I would greatly appreciate your opinions/insight/suggestions regarding the following: (will provide free shop tour and donuts for all that take the time to read this post!)

We occasionally (like 6 parts per year or less) manufacture a part that requires welding an AISI 4340 shaft to an ASTM A572 Grade 50 plate. The last time these were made at our facility was before I began working here. These parts "snuck up on me"; I recently discovered they were being prepared for welding, and we do not have a WPS for this 4340 material. Now, before we go any further, let me clarify the function of the weld between these parts:

--The 4340 shaft is machined to a diameter of 1.50" with a tolerance of +-.002". It is then inserted into a hole that is machined through a 3/4" thick, A572 plate. This hole is also machined to 1.50", with a tolerance of +-.026". This is an "interference fit" which requires heating of the 3/4" plate in order to get the shaft inserted through the hole, which then protrudes 1/4" past the back of the plate. This is where the connection of the two parts gets its strength. In order to keep this shaft from ever moving or rotating, they then require a 1/4" fillet weld (see attached pic) around the shaft where it protrudes out the back of the plate.

--Our Engineers (Mechanical, Civil, and Electrical, we do not have a Welding Engineer on staff) have specified on the drawing that the parts should be pre-heated to 250 F before welding, and then welded with E11018 using the SMAW process (this is the only part we ever use this process and electrode for). There is no requirement for PWHT or stress relieving of any kind, no instructions for slow-cooling, nothing.

--Our Engineers do not consider this weld to be critical, it is merely a way to "retain" the shaft in its position in the 3/4" plate, and I get the feeling they think I am worrying about nothing when I questioned their preheat, and lack of post-heat, requirements. Additionally, they claim they've had zero issues with these parts in the past. Not once has one come back with cracked retainer welds. So, here are my questions:

1--Do I worry about creating a WPS (which would have to be qualified by testing) for such a non-consequential weld that is performed maybe 6 times per year?
2--If I do create and qualify a WPS, what code do I use? Note: there is actually another weld that is performed on this 3/4" plate earlier in the manufacturing process. It is on standard AWS recognized materials and is done per AWS D1.1 (yes, we do have a WPS for this)
3--If I do run a test weld to take to the lab, where would I find AISI 4340 material in plate form? Other options? Am I missing something?
4--Do you agree with the pre-heat and post-weld cooling methods? What about the use of E11018?

I like the idea of undermatching the filler metal and I like the idea of using higher preheat.

The idea of running a few fillet weld samples would be the least testing I would do. Use the materials you use for production. Weld the samples replicating the joints used for production. Allow the welds to sit for a couple of days and then cut them and prepare them for macroetchs. Use fluorescent penetrant to check for underbead cracking. The macroetch and fluorescent penetrant testing should determine if there is cause for concern with underbead cracking. You can also have a could of cross sections tested for hardness using a microhardness test to determine whether or not you are developing martensite in the HAZ.

Al