Investigation of weight-on-wheel switch failure in F-16 aircraft

Transcription

1 Engineering Failure Analysis 12 (2005) Investigation of weight-on-wheel switch failure in F-16 aircraft Cher Ming Tan a, *, Kok Meng Ang a, Liang Huat Tay b, Tan Yeow Meng c a Nanyang Technological University, School of Electrical and Electronic Engineering, Nanyang Avenue, Singapore b Republic of Singapore Airforce, Tengah Airbase, Singapore , Singapore c Singapore Institute of Manufacturing Technology, Joining Technology Group, 71 Nanyang Drive, Singapore , Singapore Received 13 September 2004; accepted 14 September 2004 Available online 6 November 2004 Abstract A F-16 weight-on-wheel (WOW) switch failure investigation was conducted. It was found that the root causes were broken wires that lead to the failure of the switching mechanism of the two micro-switches in the WOW switch. Detailed investigation showed that the defective WOW switch does not have a stainless steel casing for the wires, and the routing of the harness of the WOW switch was too loose, causing frequent and severe flexing. This rendered the wires close to the bottom of the switch housing to rub against it, resulting in a cut in the wire. Furthermore, scanning electron microscope revealed that the cross-section of the cut-opened wire has a smooth flat surface, probably indicating that the copper wire has failed by fatigue. For the other failure, a bending mark was observed in the wire. Scanning electron microscope showed that the insulation had an obvious twisted crack, showing that the wire has been subjected to torsion. The cause of the torsion was found to be due to the improper combing of the wires before an outer rubber tubing was encased on it as evidence from a separate experiment. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Electronic-device failures; Aircraft failures; Switch failures; Fatigue; Copper-conductors 1. Introduction Failures in aircraft can be classified as mechanical and electrical failures. For electrical failures, it is common that the failed electrical items are identified and replaced, with no further analysis being performed on the root cause of the failed items. However, such a practice could lead to the recurrence of the failures that not only present a potential hazard to the personnel in the aircraft, but also increase the maintenance cost of the aircraft. * Corresponding author. Tel.: ; fax: address: ecmtan@ntu.edu.sg (C.M. Tan) /$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi: /j.engfailanal

2 C.M. Tan et al. / Engineering Failure Analysis 12 (2005) On the other hand, one can view the electrical failures as an opportunity for improvement in both safety and maintenance practice. With this paradigm shift, root causes investigation of the failed items becomes necessary. The outcome of the root causes analysis will provide the needed improvement on the maintenance of the aircraft, thus eliminating or minimizing the recurrence of the same failure. This will improve the reliability of the aircraft and reduce the maintenance cost. Among the various electrical failures found in aircraft, it is reported that interconnections, instruments and electromechanical devices accounted for a total of 72% of the electrically related aircraft incidents; and further breakdown shows that wires accounted for 18%, switches/relays accounted for 28% of the total failure accidents [1,2]. Therefore, in this work, we choose an electromechanical switch/relay failure for detail investigation as this failure involves all the three failure modes mentioned above. The switch chosen in this work is the weight-on-wheel (WOW) switch. It was found defective on aircraft number 885 during pre-flight checks on September Background information 2.1. Purpose of the WOW switch The WOW switch shown in Fig. 1 is an electro-mechanical switch with a 4 pole double throw (4PDT) circuit. The transfer function of the WOW switch is achieved in two parts. The first part is done mechanically by the actuation of the pin plunger. The second part is operated electrically by means of the 4PDT circuit; which is actually the four sets of micro-switches found within the stainless steel housing of the WOW switch. In the case of the F-16 aircraft, the WOW switch is engaged by supplying the switch with 28 VDC. This in turn activates a number of relays that control several flight control systems. Examples of the flight control systems are: landing gear wheels and brakes, electrical power system AC power generation, flight environment systems, engine power control, nose wheel steering system, etc Mechanism of the switch Radiographic analysis was conducted on a good WOW switch to understand its mechanism. Fig. 2 shows the X-ray image obtained. The pin plunger of the WOW switch is held in position by a one-piece stainless steel housing and a spring. When the pin plunger is not actuated, the four sets of micro-switches Fig. 1. Defective WOW switch.

