KyleT said:
the materials welded are in compression, not tension or shear. the weight of the roof sits vertically on the a-pillar and is welded.
OK, I had a feeling that might be what you were trying to say. But your assessment of the loads involved is only partially correct and misses the real causes of the damage. Yes, the static loading is compressive, but materials rarely fail under a static loading scenario. When you consider dynamic loads (under acceleration, deceleration, turning and driving on sloped surfaces) however, you do get large amounts of shear and tension, torsion as well as compression on that small and focused weld area, which is why I suggested to bridge the cutout where the post cuts through the area below the drip rail. Dynamic loads on the weld would be increased (probably exponentially) with any serious amount of weight on the roof. Simply welding the existing crack will not address the load distribution problem is you want to continue carrying cargo up there. If you retain the original design and only repair the existing weld, you will need to greatly reduce the amount of cargo you put on the roof, or drive the truck "on eggshells" when the roof is loaded up with gear. The only way to ensure that weld is subjected to static loads only is to keep the truck parked.
Kev