This study was conducted to examine the effect of flashing conditions on the resulting temperature profile during flash welding automobile door frames. Previous work on temperature profiles of flash welds has shown that at some point in the welding cycle a steady state temperature is reached, minimizing the need for further flashing. The indication of such a minimum flashing time allows flashing conditions for any application to be optimized. Unfortunately, previous work has been limited to rather heavy section materials, and the results could not be directly applied to the flash welding thin sections typical for door frames.This program was a preliminary study to examine the effects of initial flashing velocity and flashing acceleration on the resulting temperature profiles in U-shaped channel sections. Work was done on a cam driven flash welding machine supplied. Flashing conditions were varied by using cams with different profiles. The first cam was a linear cam typical of that used for production of this part. In addition, three new cams were prepared. These included a cam with a linear flashing profile using twice the initial flashing velocity of the production cam, and two cams with parabolic flashing profiles. Both parabolic cams profiles used half the initial flashing velocity of the original cam. All cams used roughly the same flashing time, resulting in different degrees of burn-off.Temperature measurements were made by locating thermocouples at increasing distances from the original interface. During welding, temperature was recorded from these thermocouples as temperature versus time data. For each set of flashing conditions, separate experiments were run to characterize temperature profiles on the three sides of the channel section. Resulting temperature versus time data was then transformed to temperature versus distance from the flashing interface data. This data was curve fit to an exponential expression, and the results of that curve fit used to calculate effective forge depth of the workpiece.The results of this program show that the parabolic cams reach steady state temperature faster than the linear cam, and are consistent with the published literature on heavier section flash welding. However, it was also noted that excessive flashing velocities (as a result of either a high initial flashing velocity or too high a flashing acceleration) tended to overheat the part. Higher flashing velocities result increase the rate of heat generation during flash welding. It is believed that the relatively long stick-out lengths used in this application, and the relatively poor conductivity of the dies prevented heat from being extracted at a sufficient rate to maintain a stable temperature profile. Apparently, heat generation needs to be balanced between that necessary to achieve stable temperature conditions in a reasonable time, while not overheating the workpiece.