Search Results

New high-temperature Mg alloys are being considered to replace 380 Al in transmission cases, wherein bolt-load retention, and creep, is of prime concern. One of these alloys is die cast AE42, which has much better creep properties than does AZ91D but is still not as creep resistant as 380 Al. It is thus important to investigate bolt-load retention and creep of AE42 as an initial step in assessing its suitability as a material for transmission housings. To that end, the bolt-load retention behavior of die-cast AE42, AZ91D and 380 Al have been examined using standard M10 bolts specially instrumented with stable high-temperature strain gages. The bolt-load retention test pieces were die cast in geometries approximating the flange and boss regions in typical bolted joints. Bolt-load retention properties were examined as a function of time (at least 100 hours), temperature (150 and 175 °C) and initial bolt preload (14 to 34 kN).

The use of die cast magnesium for automobile transmission cases offers promise for reducing weight and improving fuel economy. However, the inferior creep resistance of magnesium alloys at high temperature is of concern since transmission cases are typically assembled and joined by pre-loaded bolts. The stress relaxation of the material could thus adversely impact the sealing of the joint. One means of assessing the structural integrity of magnesium transmission cases is modeling the bolted joint, the topic of this paper. The commercial finite element code, ABAQUS, was used to simulate a well characterized bolt joint sample. The geometry was simulated with axi-symmetric elements with the exact geometry of a M10 screw. Frictional contact between the male and female parts is modeled by using interface elements. Material creep is described by a time hardening power law whose parameters are fit to experimental creep test data.

In this paper, the effects of fluctuating torque on loosening of a tightly seated bolt are investigated. Tests over a wide range of bolt stresses and loosening torques are reported and equipment developed for determination of such effects is described. It is shown that a definite functional relationship exists between the stress on a typical bolt, the oscillatory loosening torque that is applied, and the number of cycles before the bolt becomes loose. The effects of these relationships follow a clearly defined law, although they are, of course, influenced by a number of additional variables.