Use partial penetration weld.

The larger weld will have to be done in multiple passes to stay code compliant (assuming AWS D1.1 or something similar), this limits the heat input and helps alleviate the issues you bring up. The bigger issue is simply the economy of the weld, which is a function of the amount of weld filler metal and time it takes to complete the weld (as well as material preparation).

If you do need to develop the member capacity and a CJP weld isn't required for fatigue purposes you can provide a PJP weld with reinforcing fillet or just a fillet weld. Each shop may have it's own preference on this (so you may want to leave it up to them to decide, simply call for a weld that develops the strength and let them decide). For the sizes we are talking here, if the shop had to grind the tube end down to prep it for the PJP weld I suspect they would prefer just a fillet weld. However if the shop has automation (tube laser) that does the end prep for them, I suspect they would prefer the PJP with fillet overlay as this results in less weld passes than the fillet alone. Below is a Figure from the AISC Design Guide 21 that shows the amount of filler metal for the different options all with the same strength quite well (keep in mind only the 2nd and 3rd images are options if the full strength is needed):

I believe that in most cases you cant develop full HSS member capacity with a fillet weld alone, no matter how large the weld is, as the base material controls. Take a look at effective size of a fillet (Deff) defined in section 2.3 of the AISC design guide 24. For 5/16 thick 46 ksi HSS and 70ksi electrode Deff=5.44/16. Once that value is reached, no matter how much more you increase the fillet, it does not add to the weld strength, as shear rupture of the HSS member controls.

T-joint between HSS members is even worse, as on top of the weld strength limitations, there is only a certain "effective weld length" you can count on (AISC TABLE K4.1)

Quote (RabitPete)

I believe that in most cases you cant develop full HSS member capacity with a fillet weld alone, no matter how large the weld is, as the base material controls.

This is essentially the definition of developing the member capacity. The weld is sized such that the member will fail before the weld does. There is no reason to provide a larger weld as that is the size it takes to develop the member capacity.

Quote (RabitPete)

T-joint between HSS members is even worse, as on top of the weld strength limitations, there is only a certain "effective weld length" you can count on (AISC TABLE K4.1)

This is certainly a valid point, however it's not quite the same thing that the OP is asking about. This is a limit that is caused by the flexibility of the chord sidewall, its not really a question of the weld strength, but rather of the connection as a whole. Just because the branch member and weld can deliver a force to the chord, that doesn't mean the chord can handle it. Part of designing the connection is making sure that all members can develop the forces required, this may mean using members with thicker flanges/sidewall or adding stiffener plates or some other type of reinforcing (A chord with a sufficiently thick sidewall will result in the effective width of the weld being equal to the actual width of the weld).

For the example used in the OP of an HSS being welded all around to a base plate, a fillet weld can most definitely be provided that will develop the full capacity of the HSS.

Quote (dauwerda)

This is essentially the definition of developing the member capacity. The weld is sized such that the member will fail before the weld does. There is no reason to provide a larger weld as that is the size it takes to develop the member capacity.

Correct me if I am wrong, but essentially once fillet weld is added, a different failure mechanism occurs in the base metal. It is no longer governed by tensile strength, but by shear fracture, so the member will never develop an original capacity.

In the example I provided, fillet weld on 5/16 HSS will develop maximum tensile ϕRn =7.58 kip/in, regardless of how much larger than 5.44/16 you make it, while the actual HSS wall has a tensile resistance of 10.03 kip/in. And that is assuming the same resistance factor (ϕ=0.75). For bending ϕ=0.9, results in 12.0 kip/in capacity, while weld is still limited to 7.58 kip/in.

That example (and that section of the design guide) is for the situation of applying welds to the face of an HSS. Such as a shear tab welded to the side of an HSS column. It is comparing the weld strength to the shear yielding and shear rupture capacity of HSS sidewall. This does not apply to attaching a base plate to the end of the HSS as you are designing for tension in the HSS wall, not shear.

dauwerda, what you are saying makes sense. I might be influenced too much by RISAconnections and how they calculate things which is way too conservative in most cases. Base Material Proration probably should not be applied to HSS-thick base plate welds, only to HSS-HSS ones.