Air Asperation In Argon System? 1

[MetalsResearcher]

My company manufactures automated welding systems. In our TIG system we preheat the filler wire with an electric current, the wire feed tube uses Argon to prevent the filler wire from oxidizing prior to entering the weld pool. Even though the feed tube is slightly above atmospheric pressure it seems that any leaks in the Argon system allow air (with Oxygen) to get into the shielding gas and contaminate the weld with porosity. Does anyone know why even at positive pressure this aspiration occurs? I am looking for a source or document that explains this phenomenon, something written that I can refer to. I need to understand this air ingress at positive pressure. Thank you in advance, Dale dale.flood@yahoo.com

 

[Compositepro]

It is far more likely that the gas lens is not perfectly shielding your welding operation. Welder technique and drafts can easily cause this. How are you preventing air from entering your feed tube with the wire? You may need more than "a slight positive pressure" to flush the boundary layer of air off the wire.

What you suggest is theoretically possible with rubber or plastic hoses. Oxygen will diffuse though the walls at a higher rate than argon.

 

[MetalsResearcher]

Thanks Compositepro!
But my question was more general than the solution you are trying to provide. We already solved the porosity issues by sealing every leak in the argon systems. They were detected using liquid leak detector. There are two, one for the torch and one for the hotwire heating chamber/tube.

But what I don't understand is if there is pressure, and thus flow, in a gas (argon) line and some gas is leaking out, how would that cause air to get in the line that causes the porosity. It would seem that the internal pressure would always cause the gas to leak out, not in. And more, is there any literature about why a leak out is a potential spot for air to leak in...

 

[btrueblood]

Are you sure the porosity you observe is due to oxygen contamination? There are many other sources for contamination that can cause porosity.

 

[MetalsResearcher]

Hi btrueblood, I'm sure we had porosity. A lot of very fine pores scattered throughout the weld. They were so fine they passed radiography and sheer wave ultrasound but when sectioned and polished they were visible. I'm also sure that the only thing we did to get rid of it was "snoop" the fittings and eliminate all leaks. But why would air get in a line that has some amount of pressure? Is there literature that discusses this phenomena?

 

[Compositepro]

I think the concept you are looking for is diffusion. When gas molecules move through porous materials where the pore size is smaller than the mean-free-path of the molecule the flow is not controlled by pressure but concentration (of course pressure affects concentration). But diffusion will occur against pressure gradients.

 

[btrueblood]

Metals,

My point was - weld porosity is more commonly attributed to dirty parts, or contaminants in the base metal itself, and less commonly to O2 (or water vapor, or hydrogen for some metals) in the shield gas. Oxygen in the cover gas should show up as metal oxide (dross) or slag in the weld, not so much porosity in the form of bubbles. I think nitrogen would tend to make more bubbles than oxygen, if atmospheric contamination is really the problem. A fairly common weld mix for carbon steel is Ar + 5% O2, and even stainless steels are welded using Ar + 1 to 2% O2. Weld gas typically contains a small percentage of oxygen to improve heat transfer in the weld pool, and well, because it's hard to distill out ALL of the O2 in a liquid air plant.

How are the parts cleaned/treated prior to welding?

What metal(s) is the porosity occurring with?

Does the problem occur with other shield gases and metals? Have you tried sourcing/using purer Argon (i.e. technical grade vs. welding/industrial grade).

Is the backside of the joint (inside of the tube and fitting) being purged with argon also?

 

[MetalsResearcher]

Hi btrueblood, Thanks for your efforts again. The current metal that was exhibiting the porosity was A53 and A106 Gr. B. It was ground with abrasive wheels to remove all surface impurities and then cleaned with acetone just prior to welding. We have QC inspections on cleanliness. For GTAW we almost always us "weld grade argon" for carbon, stainless & nickel alloy steels and yes, when appropriate we purge the backside. These welded components are for nuclear powerhouses.

The problems isn't getting rid of the porosity, which passed radiography and sheer wave ultrasound but was found when we sectioned and polished samples. The only thing we did to get rid of it was "snoop" the fittings and eliminate all leaks. But why would air (or as you suggest, "nitrogen") get in a line that has some amount of pressure? Is there literature that discusses this phenomena?

 

[btrueblood]

Well, it could diffuse back into the line, i.e the osmotic pressure of the nitrogen (trying to get into the zero-N2 space in the argon line) might be higher than the differential going the other way (line pressure). What Compositepro has been talking about.

I'd discount this as being a significant source of weld contamination until I had absolutely ruled out every other possible source of contamination.

Are you sure you are getting air into the shield gas upstream of the weld puddle? Can you capture a sample of the shield gases emanating from the torch head, and have the sample analyzed for these contaminants (both to see if any are there, and to find out what they are). A pretty simple test - you get an evacuated test cylinder from a lab, poke the nozzle into the torch head, run the wire feed and gas purge for a few seconds, open the sample cylinder valve, close it, send the bottle to the lab. Oh, and take a sample straight from the gas supply header (K-bottle) too. It might make sense to check a weld head that has been sitting idle for a long period of time (to allow a maximum amount of diffusion to occur) versus a weld head that has had a fairly continuous stream of fresh gas going through it.