3 510 C.M. Tan et al. / Engineering Failure Analysis 12 (2005) are contacting the normally closed (NC) contacts. When the pin plunger is actuated, it pushes against the spring mechanism. Subsequently, the movement of the spring mechanism causes the metal plate to move down, causing the mini-plunger to be depressed. This in turn causes the four sets of micro-switches to switch to the normally open (NO) contacts within 80 ms of the first circuit transfer. 3. Procedure of root cause investigation To investigate the root cause of the WOW switch, a systematic procedure is followed as described below Fault isolation On 16th September 2003, the air data computation (ADC) and ÔYÕ light came on during the pre-flight check of aircraft number 885. The flight control system (FLCS) self-test was unable to reset after 5 min and this confirmed the failure. The mission of the pilot was aborted as a result. The technicians did simple contact checks on the WOW switch using a multimeter and the switch was isolated to be the problem. Subsequently, the WOW switch was replaced and follow-up checks on the aircraft were found to be good Preliminary visual examination Preliminary visual examination of the device under analysis (DUA) was performed and no visible cracks or openings were found on the stainless steel housing. The protective rubber tubing was already removed from the wire leads by the technicians and only a short length of the wire leads remained after the DUA was removed from the aircraft. Although all the wire leads were intact, there were three wire leads that show signs of insulation deformation. Fig. 2. X-ray of good WOW switch.

4 3.3. Preliminary contact testing The wire leads were separated into the four branches of the 4PDT circuit. Initial contact testing shows that two of the branches were not working and hence confirmed that the DUA is a true failure. It was also observed that these are the two branches contained deformed wire insulations. Thus, one of the potential causes of failure could be related to wire anomalies Dimensions, specifications, mechanism, etc. Detail background information of the defective item is necessary in the process of failure analysis. In this case, detailed specifications, dimensions and sectional drawing of the WOW switch were obtained from its manufacturer. Fig. 3 shows the sectional drawing of the WOW switch obtained. Characterised of the WOW switch such as electricals rating, physical range and the mechanical features provided a better understanding of the limits of the switch. For example, it is possible to determine whether the wear-out phase of the DUA has already exceeded its mechanical life by referring to the mechanical features of the WOW switch Further non-destructive failure analysis C.M. Tan et al. / Engineering Failure Analysis 12 (2005) Brainstorming and fish-bone analysis show that the likely causes of the failure could be due to vibration effects of the aircraft, wire anomalies, routing of the harness, environmental effects, constant switch depressing and its wear-out phase exceeding its mechanical life. The mechanical life of the switch is stated at 50,000 Fig. 3. Sectional drawing of WOW switch.

5 512 C.M. Tan et al. / Engineering Failure Analysis 12 (2005) cycles. As up to date, the WOW switch would have been subjected to only approximately 4000 cycles. This is less than 1/10 of the stated mechanical life of the WOW switch. Furthermore, the ADC light disappeared when switch was tapped. This shows that when the switch was subjected to mechanical shock or vibration effects, the failure ÔdisappearedÕ. Thus, it is highly possible that the vibration effects of the aircraft on power up caused the signals from the WOW switch to be cut off intermittently. Hence, possible root causes of the switch include failure of the internal mechanism of the switch and presence of intermittent wires Mechanism failure Radiographic inspection was carried out to determine the internal state of the DUA. Figs. 4 and 5 shows the X-ray images obtained. It is observed that there are no abnormalities in the spring mechanism of the DUA from the X-ray images. The X-ray images confirmed that all the micro-switches transferred to the NO contacts after the pin plunger was actuated. Thus, it is unlikely that mechanism failure is the root cause Wiring failure Constant flexing of wire leads Aircraft components like the WOW switch are subjected to a lot of vibrations in the aircraftõs operating environment. In addition, there is free movement of the harness along the part where it exits out from the WOW switch to the part where it is locked on to the strut of the main landing gear (MLG). This constant vibration of the harness along this part causes frequent flexing at the part where the wire leads exit from the main body of the switch as shown in Fig. 6. Since the wire leads that exit from the base of the WOW switch are in close contact with the opening of the stainless steel housing, frequent flexing of the wire leads against the metallic opening could cause wire chafing. From the optical micrograph of the wire leads as shown in Fig. 7, the wire insulation at the opening of the housing has exposed the wires within and it is very likely that additional flexing has broken the copper wires. Fig. 4. Before actuating pin plunger of DUA.

