We all have either heard about or personally experienced the perversity and
seemingly organic behavior of machines. As though haunted by some infernal
spirit, machines sometimes do the unexpected, as was the case here. Here is an
example of the latter.
Scene of the Crime
In a Texas city, the public works department used a wheeled, skid-steer loader for grounds-keeping at a city park. On this loader, direction and speed were controlled by left-hand and right-hand levers that varied hydraulic fluid flow to two hydraulic motors.
Pushing both levers forward allowed forward travel, with speed proportional to lever position, and similarly, pulling both levers back allowed travel in reverse. Pushing or pulling one lever more than the other varied turn radius according to the difference in the movement of the two levers. Pulling one and pushing the other made the machine turn in its own length. Bucket lifting and tilting functions were controlled by a handle at the top of each lever. When released, and with the steer levers in a neutral position, these handles engaged a neutral detent. A hand throttle lever set engine speed. After sitting down, the operator engaged a restraint bar that released a parking brake.
While stripping turf and soil, the loader driver stopped to help the grounds crew move some bags of mulch and flats. Leaving the engine running at above idle, the driver released the control levers, raised the restraint bar, and climbed out.
As he stepped down, the loader became the mechanical embodiment of perfect perversity, running in a circle, striking the driver and mangling his leg. The driver was quickly pulled to safety, and with whoops and shouting all-around, the work crew all contributed ideas for stopping the machine, but no one with courage (or lack of good sense) stepped forward to try. Finally, the crew boss grabbed a rake and as the loader circled by hit the hand throttle to idle speed position. With the loader at barely walking speed, the crew leader made like an ancient Minoan bull-leaper and avoided the moving wheels, jumped in and shut off the loader.
The Investigation
The injured driver filed a suit naming the loader manufacturer and selling/servicing dealer. His attorney retained me to inspect the machine to try to determine the cause of the runaway. With operator's, maintenance/repair, and parts manuals, and in the company of other experts and attorneys, I began my inspection. At the request of those present, a park facilities driver ran the machine. This test confirmed the job site runaway, except that our driver stayed aboard and prevented the machine from running wild.
With the floorboards and covers removed, I could see that dirt and debris had filled the space below. It prevented the handle on top of each steering lever from engaging the neutral detent. This explained the uncontrolled travel, but not the circular travel.
The service and parts manuals illustrations showed the parking brake depended on spring-loaded pins that engaged holes in left and right hand discs. Each disc was on a shaft turned by a left and right hand hydraulic motor with each shaft coupled to the two wheels on that side by a chain drive. Each pin, which moved in and out of a bushed hole, was cable-operated by the restraint bar.
Further inspection revealed that one of the locking pins was jammed in the bushed hole due to debris. The resulting friction overcame the spring force and prevented the pin from fully engaging or engaging the disc at all. When parking, only the nose of the pin engaged the disc and then popped out allowing the wheels on that side to turn, explaining the loader's circular path.
|
Click
here to view a detailed drawing of the
original brake and the proposed
changes.
|
The Culprit
My opinion report cited faulty design of the enclosures that allowed dirt and debris, a normal part of a loader's environment, to jam the brake and transmission controls. This dirt had not been cleaned by technicians during routine maintenance. In my report, I proposed that the operator's seat be linked to positively move the pump swash-plates to a neutral position, or to use a seat-operated valve to dump fluid flow to the hydrostatic pump units to make return-to neutral a virtual certainty with the seat unoccupied. Also, rather than cables and spring-loaded pins, I proposed solid linkages from the restraint bar to the brake pins.
There was no infernal deus ex machina at work here, only a design failure. This case settled prior to trial.
Reach Myron J. Boyajian, P.E., at ntesla@ieee.org.