 
 
		
		
				
				
		
			
				
					I find myself mixing in some Helium for when I weld aluminum about 1/8" or thicker, especially if extra penetration is desired.  I don't do it for stainless or mild steel.  But on aluminum, the He % allows concentrating the heat more, and I find I can often get away with 1/16" tungsten (or a "sharper" tip grind), where with argon I'd have needed a 3/32" tungsten (and/or a "blunter" tip grind.)  This makes a deeper, narrower puddle which is often a good thing.  There can be less heat input into the work.  Being able to penetrate deeper saves me prep work (don't need to bevel as deeply to get the same penetration.)  And as a result, I find I can get my aluminum jobs done more quickly.  What would have normally taken me 45 minutes, I am getting done 10 minutes quicker.
So really, I use some Helium % more often than not on any aluminum work.
Since I often run a light arc over the surface before I start to weld, to check the cleanliness of the surface (especially on repair jobs where things aren't clean to start with, and dirtiness will be detected and can corrected before porosity is introduced), I like to use a small enough He % that it doesn't adversely affect arc starting and stability under the light arc needed for cathodic etch cleaning, but yet still give me the heat boost I want when it come time to weld, and penetrate.  So I put it at a small He% flow, so I don't have to vary the gas mix after cathodic etch cleaning is done, before I go for the deep puddle melting action.
I often find myself running about the minimum He flow I can get away with while having the mix remain stable.  It is not totally scientific, and I am still experimenting, but would say, maybe it's somewhere in the ballpart of 2-5cfh Helium.  Compared to maybe 10 cfh argon if using a #5 cup, to 12-15 if using a #6 or with more stick-out.
If the He % varies after the solenoid opens (I can hear it in the sound it makes coming out of the cup), I can abort the arc start and keep the gas flowing long enough (say, 3-4 seconds) to let the mix stabilize before I triggering the arc start.  If the mix is varying and I start the arc, it can make hard arc starts and unexpected heat input until it stabilized.
Best of luck...
				
			 
			
		 
			
				
			
			
				'13 Everlast 255EXT
'07 Everlast Super200P