This month's issue of Gear Technology covers a subject engineers love to learn about from others but hate to learn about through firsthand experience: gear failure. In a broad sense, all engineering is concerned with failure. Eventually, all parts fail, so engineers need to determine the limits of their design and ensure that it will meet the requirements of the particular application.
I would appreciate if you could assist with a gear failure (occurring)
after just seven weeks in service, post installation. This driving gear
wheel has been installed in a medium-speed engine with backlash
present at four different positions; with additional backlash checked on
the mating surfaces. All backlash was found within (OEM)-recommended values. Please note included photos - it seems that the crack has started at the root fillet. Any comments would be appreciated.
Point-surface-origin (PSO) macropitting occurs at sites of geometric stress concentration (GSC) such as discontinuities in the gear tooth profile caused by micropitting, cusps at the intersection of the involute profile and the trochoidal root fillet, and at edges of prior tooth damage, such as tip-to-root interference. When the profile modifications in the form of tip relief, root relief, or both, are inadequate to compensate for deflection of the gear mesh, tip-to-root
interference occurs. The interference can occur at either end of the path of contact, but the damage is usually more
severe near the start-of-active-profile (SAP) of the driving gear.
Effective gear designs balance strength, durability, reliability, size, weight, and cost. Even effective designs, however, can have the possibility of gear cracks due to fatigue. In addition, truly robust designs consider not only crack initiation, but also crack propagation trajectories. As an example, crack trajectories that propagate through the gear tooth are the preferred mode of failure compared to propagation through the gear rim. Rim failure will lead to catastrophic events and should be avoided. Analysis tools that predict crack propagation paths can be a valuable aid to the designer to prevent such catastrophic failures.
In his Handbook of Gear Design (Ref.1), Dudley states (or understates): "The best gear people around the world are now coming to realize that metallurgical quality is just as important as geometric quality." Geometric accuracy without metallurgical integrity in any highly stressed gear or shaft would only result in wasted effort for all concerned - the gear designer, the manufacturer, and the customer - as the component's life cycle would be prematurely cut short. A carburized automotive gear or shaft with the wrong surface hardness, case depth or core hardness may not even complete its basic warranty period before failing totally at considerable expense and loss of prestige for the producer and the customer. The unexpected early failure of a large industrial gear or shaft in a coal mine or mill could result in lost production and income while the machine is down since replacement components may not be readily available. Fortunately, this scenario is not common. Most reputable gear and shaft manufacturers around the world would never neglect the metallurgical quality of their products.
Gears are manufactured with thin rims for several reasons. Steel gears are manufactured with thin rims and webs where low weight is important. Nonmetallic gears, manufactured by injection molding, are designed with thin rims as part of the general design rule to maintain uniform thickness to ensure even post-mold cooling. When a thin-rimmed gear fails, the fracture is thought the root of the gear, as shown in Fig. 1a, rather than the usual fillet failure shown in Fig. 1b.