By 803056 
Date 09-08-2007 13:04
Edited 09-08-2007 13:36
I'm not an overly scientific person, but I do like a little "show and tell" demonstration once in a while.
Take a sheet of aluminum that has a shine to it, not heavily oxidized or coated with hydroxides, and wipe one area with acetone and one area with alcohol. Which seems to be clouded and which still has a shine? I do agree that the acetone is better at removing grease and oil, that's why I use acetone first before removing the oxides or preparing the edge for welding. The alcohol is use just before welding to remove any residues like fingerprints, lint, contamination from shop air, etc.
You are correct in saying that most metals begin to oxidize when they come into contact with air (oxygen being the element of concern). However, is it common practice to clean steel with iron oxide? No, of course not for several reasons, one of which is the iron oxide is the contaminant we are attempting to remove. Other reasons include: it breaks down and becomes dusty easily and it isn't as hard as steel.
The advantages of aluminum oxide are that it is hard, doesn't become dusty (break down into smaller particles), it has a high melting point (this is not always an advantage, oops, wrong column), has a lower density than iron and most of its alloying constituents. It works well as an abrasive for most metals.
However, in the case of aluminum, the density of the aluminum oxide abrasive is nearly the same as for the base metal so it doe not have a tendency to float out of the weld, thus it causes the weld to be brittle. It has a melting temperature that is nearly three times higher than the base meal (1200 versus 3200 degrees F), so there is little tendency to melt or to disassociate in the arc.
How much aluminum oxide can be present? How good do you need your weld to be? The more oxide present, the more severe the embrittlement problem becomes. Excessive amounts of oxide tend to reduce the wetting action of the molten aluminum.
Try to weld aluminum that is heavily oxidized. The aluminum base metal will melt, but the heavy oxide layer will prevent the metal from flowing. Instead the oxide acts like a shell mold and holds the original shape until it is heated to the point it melts and the aluminum starts to flow, by then it is superheated to the point other bad things begin to happen, such as those materials with low vaporization temperatures may begin to volatize (magnesium comes to mind).
As you said, all metals have some oxide present, but in the case of most metals we can add deoxidizers to the filler metals. There are a limited number of deoxidizers used with aluminum and they tend to be very chemically active. So we are left with the option of cleaning the surface of the aluminum immediately before welding to reduce the oxide layer to a minimal amount. We also use AC to reap the benefit of cathodic cleaning (which reduces the efficient utilization of the arc energy) and we live with the reduced penetration and lower travel speeds.
As mentioned in another post, you can scrape the surface of the aluminum just before welding to reduce the oxide layer as well as brushing it or draw filing it. Whatever method is selected to remove the oxide, we don't want to add more oxide to the surface nor do we want to imbed the aluminum oxide into the surface of the metal. Notice in my previous post, I said to wire brush the aluminum using a unidirectional stroke. You do not want to use a scrubbing (back and forth) motion because the moment you reverse the stroke you tend to push the oxide into the surface of the metal.
Many of my suggestions are "lessons leaned" by qualifying many (many, many) aluminum procedures.
Maybe that is why I believe that the contractor needs to go through the procedure qualification process to truly appreciate the complications involved in obtaining good welds. Do you know how many times I've heard, "We never had a problem until you made us qualify the procedure"? What they fail to realize is the problem was always there, but they never knew it!
Best regards - Al
