I think you're essentially right, but I feel it's worth pointing out that that welds/ joints are not the only cause of stress concentrations, so what you're saying is incorrect. Also, it is already possible to have a stiff front end and compliant rear. Geometry can vary the stiffness of the component and create stress risers all of its own.
I was saying that stress concentrations are almost always the cause of failure. I think this is pretty much correct. I know geometry can cause them too, but welds will have horrible residual thermal stresses and any mechanical join will have similar peak stresses.
I was saying that stress concentrations are almost always the cause of failure. I think this is pretty much correct. I know geometry can cause them too, but welds will have horrible residual thermal stresses and any mechanical join will have similar peak stresses.