CEN11IA379

The failure of the right main landing gear (MLG) to extend normally, which resulted from the combination of a slightly worn uplock pin, the operation of the hydraulic pumps below their specified operating pressures, and contamination within the landing gear hydraulic system. Also causal to the accident was the failure of the right MLG to extend manually using the alternate gear selector handle due to the improper use of lubricant within the alternate release handle assembly, which prevented the handle from remaining in its fully extended position.

During final approach, the transport-category airplane's right main landing gear (MLG) did not extend when the landing gear selector was placed in the "down" position. Shortly after, the flight crew noted a "gear disagree" warning message displayed on the engine indication and crew alerting system showing that the nose gear and the left MLG were down and locked but that the right MLG was in transit. The flight crew followed the quick reference handbook directions to troubleshoot the landing gear issue without success. The flight crewmembers then tried to manually extend the landing gear by pulling the landing gear alternate release handle multiple times; however, the right MLG failed to extend, and they subsequently landed the airplane with the right MLG retracted. During postincident activities, the airplane was lifted and an examination revealed that the right MLG remained in its full-up position within the wheel well. The airplane's right MLG uplock pin exhibited signs of slight wear, consistent with in-service usage, and no abnormalities were found with the right MLG uplock assembly. Normal wear patterns were observed on the uplock latch with no discernable depth at the pin contact locations. The uplock pin's slight wear indicates that it might not have always rotated freely when it was in contact with the upper and lower surfaces of the uplock assembly latch as it should have per the system's design. Further, the worn pin could have increased the friction forces between the pin and the latch. Functional testing of the incident airplane's 3A and 3B hydraulic system pumps showed that their output pressures were lower than specified. Analysis of hydraulic fluid samples taken from the No. 3 hydraulic system, including the right sidestay actuator and selected landing gear components, revealed that some of the fluid samples contained particles and fibers that exceeded the in-service size limits specified in the airplane's maintenance manual, and these contaminants likely restricted the hydraulic flow within the hydraulic assembly. It is unlikely that any of these factors (the slightly worn uplock pin, the operation of the hydraulic pumps below their specified operating pressures, or the contamination within the landing gear hydraulic system) would have individually prevented the right MLG from extending; however, it is likely that the combination of these factors prevented the right MLG from extending. In addition, the overall landing gear system design, including modifications made in accordance with an airworthiness directive, did not preclude the consequence that the intermittent combination of these factors prevented the right MLG from extending normally. Postincident examination of the cockpit revealed that the landing gear alternate release handle remained extended about 7 inches. Pulling up on the handle resulted in it moving about 3 additional inches to its fully extended position. The additional handle displacement resulted in the right MLG extending out of its wheel well. When the handle was released from its fully extended position, the handle automatically began to slowly retract, which caused the right MLG to stop extending. The handle had to be pulled and manually held in its fully extended position for the right MLG to extend to its down-and-locked position. During postincident functional testing of the landing gear alternate release handle assembly, the handle's locking mechanism failed to maintain the handle in its fully extended position when a specified retract load was applied to the assembly. Disassembly of the handle assembly did not reveal any mechanical discrepancies. However, the assembly's inner housing and outer slider were found coated with an oily material consistent with lubricant; the origin of the oily material could not be determined. The alternate release handle assembly design specifications do not call for the application of lubricant on the inner housing or the outer slider; therefore, it is likely that the lubricant prevented the handle from remaining in its fully extended position.