6 C.M. Tan et al. / Engineering Failure Analysis 12 (2005) Fig. 5. After actuating pin plunger of DUA. Fig. 6. WOW switch at MLG Temperature cycling of wire leads The DUA was subjected to temperature cycling in the aircraftõs operating environment. Temperature is lower at higher altitudes, resulting in the contraction of material. Conversely, temperature is higher at lower altitude, leading to expansion of material. When the epoxy at the base of the switch housing contracts, it causes the wire leads to be pulled inwards and vice versa. Hence, the copper wires are subjected to a series of cyclic loading due to this temperature cycling of the aircraft. This could result in fatigue Routing of the harness When routing the harness of the WOW switch on the MLG, installation standards must be followed to ensure that the harness are not entangled or squashed. However, a required length of the harness (Fig. 8) is not stated in the techniciansõ job guide. Hence, if that part of the harness is routed such that there is greater

7 514 C.M. Tan et al. / Engineering Failure Analysis 12 (2005) Fig. 7. Deformed wire leads of DUA. Fig. 8. MLG WOW switch. free movement, the wire leads will experience greater degree of flexing due to the vibration effect of the aircraft. On the other hand, if there is little free movement of the harness as shown in Fig. 9, the wire leads will be subjected to higher bending stress. Both scenarios could result in wire chafing upon prolonged vibration of the wire leads. Microscopic images of the 2 intermittent opened wire leads show that one of them has an insulation defect which occurred at a short distance away from the base of the WOW switch (Fig. 10) while the other has defect occurred at the opening of the switch housing (Fig. 7). These two parts of the wire leads experience higher bending stress as compared to the other parts due to the routing of the harness (Fig. 8) Production related Typical WOW switch has a stainless steel casing for the wire leads with a polymeric material surrounding it as shown in Fig. 11. This feature, which provides added protection against chafing, was found to be missing from the DUA. This design deficiency is believed to have contributed to the wiring failure. Improper combing of the wire leads before being enclosed in the protective rubber tubing could have resulted in kinking. This induced torsion to the twisted portion of the wire. The vibration effects of the aircraft and the

8 C.M. Tan et al. / Engineering Failure Analysis 12 (2005) Fig. 9. Routing of harness of WOW switch at MLG. Fig. 10. Deformed wire lead of DUA. Fig. 11. Good WOW switch with stainless steel casing. routing of the harness could cause further straining of the wire leads, hence resulting in wiring failure. Another effect of improper combing is the entangling of the wire leads. Example of such an improper combing was found in another non-faulty WOW switch as shown in Fig. 12. A simple experiment was conducted by

9 516 C.M. Tan et al. / Engineering Failure Analysis 12 (2005) Fig. 12. Improper combing of wire leads. Fig. 13. Side view of experiment set-up. Fig. 14. Sample wire lead used in experiment. clamping three wire leads. Similar marks of bending stress were observed on the entangled wire lead (Fig. 14) when the wire leads were bent as illustrated in Fig Root cause identification & verification 4.1. Radiographic inspection of the wire leads Failure analysis up to this point shows that it is likely that the wire leads of the DUA have failed. Hence, radiographic inspection was done on the wire leads of the DUA. Results of the X-ray inspection (Fig. 15)

