A series of six whole-body sled tests were conducted with unembalm-ed cadavers to study knee impact phenomena under simulated car crash conditions. The purpose of the study was to investigate the occurrence of upper leg fracture types not seen in stationary subject knee impact tests (see UM-HSRI-76-33) using a moving-mass impactor technique, and to extend the pulse durations of the knee forces to times greater than 20 ms, The tests were successful in achieving both goals. Major conclusions drawn from the test series were:​

• Proximal upper leg injuries (femoral neck and shaft and pelvic fractures), which account for most upper leg injuries in real crashes, occurred with the whole-body knee impact test conditions used in this study.
• Femoral neck and acetabular fractures appear to produce a change in the rate of load transfer from the knee to the pelvis, while femoral shaft and extensive condylar fractures produce an abrupt drop-off in the load transferred from the knee to the pelvis. In these tests, neck and acetabular fractures, when they occurred, preceded the shaft and condylar fractures.
• The forces necessary to produce proximal leg fractures are similar in magnitude to those that produce distal knee injuriesin impactor tests.