WPR11FA316

The pilot's decision to take off toward a nearby low cloud layer and the subsequent turn, stall, and spin during the pilot’s attempt to avoid the cloud layer. Contributing to the accident was the pilot's failure to avoid the stall. His ability to avoid the stall was hindered by an inaudible stall warning system of questionable accuracy.

The pilot departed from his home airport situated about 3 miles east-northeast of the ocean with a low-lying stratus cloud layer.  The takeoff was conducted while it was still daylight. Eyewitness and photographic evidence indicated that the stratus layer was nearby, to the southwest, south and southeast of the airport at the time of the takeoff.  The airport was non-towered, and was equipped with two similar-length runways, designated as 2/20 and 8/26. Airplane performance, and terrain and obstacle clearance considerations did not preclude a takeoff from any of the four possible runway options. However, the takeoff was conducted from runway 20, directly towards the cloud layer. Eyewitnesses and recovered GPS data indicated that the airplane began a sharp left turn prior to reaching the end of the runway, at an altitude of about 400 feet above ground level (agl). That turn was consistent with an effort to avoid the cloud layer, but contrary to published airport noise abatement guidance that prohibited departure turns prior to the airport boundary, or at altitudes below 900 feet agl. The airplane did not enter the cloud, but during the turn, the airplane stalled, entered a spin, and descended rapidly to the ground. The airplane struck a parking lot and building less than 700 feet from the departure runway. Post-accident examination of the airplane and engine did not reveal any anomalies or failures that would have precluded normal operation. At least two headsets, one of which was a noise cancelling unit, were located in the wreckage. According to the airplane co-owner, the vane-activated, electrically-powered stall warning horn was inaudible to a pilot wearing a headset, and the owners' attempts to rectify that situation were unsuccessful. Post-accident testing of the vane switch and warning horn indicated that they were functional, but the horn volume was not measured or compared to any known standard. During airplane manufacture, the final position of the stall warning vane and switch assembly on the wing is determined during the production flight test of each individual airplane, in order to ensure system activation at the proper angle of attack. No records of the as-delivered vane position were available, and the as-delivered position of the vane could not be discerned by examination of the wreckage. Examination of the vane assembly revealed that it had been modified, and was not installed in accordance with the manufacturer's design drawings. In addition, no information regarding the accuracy of the modified stall warning system was located. The investigation was unable to determine whether the system would have provided sufficient, or even any, notification of a stall, presuming the horn was audible to the pilot, which in this case it was not. Despite three other runway alternatives, the pilot knowingly and intentionally decided to depart from the runway most closely aligned towards the stratus layer, with the apparent plan to turn to avoid it once airborne. While his runway choice may have been influenced by habit pattern, existing traffic, or a previous taxi event at that airport, the investigation was unable to determine why the pilot chose that runway, instead of using any of the other three alternatives which would have taken him away from the cloud layer. He then inadvertently stalled and spun the airplane during the avoidance turn, at an altitude which did not allow recovery.

WVI was a non-towered airport situated about 2 miles northwest of Watsonville, and about 3 miles east-northeast of the Pacific Ocean. Airport elevation was 163 feet above mean sea level (msl). WVI was equipped with two paved runways, designated as 2/20 and 8/26. Runway 2/20 measured 4,501 feet by 150 feet, while runway 8/26 measured 3,999 feet by 100 feet. The full length of each could be used for takeoff. Runway 2/20 was oriented approximately perpendicular to the local shoreline, and therefore aligned approximately towards the source of the stratus layer. All runways were designated as left traffic, and runway 20 was designated as the "preferred calm wind runway" in the noise abatement guidance published by the airport. The guidance also stated "no turns before crossing the freeway" [California highway 1] for departures from runway 20. Highway 1 was located about 1/4 mile beyond (west-southwest) the departure end of runway 20; the accident site was approximately abeam the departure end of runway 20. The guidance prohibited departure turns below 900 feet above ground level (agl), and advised pilots that "Safety always supersedes noise abatement procedures." According to one witness, a Piper Archer had departed runway 20 just prior to the accident airplane, but the investigation was unable to determine whether any other aircraft departed or arrived in the period surrounding the accident time. According to the airport manager, some pilots avoid using runway 8 due to the deteriorated condition of the taxiway normally used to access it. The investigation was unable to determine whether the previous landing gear door damage incurred by the accident pilot during taxi occurred on this taxiway, or elsewhere on WVI. In July 2013, the manager reported that an approximate 8-year effort to obtain required approvals to repair the taxiway had recently been successful, and that the repair project was moving ahead. That recent approval was not related to, or influenced by, the accident. The airport manager noted that due to an attempt by the City of Watsonville to close runway 8/26, the pilots increased its utilization, particularly when the fog/stratus layer was approaching the airport. He also noted that "recently" (prior to the accident) although runway 8/26 is designated as left traffic, the "fog was forcing a right pattern." An autopsy was performed on the pilot by the Santa Cruz County (California) Sheriff-Coroner Office. The autopsy determined that the cause of death was "injuries sustained in the plane crash." Forensic toxicology was performed on specimens from the pilot by NMS Labs of Willow Grove, Pennsylvania, for the Santa Cruz Coroner. The toxicology report stated that the examination "did not reveal any positive findings of toxicological significance." The FAA Civil Aeromedical Institute conducted forensic toxicology examinations on specimens from the pilot, and reported that no carbon monoxide, cyanide, ethanol, or any screened drugs were detected. Airplane Performance Information The accident flight calculated weight (2,650 lbs) and moment (128,709 in-lbs) values were evaluated against the airplane weight and balance envelope; the results indicated that the airplane was loaded within the allowable weight and balance limits. The Garmin GPSMap 396 that was recovered from the wreckage contained data from the accident flight. The data consisted of 11 usable points for the accident flight; parameters included time, latitude, longitude, and GPS altitude. The recovered data was used by NTSB engineering personnel to estimate/calculate some basic airplane performance parameters. Those results were then compared to the airplane manufacturer's Owner's Manual (OM) performance data. The data indicated that the airplane took the runway at 1927:02, started its takeoff roll about 7 seconds later, and became airborne about 1927:30. Between 1927:44 and 1927:56, the ground track began deviating to the southeast (airplane left). At 1927:58, the airplane was abeam the departure end of runway 20, offset about 300 feet southeast of the centerline, at an altitude of about 430 feet agl. The maximum altitude of about 440 feet agl was recorded at 1928:01, and the next GPS data point was the last recorded point. That point was recorded at 1928:05, and was located approximately coincident with the impact location. The airplane achieved a steady-state climb rate of about 800 feet per minute (fpm) from about 1927:30 to 1927:50, and then over the next 8 seconds, the rate began to decay slowly. The airplane then began a very rapid descent. Calculated airspeed values that utilized a constant ASOS-based wind speed and direction were initially about 73 mph and then decreased approximately linearly to about 55 mph. The OM specified a flaps-retracted, power-off, wings-level stall speed of 68 mph, which increases to 80 mph in a 40 degree bank. The OM did not contain any performance data regarding flight with the landing gear doors removed. Since the (input) GPS data was at such a low sample rate, the (output) calculated performance values are necessarily coarse, and should be considered approximations. Furthermore, the calculation technique for the pitch and roll angles assumes non-stalled, coordinated flight, which may not be representative of the entire accident flight. In particular, the calculation technique is not able to capture a spin, such as that reported by some witnesses to the accident. Given these and other limitations, the observed takeoff and climb performance was not significantly different from the predicted performance. Stall Warning System The airspeed indicator (ASI) is the primary device for stall avoidance in the accident airplane, while the stall warning system provides a secondary defense mechanism. The speed arc markings on the ASI were determined to be congruent with the FAA-approved and required values. During post accident functional testing of the ASI, an internal gasket was determined to be leaking. The testing revealed that ASI indications for input test speeds below 100 mph were initially no more than 2 mph high, but that a gasket leak could result in a subsequently lower indication. It could not be determined how much lower that indication would be, and it could not be determined whether the leaking gasket was a result of the accident and post impact fire. The airplane was equipped with a stall warning system, which consisted of a wing-mounted sensor vane and switch assembly wired to an audible electric horn installed in the cockpit. The M20 series aircraft were manufactured under the CAR 3 regulations, and the section specific to the stall warning system was 3.721. The system is also compliant with FAR 23.1431, paragraph 6. According to the airplane manufacturer's OM, "A stall warning horn, …triggered by a sensing vane on the left wing leading edge, will sound when airspeed drops to near stall speed. The sound becomes steady as the aircraft approaches a complete stall." At airplane manufacture, the stall warning horn in the M20F was a Mallory Sonalert Products SC628R. The volume output of the horn is dependent on the current supplied. At 12 volts, the horn volume would be approximately 82 db at a distance of 2 feet. Any degradation in the grounding of the horn wiring circuit would decrease the effective voltage, and thereby decrease the volume of the horn. According to the maintenance records, a new stall warning horn switch was installed in December 2003, when the airplane had a total time in service (TT) of about 3,442 hours. A June 2007 (TT about 3,818 hours) maintenance records entry catalogued an effort to "troubleshoot and repair [an] inoperative" stall warning horn. A February 2008 entry (TT about 3,878 hours) stated that a technician had "replaced [a] broken wire to [the] stall warning" system. Post-accident interviews with the airplane co-owner revealed that the co-owner had never determined