10 C.M. Tan et al. / Engineering Failure Analysis 12 (2005) Fig. 15. X-ray of DUA wire leads. shows that there were two intermittent opened wires (labeled Wire 4 and Wire 7). These two wire leads were two poles of the DUAÕs 4PDT circuit mentioned above. Wire 4 was the one with slight deformation but with insulation still intact (Fig. 10). Wire 7 was the one with exposed broken wires (Fig. 7). With these two poles intermittently opened, the 28 VDC that was supplied to the poles would not be able to reach the opposite side of the contacts. Thus, it is likely that the flight control system of the F-16 detected that the circuits along the two poles were not receiving the expected signals, causing the ADC light to appear Destructive failure analysis The DUA was cut to verify that there were no abnormalities with the mechanism of the switch. Contact checking for the DUA was done in the absence of the wire leads this time. Result shows that all the microswitches were functioning properly. This verifies that there were no abnormalities with the mechanism of the DUA. Scanning electron microscope (SEM) analysis was conducted on Wire 4 and Wire 7 at the points where the open intermittency was observed. Fig. 16 shows that the insulation of Wire 4 had a tear mark that is at an angle from the direction of the wire lead. The direction of the tear indicates that Wire 4 has been subjected to torsion [3]. SEM of Wire 7 shows that the exposed broken copper wires had relatively flat and smooth surfaces (Fig. 17). This is an indication of fatigue failure of the copper wires Further inspection with optical microscope Further microscopic inspection of other wire leads shows visible marks of bending stress on the insulations at the opening of the DUA housing. One of the wire leads had also tear at the bending stress point as shown in Fig. 18. This tear is similar to that of Wire 7 as both tears propagated from that point onwards Root cause identification From the above investigation, the main contributing factor leading to the failure of Wire 4 is the improper combing of the wire leads. Kinking of Wire 4 is a result of improper combing and this lead to torsion

11 518 C.M. Tan et al. / Engineering Failure Analysis 12 (2005) Fig. 16. SEM of Wire 4. Fig. 17. SEM of Wire 7. Fig. 18. Visible bending marks on wire leads.

12 C.M. Tan et al. / Engineering Failure Analysis 12 (2005) acting on the twisted portion of the wire lead. In addition, the twisted portion is located at the point of bending where the bending stress is the higher. Prolonged vibration effects of the aircraft have further enhanced the bending stress, leading to intermittent open wire defect. The main contributing factor leading to the failure of Wire 7 is the design of DUA. Without the stainless steel casing for the wire leads, the wire leads at the opening of the switch housing were vulnerable to chafing. Moreover, the wire leads were located at the bottom of the switch housing where the bending stress is higher than other parts of the wire lead due to the routing of the harness. In addition, this part of the wire lead experienced the effects of vibration to a greater extent due to the absence of the casing. The wire leads were also subjected to temperature cycling in the aircraftõs operating environment. Prolonged vibration and temperature cycling over time resulted in fatigue failure of the copper wires. Thus, the root causes of failure of the F-16 WOW switch are the improper combing of wire leads during production and the design deficiency of the WOW switch. 5. Proposed improvement in maintenance 5.1. Use of WOW switch with stainless steel casing All WOW switches used on the MLG of F-16 should be checked to ensure the presence of the stainless steel casing for the wire leads Improved routing of harness The length of the harness which is not specified in the job guide should be optimized such that bending stress on the wire leads are minimized when the WOW switch is installed on the aircraft. The length of the harness should be routed such that it is between that of Figs. 6 and 9. In addition, clamping that length of the harness to the strut could minimize vibration effects of the aircraft. 6. Conclusions The root causes of failure of the F-16 WOW switch are the improper combing of wire leads during production and the design deficiency of the WOW switch. Proposed improvements in maintenance include the checking of WOW switches for the stainless steel casing and improving the routing of the harness to minimize vibration effects and bending stress. References [1] Report from Norton Air Force Mishap Database, Report No ; 1986 Nov [2] Martin Perry L. Electronic failure analysis handbook. New York: McGraw-Hill; [3] Sachs Neville. Root cause failure analysis understanding mechanical failures, Reliability Magazine; June Available from