Sunday, May 6, 2018

Driverless cars don’t need steering wheels - but the rest of us sure do

















Sometimes the redesign of a product (or changes in materials and processes) results in unanticipated problems. The Ford Fusion is a mid-sized sedan which first was produced in August 2005. A second-generation model was introduced in 2013. A single 10mm diameter bolt holds the steering wheel onto the internally threaded steering shaft, as is shown above.

In USA Today on October 27, 2017 there was an article titled Probe of Ford Fusion steering wheels that may loosen, detach which reported on an investigation of ~840,000 vehicles and included this startling statement:

“A person in Georgia told the National Highway Traffic Safety Administration [NHTSA] that a steering wheel fell into their lap in a 2015 Fusion when turning into a gas station on Sept. 23.”


Another article in USA Today on March 14, 2018 was titled 10 times Ford steering wheels came loose or off, causing massive recall. The problem turned out to be bigger than expected. Ford announced a recall on that day:

“Ford is issuing a safety recall in North America for approximately 1.3 million 2014-18 Ford Fusion and Lincoln MKZ vehicles for potentially loose steering wheel bolts that could result in a steering wheel detaching from the steering column. In affected vehicles, the steering wheel bolt may not maintain torque, allowing the bolt to loosen over time, and if not serviced, a steering wheel could potentially detach and lead to a loss of steering control and increased risk of a crash. Ford is aware of two accidents with one injury allegedly related to this condition.”




















The NHTSA recall report had more details. As shown above, it said the chronology was that:

“Ford discovered a design change to the steering wheel fastening system in 2013 for 2014-2018 model year Fusion and MKZ vehicles. The amount (length) of threads inside of the steering wheel shaft decreased by 5 mm. The thread reduction is in the end of the steering shaft where the bolt first contacts the threads internal to the steering shaft. The bolts used to secure the steering shaft have a nylon patch prevailing torque feature to ensure proper torque retention. With the reduced amount of threads inside the steering shaft and if the nylon patch is located towards the head of the bolt, the nylon patch may not fully engage the threads to ensure proper torque retention.”

“The remedy bolt is 17 mm longer, providing more robust steering column thread and torque retention patch engagement. The remedy bolt also has a 13 mm longer torque retention feature to ensure proper engagement.”    

A bolted joint is held together by friction, which can be increased by adding a nylon patch. A web page for the ND patch process describes that feature:

“…fasteners are heated and sprayed with a custom nylon powder which adheres to the part. When assembled with a mating part, the engineered plastic nylon patch is compressed.  Due to the elastic memory it resists this compression and acts like a wedge, increasing the metal to metal contact 180° opposite the material. This mechanical force creates a strong, yet fully adjustable lock which will not weaken, even under extreme vibration.”















Another way to add friction to a joint is with a Nyloc nut, as shown above. When a bolt is  turned past the internal threads of the nut and into the unthreaded nylon in the locking feature, it forms new threads.

Almost two decades ago I encountered a situation where such a nut was not seated properly and led to a crash. It was on the elevator control linkage for a Glaser-Dirks DG-800B motor glider, N98NL. The crash occurred on June 21, 1998 at Jean Nevada, and the National Transportation Safety Board (NTSB) ID was LAX98LA209. The glider had four hours of flight time after manufacture.  An FAA inspector examined the aircraft. He found the bolt in the fuselage and recovered it, but could not find the nut. The NTSB Factual Report says:
   
“The bolt and exemplar parts were examined with an optical microscope at SEAL Laboratories in Los Angeles, California, on July 31, 1998, in the presence of the Safety Board. A copy of the laboratory report is appended to this file.

According to the metallurgist, the first three threads of the accident bolt did not have debris in the roots, and installation of an exemplar locknut in the hand-tightened position would cause the nut threads to wipe material from the roots of the first three bolt threads. For proper tightening of the locknut, the threads must be engaged over the full length of 6 mm (six turns) so that the blue plastic lining which provides thread locking is deformed. This would require three more turns than are indicated by the lack of debris in the thread roots. The metallurgist concluded that a nut was installed on the bolt but not fully tightened.”

In this case, meticulous handling of the evidence by that FAA inspector, who bagged and tagged the bolt, made it possible for me to provide a definitive answer.

The image of a Nylon lock nut came from Wikimedia Commons.

No comments:

Post a Comment