The airplane was equipped with a L-3 Communications model FA2100-1020 cockpit voice recorder (CVR) with serial number 000228060. This model is a solid-state CVR that records 2 hours of digital cockpit audio. Specifically, it contains a two-channel recording of the last two hours of operation and separately contains a four-channel recording of the last 30 minutes of operation. The two-hour portion of the recording is comprised of one channel of audio information from the cockpit area microphone (CAM) and one channel that combines three audio sources: the captain's audio panel information, the first officer's audio panel information, and the observer's audio panel information. The 30-minute portion of the recording contains four channels of audio data; one channel for each flight crew and one channel for the CAM audio information. The CVR was received at the recorder laboratory where it did not exhibit any heat or structural damage. The audio information was extracted from the recorder normally, without difficulty. Timing of the summary was established by correlating CVR events to common events on the flight data recorder (FDR). The airplane was equipped with a L-3 Communications/Fairchild model FA2100 FDR with serial number 000174026, which was designed to meet the crash-survivability requirements of TSO-C124a. This model records airplane flight information in a digital format using solid-state flash memory as the recording medium. The FA2100 can record a minimum of 25 hours of flight data. It is configured to record 128 12-bit words of digital information every second. Each grouping of 128 words (each second) is called a subframe. Each subframe has a unique 12-bit synchronization (sync) word identifying it as subframe 1, 2, 3, or 4. The sync word is the first word in each subframe. The data stream is "in sync" when successive sync words appear at proper 128-word intervals. Each data parameter has a specifically assigned word number within the subframe. The FDR was received at the recorder laboratory in good condition and its data was extracted normally from the recorder. The FDR recording contained approximately 119.4 hours of data. Timing of the FDR data is measured in subframe reference number (SRN), where each SRN equals one elapsed second. The incident flight was the last flight of the recording and its duration was approximately 1 hour and 42 minutes. The FDR incident data was converted from SRN to the incident local time. MKE was a field elevation of 723 feet and was five runways. Runway 13/31 was concrete-surfaced, 5,868 feet long and 150 feet wide. Runway 7R/25L was asphalt-surfaced, 8,012 feet long and 150 feet wide. Runway 7L/25R was asphalt and concrete surfaced, 4,800 feet long and 100 feet wide. Runway 1R/19L was concrete-surfaced, 4,183 feet long and 150 feet wide. Runway 1L/19R was asphalt and concrete-surfaced, 9,690 feet long and 200 feet wide. A National Transportation Safety Board (NTSB) vehicle recorder specialist chaired a CVR group and produced a CVR factual report. The CVR group reviewed recorded cockpit communications starting from the airplane's time while parked on the ramp at CVG. The report summarized communications during the flight's initial approach descent at MKE and continued through the final landing. The flight crew statements are consistent with the CVR factual report's findings. The report findings confirm the flight crew used QRH references during their six landing gear extensions attempts. A NTSB vehicle recorder specialist downloaded, decoded, and produced a factual report to include graphic plots in reference to data from the FDR. The FDR incident data, in part, indicated that about 20:52, while descending through a pressure altitude of approximately 2,745 feet, the left MLG data transitioned from "Not Down and Locked" to "Down and Locked." The NLG data transitioned from "Not Down and Locked" to "Down and Locked" while the airplane's data indicated it was approximately at a pressure altitude of 2,618 feet. The landing gear disagree warning transitioned from "Not Active" to "Active" while the airplane was approximately at a pressure altitude of 2,217 feet. About 20:54, the right MLG transitioned from "Not Down and Locked" to "Down and Locked" and the airplane's pressure altitude increased to approximately 2,869 feet. A second later, the landing gear disagree warning transitioned back to "Not Active." About 20:54, the right MLG transitioned back to "Not Down and Locked." By 20:55, both the left MLG and NLG transitioned back to "Not Down and Locked." Over approximately the next 10 minutes until 21:04, the NLG data transitioned six times from "Not Down and Locked" to "Down and Locked" and it remained at "Down and Locked" until touchdown. From 21:00:36 to 21:01:06, the right MLG transitioned to "Down and Locked" and back to "Not Down and Locked" while the airplane was at a pressure altitude about 4,150 feet. About 43 seconds later, the No. 3 hydraulic pressure decreased from approximately 2,720 pounds per square inch (psi) to 14 psi. The airplane remained at about 4,150 ft. While the airplane was at a pressure altitude 4,150 feet and over the next, approximately, 7.5 minutes until about 21:13, the No. 3 hydraulic pressure increased to about 2,720 psi, decreased to about 10 psi, increased to about 2,700 psi, decreased to about 14 psi and then increased to about 2,680 psi where it remained until touchdown. About six and one half minutes later about 21:19, the left MLG transitioned from "Not Down and Locked" to "Down and Locked" and it remained at "Down and Locked" until touchdown. The airplane's pressure altitude had decreased to about 900 feet at that time. About 13 minutes and 14 seconds later at 21:33:04, the left MLG weight on wheels data transitioned from "Air" to "Ground" and one second later at 21:33:05 CDT, the NLG weight on wheels data transitioned from "Air" to "Ground." The FDR recorder specialist's factual report is appended to the docket material associated with this case. The pressure and return hydraulic filters were examined with computed tomography scans and digital radiography. A NTSB aerospace engineer produced a computed tomography specialist's factual report based on the radiographic examination. The examination showed that there were three particles found in the pressure filter and one particle found in the return filter. There were several cracks noted in the epoxy material in the end caps of the return filter, and there were some high-density areas noted in both filters within the filter material itself. The specialist's report is appended to the docket material associated with this case. Eight hydraulic system filters were shipped to the NTSB material laboratory along with hydraulic fluid that was collected with the filters for examination. A NTSB chemist examined the filters and fluid and produced a materials laboratory factual report. The report indicated that all of the filters were comprised of an inner metallic perforated tube with two additional types of filtration media laid over the top: an outer layer of stainless steel wire screen; and a filter consisting of several layers of woven fiber mesh located between the tube and the steel mesh. A measurement of the openings in the outer wire screen found the openings to be 166 by 198 micrometers (µm). The inner fiber mesh consisted of irregularly shaped and sized openings with an average opening size between 25 µm to 50 µm. The filter mesh layers for all of the filters were examined under a 5X to 50X stereo zoom-microscope. The examination of the filters revealed no significant particulates within the mesh. Each filter was rinsed with acetone to remove any material trapped within the filter material. There was no evidence of metallic particles present in the filtrate rinse. The filtrate from both the fluid and the filters samples was further examined and it exhibited spectra was consistent with characteris