Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic Welding Industry / Technical Discussions / Why does Pulse GTAW Welding work?
- - By LarryL (**) Date 11-24-2003 15:46
Recently I purchased a refurbished Thermal Arc Pro-Wave 300GTSW. Since this is a huge step up from my old Lincoln IdealArc 250, I have been practicising with my new welder and learning how to use its modern features. Even though I understand its basic principle, I am still puzzled by how pulsing the arc at high frequency works in welding thin material, such as steel. For eample, I've been able to weld 0.035" thick sheet to 0.125" thick sheet by setting the pulse frequency to 200 hz at around 120 amps EN. This does a good job with little tendency to burn through. I admit that I also ease up on the foot control a little when the arc sweeps down to the thin material. My question is: since pulsing at this very high frequency intuitively seems to be equivalent to just using a lower amperage setting, like 75 to 80 amps, without pulsing, why doesn't using the lower amp setting without pulsing work just as well? The lower amp settings without pulsing don't work as well for me as I get more burn through in the thin sheet with them. At 200 hz, however, the arc is pulsing between high and low amps every 5 milliseconds. One would think that the puddle doesn't have time to cool down on the thin sheet between these very brief pulses from high to low current. Yet pulsing seems to work, but, why?

My Thermal Arc welder doesn't have the Dynasty's ability to adjust background current and the peak current's percentage of total cycle. Therefore I don't know what the low current end of my pulse cycles consist of. They must just be a fixed percentage of the high end in my welder.
Parent - By dseman (*) Date 11-24-2003 17:08
Larry,
Excellent question and a very astute observation. I'm no expert in this matter, but I'll make a stab and hopefully others will join in. I pulled up the operating manual for the T/A 300 and found that in pulse mode, the peak time is 35% of the cycle, leaving 65% for the background. The background level is fixed at 20% of the peak. So with your 120amp peak setting at 200hz you end up with the following:
Avg. amperage: 120*0.35 + (120/5)*0.65 => 57.6avg. amps

What is not taken into consideration with the assumption that this pulsed waveform is the same as an avg. 57.6amps DC? The answer is TIME.
The heat into the weldment is a function of both time and amperage. It takes a certain amount of time for the heat to propogate through the metal and raise its temperature. The final temperature attained is a function of the wattage input, time, and thermal conductivity of the metal which dissipates the heat away from the area of interest. Since the 120amp peak pulse is only on for (1/200)*0.35 = 1.75msec, this appears to be long enough to melt a small puddle in your joint, but then immediately the amperage is lowered to 24amps where the pool quickly freezes and thus stops any further penetration. As you stated, if you tried to weld this at 70-80amps there would more likely be melt-through. My guess is that even at 60amps you would be more likely to have problems.

Good question. I've mentioned to Miller and Lincoln that they should publish a more definitive application note or brochure on the subject but they both indicate that pulsed welding is 'highly subjective'. My counter-point would be that if they explained what I have attempted to
(that is if I'm correct!) a lot fewer people would be frustrated and it would reduce the learning curve considerably.

-dseman
Parent - - By Lawrence (*****) Date 11-24-2003 18:43
Last Tuesday I was at the "Fabtech" Exposition at McCormac place in Chicago. Nice show, good exhibits.

Miller Electric had a number of demos set up and an army of staff down from Appleton. I asked your exact question because I don't have a high frequency pulse tig powersource at the moment and was wondering.......

#1 if the performance was worth the investment.

#2 if there is some doccumented theory with practical examples, including working parameters to back up the claims of superior performance on the thin stuff.

Its no surprise they said that its worth the money, however nobody I spoke to could render an explaination of the functions/theory behind the whistles and bells.

To be fair my experience with the folks at OTC was just the same.

Larry, What does your owners manual say?

I sure hope somebody will publish soon on this topic.
Parent - - By dseman (*) Date 11-24-2003 19:21
Lawrence,
Are you the same person who works in the airline industry and has previously written some great posts on the aerowave and dynasty machines? I thought if anyone would have a good handle on parameter optimization it would be you! I was hoping that you might be able to dispense some golden nuggets of wisdom on this topic. ;)

-dseman
Parent - - By Lawrence (*****) Date 11-24-2003 20:19
Dse

I'm that guy. However, I'm out of Aerospace and into a Technical college (and loving it).

