Shredded tungstens welding Aluminum with mixed AC/DC

[Gavin Melville]

I've been using mixed arc for welding Aluminum for a while, and weld wise all is good. I run 200 ish amps DC, 60% of the time, AC at 60% of that current. The AC is 250Hz, square wave, 62.5% EN. I am welding thin to thick, say 1" to 1/8", and get deep penetration and narrow welds.

The problem is tungsten life. The ends of the tungstens are shredded, like when you get a split, except I get 20 splits. Tried all kinds, zirconium doped, lanthanated, even thorium doped. The best results are with rare earth (E3) in 1/8", one size up from what I usually use. They still shred, but last longer. Prep is a short cone, no ball as per Kemppi's recommendation. I end up regrinding every 10 minutes.

Is anyone else running mixed AC/DC on Aluminum?

Ps: in case you think you are seeing double I also posted this to welding tips and tricks.

 

[hotrodder]

I've not had a chance to have a proper play with kit thas offers mixed TIG but my thinking is that the problems are due to thermal shock? 62.5% EN is putting a lot of heat into the tungsten (with a decent squarewave inverter) but for extremely brief periods so the tip doesn't have enough time to melt a little and stabilise? Personally i'd experiment with more EN% and tweak other settings to get/keep enough cathodic cleaning
Other things maybe worth trying are grinding to a longer taper but not sharp, leave a 1mm flat at the tip. That can encourage the tungsten to round off a nats instead of splitting. Or briefly round off the tip a little using regular AC etc

 

[Gavin Melville]

When running mixed AC/DC the AC is only there 40% of the time, so the effect is less than normal AC. The DC creates a narrow arc, heats the aluminium for 0.6 of a second, then changes to AC. The cleaning action is good, about what I expect. Even if I start with a balled tungsten, the ball gets blown off. I'll try and get some photos later today.

 

[hotrodder]

I understand how it works, what i was trying to say is that at 62.5% EN the tungsten is given a massive thermal shock every time the arc swtiches to EP. More EN will reduce that but, not having had experimented with mixed TIG settings, might not be able to able to increase EN much before weldpool cleanliness suffers. Tweaks to other settings including AC frequency might fix that along with the splitting issue for example if (for whatever reason) i bias AC balance as far towards EN as the weldpool will tolerate i can't bias it as far to EN at higher AC frequencies as i can at lower Hz etc

 

[Richard.]

I've got this function on my unit and have never used it tbh. I understand the principle behind what it does but what are the advantages of mixing AC/DC.

 

[Gavin Melville]

Advantages, according to Kemppi: Deep penetration, less heat into the work piece, welding thick to thin material. http://www.mig-welding.co.uk/forum/threads/mls-2300-ac-dc-mix-mode.56543/#post-679289
To a significant extent that is what happens. I have played with most variables, and the defaults are about right, although I prefer running a higher AC frequency.

This is a not great photo of what some tungstens look like. All after - say - 30 minutes use.

at 62.5% EN the tungsten is given a massive thermal shock every time the arc swtiches to EP

I'm not convinced there is a great thermal shock. The waveform looks like this;

The EP pulse time is short enough that the electrode is only positive for 16ish percent (40% of 40% in round figures). At 250Hz the thermal inertia of the electrode shouldn't respond instantly. I have played with the AC duty cycle, unfortunately the MLS 2300 only allows from 40% to 75%.

I'm going to try and get some 2% lanthanated tomorrow and see if this helps.

 

[Tigman]

you said in your post advantages are less heat input into the weld ,therefore the heat is going to the tungsten due to the reverse polarity , is that right ? lol tell me if im chatting S@@@

 

[Shox Dr]

I don't either Richard. Takes too long.
However I think macros are way to go with the common jobs, and great joker card for those once in a bluemoon jobs, a pig to do but worth it.
A saved marco, would be the mutts.

 

[Richard.]

I've actually never used any of the gimmicks on that panel.
Micro tack, mixed, job storage etc I've just altered the standard settings. I did tap the synergic pulse button once. And it provides a usable DC pulse setting.
AC pulse I've just never got on with.

 

[Chris Stevens]

Try reducing frequency, and playing with proportion of ac to dc parts of waveform. Mixed does seem to work quite well in providing increased pen for some jobs, and I would guess someone who has used it any amount would be able to provide useful info on best way to set up. Mine screws tungstens quite quick in mixed, and wonder if you have air cooled torch rather than water, as this might make problem worse?

 

[hotrodder]

This is a not great photo of what some tungstens look like. All after - say - 30 minutes use.
I'm not convinced there is a great thermal shock.

Personally after seeing those pics i'm more convinced it's thermal shock that's killing 'em. Obviously it doesn't get anywhere near that severe with pulsed DCEN but aggressive pulse parameters in DC can also shorten tungsten life/number of restrikes you get before the tip needs redressing due to 'frosting' (manky grey deposits/build up at the tip)

I'd expect turning the AC frequency down to result in the tungsten running hotter and maybe help it melt/round off a little and as result stabilise
Obviously haven't seen the etched zone on a weld but if the cleaning action at the current settings is good then increasing the EN% might help in the other direction i.e. keep the tungsten sharper, stop the 'start to melt - freeze - start to melt- freeze' then split cycle.
In either case changing the prep might help... instead of prepping 'by the book' for AC grind it somewhere between an AC and DCEN prep; a longer taper (2 or 3D) but with a larger flat at the tip than for DCEN to try and encourage it to round off a little instead of blowing the tip into the weld and then splitting and showering the weldpool with more microscopic pieces of tungsten
Dunno if you snap the end off or cut the tungsten but IME snapping 'em = more likely to split again/split more quickly. Especially once they've been properly overloaded and gotten properly manky

 

[Richard.]

I can never get away with snapping them they always split on me if I need to I cut them.
The thermal shock makes sense as the inconsistent change between no loading then loading has got to give that tungsten a jolt . I find that if I pedal pulse On ac I can get alsorts of wacky shapes come up on the tungsten and I believe alougth it's a different kettle of fish the reasoning is the same. It gets whacked with a load of current then Backed off it doesn't get time for the ball to stabilise. This is the same thing but instead of current it's polarity. It gets whacked with ep loading on ac then flicked to dcen only and the ball doesn't get time to stabilise. I'd certainly be trying a narrower prep as hr suggests. Perhaps instead of a flat try 2 angles. I find it's simply brilliant for encouraging a ball. Narrow prep like a dc prep with a small wider prep at the tip. Almost flat but not quite.

 

[Gavin Melville]

The tungsten's are cut by grinding, and I've tried long, short, flat tipped, balled (DC+ on copper)....

Just tried lanthanated, both 2.4mm and 3.2mm, 2% doped. Results basically the same. Overall the best has been E3 with a small flat tip.

I'll try a lower AC frequency, and a software waveform shortly.

 

[Shox Dr]

Have you thought of calling Kemppi?

 

[Gavin Melville]

Actually I hadn't. The last question I asked kemppi was about why the manual was so content free. The answer was don't go there.

I also asked what setup function A23 did - not in the user manual. The kemppi rep who was standing in front of me at a trade show called the dealer 24 hour tech support number. They didn't know either.

In reality I'd rather have a great welder than a great manual, but seriously guys......

It's worth a try

 

[Chris Stevens]

I would guess experimenting with balance and frequency may well make a difference?

3. MIX TIG (mixed current AC/DC-)
With mixed current frequency and balance of alternating current are determined by
adjustments made for AC-current. Adjust by choosing sections 4. – 5. if needed.
The following QUICK SETUP parameters can be adjusted:
1. AC-time (AC 10 ... 90 %, factory setting 50 %)
2. Cycle time (CYc 0.1 ... 1.0 s, factory setting 0.6 s)
3. DC-current (DC(-) 50 ... 150 %, factory setting 100 %)
4. Balance (bAL -50 … 0 … +10 %, factory setting –25 %)
5. Frequency (FrE 50 … 250 Hz, factory setting 60 Hz)
6. Selection of AC waveform, sine or square wave (SinuS/SquArE)
7. Hot Start time for the 2T function (H2t 0.1 s…5.0 s, factory setting 1.0 s). Can be seen only on
the 2T function.
Factory setting is marked with a dot after the numeric value.
Growing DC-current increases penetration but decreases cleaning effect.