I'd be looking for other sources of contamination first. A53 is just plain old steel pipe - the spec. has been around for probably a century or so. What level of nitrogen is allowed in the base metal? What level of other possible porosity-causing contaminants (low b.p. metals can cause weld porosity too, e.g. zinc). If it IS gas porosity, nitrogen is a more likely contaminant gas (it's 80% of the atmosphere after all). Nitrogen tends to not react with the metal, but diffuses into it, and then bubbles back out as the puddle cools and solidifies. Hydrogen (from hydrocarbons, i.e. wax, grease, oil, or from water (remember that rust contains a certain amount of water too)) also bubbles back out of the molten metal as it cools. Oxygen, on the other hand, reacts with the metal (especially hot metal) to form oxides, so there isn't much free oxygen left in the molten metal to bubble back out. Nitrogen can also contaminate your weld from the backside, if you aren't purging the i.d. of the tube/fitting. Or, the purge gas may not be applied long enough, or the flow rate may be too low. Welding in a windy environment may disturb the purge gas "shield".

 

[mbt22]

How is the tube held at postive pressure? Could you have created a venturi, sucking air in through a small leak where the diameter is reduced by a fitting?

Matt

 

[Duwe6]

Another probable cause is overpressuring/excessive flow of the torch gas. Even the "Gas Lens" type Tig torches will go into turbulent flow with excessive flow.

If you are not using a "Gas Lens" diffuser, immediatly change. The old-style, cheap torch diffusers are VERY easy to put into turbulent flow and entrain air. The quality of your cover gas purge is evidenced by the color of the finished weld. Get a good welding inspector to evaluate your as-welded colors. If the color is good, it is the either the filler or the base metal, like btrueblood said.

N2 in your hose is a very, very probability after the argon has been run long enough to purge out the lines - 2 or less volume changes.

 

[MetalsResearcher]

Thanks again btrueblood, mbt22 and Duwe6!

All of you are providing great tips on how to stop the porosity but as I already stated twice, "The problems isn't getting rid of the porosity. The only thing we did to get rid of it was "snoop" and tighten the fittings and eliminate all leaks. Since we are using large diameter gas lenses on each machine there would be some amount of back pressure albeit low. So why would air (or nitrogen) get in a line that has some amount of pressure?

Is there literature that discusses this phenomena?"

 

[MikeHalloran]

I fail to understand how the pipes could aspirate air, but the gas lenses could, especially if a nearby pipe leak disturbs the shape of the gas lens.

I'd try to visualize what's going on in and around the lens by probing it with a gentle flow of cigar smoke. Good cigars, please.



Mike Halloran
Pembroke Pines, FL, USA

 

[btrueblood]

Metals,

Please re-read 2nd paragraph of my last post, you haven't answered my questions, particularly "Are you sure you are getting air into the shield gas upstream of the weld puddle?". If you have confirmed this, please elaborate; if not, I gave you some suggestions for testing to locate the source of contamination. Mike gave another.

Regarding gaseous diffusion, you could make an order of magnitude calculation to determine a reverse flow rate for various gases, provided you can find a permeability rating for the material and gas species of interest. An example: I have a 25' length of 1/4" o.d. nylon tubing, filled and sealed with pure argon, and sitting in one standard atmosphere. Matweb.com lists an oxygen permeability, P, for this material of 3.5 cc*mm/[m^2-24hr-atm]. Calculate the volume of oxygen transmitted through the tube wall in 24 hours as:

V = (3.5cc*mm/[m^2-24hr-atm])(.15 m^2)/(1.52mm)
= .35 cc per 24hr per atm.

Assuming the partial pressure of oxygen in the atmosphere is about 0.2, and the partial pressure of oxygen in the tubing is zero, and (here's the big question mark) that the differential pressure of argon and its permeation in the opposing direction does not "sweep" out any oxygen trying to penetrate the walls, the oxygen might reach 0.1% by volume in the tube, after sitting for 24 hours. I calculated the tube's volume at 60 cc, so the concentration of O2 in 24hrs would be:

concentration = (.35 cc/24 hr-atm) * (0.2 atm) / (60 cc) = 0.1%

It'd be an easy test to confirm, hook up a tube to an evacuated sample cylinder, fill tube with argon and cap it off. Wait 24 hrs, and open sample cylinder valve, then close it, and send it off to the lab for analysis. I'd bet you a beer that the test sample would be an order of magnitude, if not several, lower in O2 concentration than this calculation predicts.

So, next: is your typical purge gas flow rate more than 1 cc/sec, and do you not typically pre-purge before welding for at least 60 sec? Doing so would flush any contaminated gas out of the tube...eliminating the contamination. During a flowing purge of 1 cc/sec, the calculated influx of O2 through the wall, if correct (big caveat, it's probably less, see above), would give an O2 concentration in the gas of ~0.8 ppm, probably less than the O2 concentration in "pure" lab-grade argon in the first place.

So, what testing HAVE you done?

 

[Duwe6]

Based on btrueblood's cogent analysis, this excludes hose permeability, and leaves;

Excessive flowrate for the diffuser, causing flow sufficiently turbulent to draw in air.

Failure to, or inadequate, pre-flow immediatly prior to striking an arc, to purge out the O2 that had diffused into the hose from the "gun's" open end.

Contaminated argon supply, or header. Dewpoint needs to be -40°F or lower; no oil in the lines. Try using liquid argon, by defiunition it is dry. O2 content lower than 1.5% will not be noticed during welding, so it is not Oxy contamination, unless the ammount is HUGE, like 10% or so.

Contaminated basemetal. Usnig a clean, white cotton rag or gauze pad moisened with a cleandrying solvent [acetone, MEK, lab grade xylene or naptha, etc] wipe all weld-prep edges. You are looking for a thin film of oil.

Contaminated weldwire. Buy a small spool of a premium brand, like ESAB or Lincoln, and run it. Use a 'wipe' on the wire where it enters the thimble on the wiredrive, just after the drive rollers. A foam ear-plug with a needle-hole in it will wipe well for a half-day or so.