I have never had a powersource that was equipped with a DC pulsation frequency above 20 Hz. I can't speak to the application because I just haven't done it yet.

I like what you said, and I think the arc physics are not all that different in some ways to the pulsed mig and its ability to weld solid wire in spray transfer mode vertical up. (however, no adaptive feedback in the tig).

<I think> I have my brain around the idea, but I really like to get feedback from others on stuff like this. It really makes it easier to pass along this stuff to students when you can get the foundations of a theory comming from several different viewpoints.
Parent - By dseman (*) Date 11-24-2003 20:58
Lawrence,
Hmm. Well I know the dynasty has dc pulsing out to wazoo (wazoo is around 500hz), so I guess I thought you had as much experience on the dc side as you did on the ac side, based on your fine ac dissertations. I'm not an AWS member, so I have limited search capability of their journal articles. Maybe this will sufficiently motivate me to become one though. What I have found is the following article on the use of pulsed dc to increase penetration profiles:http://www.aws.org/wj/supplement/Ko02-01.pdf
It describes the use of higher amperage (>200amps) with sufficiently long, peak pulse durations (there is that TIME variable again) in order to increase the penetrating arc force beyond that of the avg. current values. It also appears from the original post that very small peak pulse durations are useful in producing a limitied amount of penetration. As far as textbooks go, the only one I've found whose title sounds applicable is "Pulsed Arc Welding" by J.A.Street of Woodhead Publishing ($110 US funds). I think I'll try to find that at a library. Not really too interested in purchasing blindly at that price. Let us know if you have found any other resources.
-dseman
Parent - - By LarryL (**) Date 11-25-2003 03:10
Lawrence, my Pro-Wave 300GTSW's manual only tells how to set the switch (to low or high frequency) and how to rotate the pulse frequency control to the desired value. The available frequencies range from 0.5 to 500 hz. The manual is disappointing in that it says nothing about when to use pulsing and provides no explanation of the advantages of pulse GTAW welding.

Actually, there seems to be very little information about pulse GTAW welding on the Internet. I found a little bit at Miller's site and Kempi's site. Overall, though, I haven't found any indepth treatise on pulse GTAW (there's a good deal about magnetic pulse welding, however).

I am surprised that you didn't experiment with pulsing when you had Dynastys available to you, Lawrence. The Dynasty 300DX, for example, has more adjustability of pulse values than my Pro-Wave welder. However, in my brief experience so far with pulsing, it does work well with thin material and minimizes burn through. In my first attempt at pulsing, though, I set the frequency at 20 and the pulsing at this range literally drove me crazy. Then I read the section on pulsing in the online manual for the Dynasty. It related that the frequency range from 10 to 50 is never used because the arc pulsing is disturbing to a welder.

I appreciate your explanation, Dseman. However, I find it difficult to comprehend how a molten puddle can react so quickly when a there is just a few milliseconds cycling from high current to background current. I'm convinced, though, that pulsing works when welding thin sheet.

I bought this refurbished inverter welder just to get the squarewave, balance control and variable frequency features, plus, of course, the 300 amp output (25% duty cycle). Getting pulsing was just a bonus that I found out about after I started using the welder.
Parent - - By dseman (*) Date 11-25-2003 03:52
Larry,
I too am amazed at what can happen in a few msec. But consider that for steel at a temperature of 2768 F, the metal is entirely molten. I don't know what the boiling temperature is so I won't go into what happens above 2768F. Your concern is how the pool can transition between molten and solid so quickly, right? Well, crystals of austenite first start to form at 2767F and continue until 2714F is reached. At this temperature the last drop of molten metal is now fully crystallized and stays in this form until 1560F is reached, at which time ferrite crystals seperate from the austenite. So, as far as pulsing and pool control goes, I think that a weld pool can transition from say 3000F to 2714F in a rather short period of time--thus exhibiting the quick weld pool 'freeze'.

Hope this helps with your understanding. I do rather enjoy these arc physics discussions!

-dseman
Parent - By LarryL (**) Date 11-25-2003 16:23
Dseman,
I'll gladly accept your explanation since I nearly flunked out of my metallurgy course while studying for my chemical engineering degree at Stanford, eons ago. The word around our chem. eng. circles was that metallurgy was one dull course that should be cut often. Poor Dr. Shepard (I still remember his name!) was the butt of pranks sometimes. One story was told how a student, during the course's final exam, rose to his feet shouting "Dr. Shepard, you've got me this time! He then threw his exam papers down and stormed out of the class. It later turned out that he wasn't even enrolled in the class. Be that as it may, I cut a lot of metallurgy classes. Now that I'm wiser and have an interest in welding and metalworking, I wish that I hadn't cut those classes!

I guess that the molten pool of metal can react very rapidly as you say. I myself am impressed that pulsing works in welding thin to thin or thick to thin. Yesterday, for example, I was working on a new welder cart (isn't that always the first project after buying a new welder?). At 120 amps EN and pulsing set at 200, I was amazed at how nicely and quickly I could lay down a bead between 0.125" thick square tubing and 0.035" thick rectangular tubing (scrap from Chinese pallet). Previously when I tried welding these two together with 75 amps EN without pulsing, I'd burn through the thin stuff at least once just about every time.

Lawrence, you need to try pulse GTAW on thin material for yourself and give us your opinion on it.

Parent - - By Niekie3 (***) Date 11-30-2003 12:57
Hi dseman

I too am sceptical of the explanation that the weld pool solidifies in such a short space of time. I say this because when welding with a low frequency pulse (0.5 - 1 Hz) one can clearly see the weld pool solidifying between pulses. This solidification taking place rather slowly. If we argue that at a low average amperage, the temperature will be very close to the solidification temperature, then the weld pool size will also be correspondingly smaller. If it is able to solidify and again melt within the space of a few micro seconds, then the weld pool will also be so small that it will not be practical to weld with. (How do you dip your filler into a weld pool that is 0.01mm in diameter?)

I do however not have a definitive answer, but will present another theory that somebody else will probably shoot down!

I believe that at these higher pulse frequencies, the welder merely has better control of the heat input and penetration characteristics than with an arc that does not pulse. In essence however, it is the same as welding with a lower, non-pulsing current. Let me explain where I am coming from:

There is a general misconception that when welding with the same heat input, the penetration into the base metal will be the same. This is incorrect. For example, welding with the following parameters will give the same heat input:

CASE 1: 100A, 10V, 20mm/min
CASE 2: 150A, 10V, 30mm/min

However, case 2 will give a substantially higher penetration into the base metal than for case 1. To understand this, we need to think about what happens to the heat that is introduced into the metal by the welding arc.

A great deal of the heat (energy) is conducted away from the area where the arc inpinges onto the plate. Some of the energy is used to melt the base metal, and some is used to melt the filler metal.

According to conventional wisdom, by reducing the amperage you can just make up for it by reducing the travel speed. However, when the amperage becomes low enough, no melting of the base metal will occur, even if you stand completely still. (Infinite heat input!) (Ever try to weld a thick piece of Copper with a low amperage TIG torch?) By increasing the amperage, the same heat input will be achieved by increasing the welding speed, but we must remember that the rate at which the energy is conducted away from the weld pool is very much a function of the weld pool temperature and thermal conductivity of the base material. As such, as the amperage is increased, and welding speed also increased the penetration into the base metal increases, becausue the thermal gradients in the base metal increases. This is a good situation when welding thick materials, but is a liability when welding thin materials. As such, when welding thin materials we need to reduce the amperage so that the percentage of energy that is "lost" into the base metal through thermal conductivity is increased, and the corresponding weld pool is much smaller. This reduction in amperage is however limited to the point where the arc stability is affected. Here is where the high frequency pulsing helps, because a stable arc can be maintained at a much lower average amperage setting.

I hope you could follow my reasoning here.

Regards
Niekie Jooste
Fabristruct Solutions
Parent - - By dseman (*) Date 11-30-2003 22:17
Nieke,
In your example of low freq. pulsing you noted that you saw the weld pool freezing. I guess what I should have stated in my earlier post is a better definition of a 'frozen pool'. Based on your example, I believe you were indicating 'frozen' to imply a state where a solid object could no longer penetrate, as in a filler rod or chipping hammer. My definition would be where 'sufficient grain growth occurs at a temperature, where the heat flow out due to thermal conductivity is greater than the heat flow in to support a molten state'. The temperatures I provided were given as a concrete example of how narrow that temperature band is for steel. The issue becomes how can the cooling rate occur so quickly.

I agree with your caseI, caseII examples describing the heat into a weldment where Hinput = ( I * E )/ V .
I = amperes, E=volts, V = travel speed
where Hcase1 = Hcase2, yet case 1 had lower penetration and a wider HAZ than case 2.

Linnert, from "Welding Metallurgy" has a similar example:

Case 3: 800amps,26volts,27.4 in/min produces Hinput = 45.5KJ/in
Case 4: 135amps,26volts,4.7 in/min produces Hinput = 44.6KJ/in

the kicker is that the weld nugget area was 0.185 sq.in. for Case 3
and only 0.044 sq.in. for Case 4. He indicates that a greater amount of heat was absorbed from the arc for Case 3,indicating a greater transfer efficiency over the thermal conductivity and cooling of the nearby metal. Hot and fast produces greater thermal gradients in temp over shorter distances away from the bead, and this implies less heat loss due to conduction.

I do disagree, (however quite friendly!!) , on your theory that an averaged pulsed waveform (consisting of high amperage, short duration pulses and a longer duration pulse of low amperage) has a Hnet equivalent to E * I from a non-pulsed waveform, and that the problem is mainly an issue of arc stabilty at lower amperages. My experience with inverters has indicated to me that they can be very stable at amperages as low as 5 amps. The good doctor's example had a high pulse of 125amps,a low pulse of 24amps, and I just can't believe this is really that much more stable than his averaged pulse-waveform value of 58 or so amps.

The high heat and very short duration of the pulsed waveform would seem to me to also have very high thermal gradients. Since the arc is only on for a brief period of time it should be very concentrated and have a narrower distribution of peak temps then say a constant DC arc. Steep temp distributions would tend to indicate a rapid cooling of the pool. If travel continues at too slow of a pace however, I can see the HAZ increasing as you continue to pour heat into relatively the same physical spot. To be sure, if you want to reduce penetration, you will want to reduce the heat transferred immediately after pool establishment and prior to it sinking downward and creating a nugget. The fastest way to do that is to cut the amperage low enough so thermal conductivity is greater than the heat input. I don't quite have a good mathematical expression to show for how pulsing differs from the continuous amperage cases that you and I both have found. I need to get a better understanding of how the short term , almost instantaneous, heat transfer differs from the long term found with a constant amperage arc. If and when I find out more, I will post.
I think I'm off to the library.....
-dseman
Parent - By Niekie3 (***) Date 12-01-2003 20:38
Hi dseman

Let us know what you find in the library. It should be quite interesting.

Regards
Niekie Jooste
Parent - - By Lawrence (*****) Date 11-25-2003 13:24
My Dynasty was a quazi Beta model and had no high frequency pulsation. It had knobs instead of the euro-style touchpads.

Some folks find switches and knobs intimidating but love touchpads. I'm just the opposit.

:)
Parent - By awill4wd (**) Date 11-26-2003 11:24
Lawrence, I have to agree with you wholeheartedly about knobs versus Led's/touchpads.
My OTC has LOTS of knobs and buttons but after learning just what each knob/dial does I find it's second nature to me now to just quickly glance at the machine faceplate and if I feel I'd like a change in pulse/AC balance/frequency etc etc it's really simple just to reach over turn a knob/dial and continue welding rather than have to access a menu then change the parameters via an led screen.
To others it looks pretty intimidating but OTC really got it right in not following every other manufacturer and changing to an led/menu system.
Regards Andrew.
Up Topic Welding Industry / Technical Discussions / Why does Pulse GTAW Welding work?

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill