I posted a topic at the welding and metal fabrication forum on one way to restore a bore in a stamped steel piece: "Weld Mold 26-C 'How-to' Oxy-Acetylene Repair: Restoring a Stretched Bore in Stamped Steel". In that 'how-to', I mention heat treated metals and also refer forum members to one of the magazine's slideshow video presentations.
The slideshow discusses a major concern when welding, brazing or silver brazing (hard silver soldering) near heat treated parts. Many automotive parts, especially wear points like gear teeth, shafts, splines, thrust washers and running surfaces, have been heat treated to the depth and hardness required.
When we weld near any heat treated alloy metals or forged parts, there is always concern about damaging the heat treatment. Even the use of specialized, hard alloy filler materials (some as high as 140K or more tensile in the weld) will not prevent problems at the nearby heat treated areas. If you raise metal temperature high enough during the welding process, any through- or case-hardening will be lost. This means that the metal will soften and be rendered either unsafe or no longer capable of handling its intended function, especially wear points like splines, gear teeth, shafts or thrusts. It is absolutely certain that adjacent to a metal fusion weld, any heat treatment or case hardening will be lost.
If parts like a gear or shaft are heat treated or "case hardened" and need welding, you must first "normalize" the metal. This is similar to annealing, but is intended to simply reduce the hardness in the case area, typically the surface 0.030"-0.040" zone if we're talking about common automotive transmission gears and shafts. A ring-and-pinion gear set, due to the size of the gear, is often much deeper case hardening, and some components, especially hardware and fasteners, even get "through-hardened" as opposed to case hardened.
Once normalized, a gear or shaft can be welded with an appropriate filler that matches the base metal material. If the match is correct, the part can be machined before re-heat treatment, then heat treated to the component's original Rockwell hardness (prior to normalizing) and case hardening depth if dealing with a case hardened piece. For selecting niche filler materials on alloys and exotic metals, I turn to Weld Mold Company.
One way to know the original hardness is a Rockwell C test before normalizing. Determine the depth of case hardening and adjust the final heat treatment accordingly. In the magazine slideshow video, I show and talk about an 8620 cluster gear repair process.
There is much more to say about metal prepping, niche welding filler materials and heat treatment...Looking forward to a dialogue in this subject area, we're just beginning! If you have questions, please share.