The fact that the surface becomes black sounds to me like you are getting some contamination into your gas. There are so many possible origins for the contamination that I am not going to even try to mention them all. (Not only will I forget half, but it will take me all night.)
The fact that the material just suddenly "falls through" is a typical problem when welding light gauge Al. Al alloys are typically designed to have a relatively small melting range. (As this decreases the cracking tendency.) This means that when it starts to melt, it very suddenly becomes fluid. In addition, Al has a very high thermal conductivity. This means that you need to add a great deal of energy befor the melting starts. The third complicating effect is the oxide layer. This oxide layer melts at an extremely high temperature. Much higher than the melting temperatue of the Al. This means that if you should have excessive oxide on the surface, you will not see anything happening untill all the base metal beneath the oxide is moulten, at which stage the "falling through" bit happens. When one combines these three factors, you end up with a material that is extremely difficult to control during welding.
You state that you are cleaning the Al, and also using HF AC current to weld it. This should help with the oxide removal, but again, there are many possible problems that can creep in here.
Personally I believe that you are being somewhat optimistic in believing that you can weld such a thin Al. Possibly somebody with a great deal of experience welding Al can do this using a pulsing power source. Even then, s/he will have trouble.
Regards
Niekie Jooste
Forum
