I like the idea of 4340 with 40-45 HRC. Your aim is right on the mark.
If this is really going to take a beating you might want use the aerospace remelted grade of steel.
We used to make tooling in-house out of 4335V VAC+ARC and HT into this range.
They lasted forever.
These will need very heavy walls, and don't forget the internal fillets at the corners.
Rough machine, then HT, and then finish machine.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

That big socket and nut are going to have a lot of inertia and will absorb a lot of energy from the impact and little will translate into tightening. Before you go the custom socket route have to considered off the shelf options like Superbolts?

Ed,
Thank you, I'm not familiar with 4335 but I'm thinking the 4340 is probably best for strength as well as impact/fracture resistance. The walls are 9/16" thick. I found an old ASM book on heat treatment and it looks like normalizing at 1600F, quenching from 1525-1575F and tempering 2 hours at 900F will give HRC41-43 as well as toughen it for impact.

Tug,
That's a good point but I don't think there's much that can be done about it. Probably part of why they don't make impact sockets this size - diminishing returns, along with weight, cost, demand, etc.
Thanks for the idea, Superbolts are interesting but this is for an existing design and they won't fit. Worst case, we go back to using the impact spanner and hammer, my main concern is this thing taking 3000 ft-lb torque...

I might go 950F on the temper.
There are tables that you can use that relate the as-quenched hardness to correction in temper temperatures.
This is a ref pulled from and AMS spec.
https://res.cloudinary.com/engineering-com/raw/upload/v1713552062/tips/steel_HT_opyatz.docx

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

For comparison, a 1-1/2 - 6 thread is torqued to 3000ftlbs and has a 2-1/4 inch nut. At 7 inches your stresses in your socket are not going to be very high.

Interesting Ed. What I'd like to know now is how temper time affects toughness. Hardness tends to drop the longer you temper, so does toughness increase correspondingly? And does it make sense to double temper something like this?
Tug are you implying our 1-1/2" impact driver isn't beefy enough?

Toughness is the amount of energy it takes to deform the material.

Make it soft so it can deform a lot, but has no strength = low toughness
Make it hard so it cannot deform, but it has high strength = brittle

From one end to the other the toughness increases to a maximum and then decreases again.

There could be a large or small range of suitable toughness, depending on the required toughness.

Temper too long from the low-yield/high strength brittle condition and you get to the high-yield/low strength condition and it goes from shattering to just plastic deformation.

sublime
I would suggest a shock resistant tool steel, and I would make extras to destruct test. to obtain
actual value, an other suggestion would be A-2 tool steel, very little distortion during heat treat.
but are only suggestions and you need to test out. I made all my tooling mostly with A-2, the second choice would be 4340M and I would core harden then induction harden the wear surfaces.

Have we heard what the expected torque will be?

If there is room for a socket with a comically large OD and resulting immense wall thickness I think it could be made of stuff not much fancier than the bolt/nut.

Dave what I'm referring to is the Izod impact energy, which generally increases with tempering temperature.
mfgen 4340 is impact resistant, more so than A-2 I believe. Induction hardening may be a good idea but do impact sockets typically get any surface hardening?
Tmoose, it can't be more than 3000 ftlb. We don't know exactly what the nut is made of though.

Often impact sockets are surface hardened.
It prevents fretting and scuffing that roughen the surface.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

True surface hardening probably wouldn't hurt, but not sure if we need it for something that won't get used extensively (about 4 times a month).

But now I have another question: should the part be stress relieved after tempering, and how/when? Will the hardness or strength be affected if, after tempering at 950F and cooling back down to room temperature, it is then subjected to a couple more hours in a furnace at about 890F for stress relief? I imagine this will happen just before finish machining.

I would just double temper.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

Again at 950F for a couple hours, one right after the other? Or maybe just 1 hour the second time. The thickest part is approximately 2 x 1 inches cross section. But what'll a second temper do to the hardness/strength/impact resistance? Pretty sure the hardness and strength will drop further. If I do any surface hardening should that be before the second temper then?

Just pick a material that work hardens a lot. If it deforms at all it will get stronger.

3000ftlbs on an 8 inch socket? You could get away with low carbon steel. This is the diameter of nut we used to use on our boat propellers that get tightened with torch cut A36 wrenches.

Your problem is going to be on the drive side of the socket. You mentioned something about 1-1/2 inch. That's much smaller than 8 inch and is going to be your point of failure.

However, you are wasting your time. You will not achieve your torque values with an impact with a socket that has so much inertia.

It looks like we won't need the second temper for this material:
https://www.eng-tips.com/viewthread.cfm?qid=172445

You're right about the point of failure Tug, but it's also thicker at that end. We're going to try it, hopefully it's not a waste.

A second temper isn't really needed, but I always like when I am up at these strengths.
If you are worried about strength, then use 925F for your tempers.
And yes, I have been thinking about the drive end all along.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed