[tshaw]

Gentlemen,

I am doing an automated (circunferential) TIG weld on Titanium alloy. The weld is done inside a chamber that is filled with Argon(30%) and Helyum(70%) so that the shielding is superb (<10 ppm O2) . Wall thickness is 2mm and the joint edges are machined in such a way that they provide some material for addition during the weld. Doing the weld with 65A and 1.2 mm/second from the start to end (including some overlap, there is a ramp down at the end of the weld) ) gives a satisfactory penetration, except for about 25mm following the start of the weld. Please have a look at the attached pictures with test specimens.
The behavior is that when the arc is opened the part rotates 25mm until full penetration starts then penetration is very good for most of the part until the part rotates 360 degrees and the arc coincides with the start of the weld. From then on (during the overlap) the 65A is not able to melt the material that has not been melted at the very start of the weld (those 25 mm already mentioned).
The equipment allows me to increase the current at any angular position I want and I have tried to do so at the end of the weld and this solved the penetration issue but I end up with a rather big weld width, what is not acceptable in this case. Other possibility would be to start the weld with an increased current and then decrease it to the 65A and maybe increase again at the end of the weld. I have not tried this.

Would anybody that experienced this problem has and advice?

I have just one side access, so I need full penetration welding from one side only.

Thanks in advance

Thomas Shaw

[weldin mike 27]

I know almost nothing about titanium and automation, but I think that if you experiment with different angles on your tungsten electrode and different arc gaps, you may achieve satisfactory results. Any particular reason for the argon /helium mix? Straight argon is more common.

[cj737]

In lieu of increasing the amperage at the start, I would suspend adding filler for the initial 25mm rotation. Then, when you come to the "overlap" the 65 amps in only heating the base metal for filler addition, not the base plus the initial weld with filler. Essentially, you are reducing the mass with this approach.

I don't have any automation experience either, but I have used a similar approach on tubing with a positioner.

[LtBadd]

Thomas, welcome to the forum
So either increasing the amps at the start, and/or start the arc, and wait a short period of time before you begin the rotation.
Either way you need to get penetration before you start moving.

We used to use high purity argon in the chamber, and then mix the helium in the torch shielding gas

Post back with your progress on this, very interested to follow

[tshaw]

I know almost nothing about titanium and automation, but I think that if you experiment with different angles on your tungsten electrode and different arc gaps, you may achieve satisfactory results. Any particular reason for the argon /helium mix? Straight argon is more common.

Hi Mike,
Thanks for your answer. We already tried different gaps and angles but there was no improvement from the gap and angle we are using. We are using argon/helium (also known as "Navy") becasue it gives better penetration (for the same current).

Thomas

[weldin mike 27]

No problem mate, best of luck

[tshaw]

In lieu of increasing the amperage at the start, I would suspend adding filler for the initial 25mm rotation. Then, when you come to the "overlap" the 65 amps in only heating the base metal for filler addition, not the base plus the initial weld with filler. Essentially, you are reducing the mass with this approach.

I don't have any automation experience either, but I have used a similar approach on tubing with a positioner.

Hi CJ, Thanks for your response, If I understood well, you are suggesting to grind off the addition material (remember that the additional material is machined at the edge of the weld joint, we don't add material as usual, with a rod) of the initial 25 mm of the weld and then add material during the overlap. I am afraid that I can not even try this because we are welding inside a chamber and it will be very clumsy (or even impossible) to add material with a rod using the chamber gloves during the automatic process.
Have a look at the pictures
Thomas

[tshaw]

Thomas, welcome to the forum
So either increasing the amps at the start, and/or start the arc, and wait a short period of time before you begin the rotation.
Either way you need to get penetration before you start moving.

We used to use high purity argon in the chamber, and then mix the helium in the torch shielding gas

Post back with your progress on this, very interested to follow

Hello Richard,

We have already proceeded like you said, however with pure argon and no helium on the torch. After some adjustment of the initial current and the time stopped it does get good penetration but the width of the weld at this position gets too big and is not acceptable according to standard I have to follow (it is an european aerospace standard - ECSS). See the pictures of this weld below.

Then we tried welding only with helium inside the chamber and penetration is the best (same penetration with half the current compared to pure argon) but the arc became very unstable. During the first welds it was ok, the arc opened right at the weld position, as expected, but afterwards, instead of opening at the welding point the arc started jumping everywhere inside the chamber between the cabling, torch holder (positive) and the chamber (ground). See the pictures.

Next, still with pure helium inside the chamber, we tried to open the arc flowing argon through the torch and after the arc opened we closed the argon did the rest of the weld in pure helium, what did work well but we are afraid that during the process it may become unstable again and damage the product (big big loss $$$$).

Your idea of filling the with argon and flowing the helium through the torch seems great. We will try it.

Thanks very much,
Thomas

[cj737]

Hi CJ, Thanks for your response, If I understood well, you are suggesting to grind off the addition material (remember that the additional material is machined at the edge of the weld joint, we don't add material as usual, with a rod) of the initial 25 mm of the weld and then add material during the overlap.

Not exactly what I was proposing. I was unaware of "the additional material is machined at the edge..." I understood your original post to be the filler was added as the rotation began, and it was this initial section (25mm) of original weld was holding the overlap from reaching full penetration.

My mistake apparently-

[trainingGrounds]

Is there no way to make it slower for the first few mm until adequate penetration is achieved then bump it up to the 1.2mm/s mid weld?

[tshaw]

Hi Gentlemen,

Finally we have made substantial progress with the lack of penetration at the weld start (actually, overlapping the weld start). First we changed the shielding gas from argon to a mixture called NAVY that is 30% argon and 70% Helium. This improved the penetration without the side effected that I mentioned previously when arcs were opening everywhere inside the chamber (with pure helium). Then we adopted a current profile that at about 90 degrees before reaching the start of the weld, we made a ramp up so that at about 1 inch from the weld start (where we had the lack of penetration) the current stabilized at about 120% of the steady state value overlapping the start of the weld and getting the full penetration during this overlap. I believe that the key actions were the NAVY gas and starting the current ramp up a way before overlapping the weld start so that more heat was delivered to help the penetration. Note that this additional heat was delivered overall (during the ramp up), not locally as we were doing before, what produced that massive local lump.

Thank everybody for the help.
Thomas

[LtBadd]

Sounds good Thomas, thanks for posting back. I would be interested to see any photos you can share of the weld and/or your setup

[tshaw]

Hi Richard,

Pictures attached.

Thomas

[pgk]

Looks much better!

[LtBadd]

Very nice Thomas, looks like you are crushing it now, thanks for the equipment and weld shots!