacent to the rudder arm assembly body was also examined. The broken surfaces were sectioned from the rudder arm assembly and cleaned to remove much of the post-fracture surface contamination on the component. Much of the fracture surfaces were flat and 90° in orientation relative to the fractured arms, though parallel to the rudder arm cylinder. The fracture surfaces exhibited a dull luster and rough texture, and detailed examination confirmed that the fracture features were consistent with overstress fracture. The composition was examined using EDS and was also unremarkable.
The operator reported that their training manual descri...
Data Source
Data provided by the National Transportation Safety Board (NTSB). For more information on this event, visit the NTSB Records Search website. NTSB# ERA23LA245