Bend test indications

By Joseph P. Kane (****)

Date 06-01-2007 12:28

I think they look like typical lack of fusiion defects.

By jwright650 (*****)

Date 06-01-2007 12:30 Edited 06-01-2007 12:54

Me too.

edit: especially this one along the groove face.

By Stephan (***)

Date 06-01-2007 12:45

cudaxtreme,

I fully would agree with Joseph and John!

Would you please allow to kindly ask a serious question?

Currently I am preparing a paper where I am describing that also the GMA-Pulsed-Welding can - under particular circumstances - produce some critical weld-discontinuities.

I have named "Undercut" and "Lack of Fusion" as two of the most common known ones.

Would you permit me to use one of those excellent jpg's for integrating it into my paper?

Thank you in advance and regards,
Stephan

By cudaxtreme (*)

Date 06-01-2007 12:59

No problem Stephan. Feel free to use these photos.

These joints were machine welded. There is no weaving on the hotpass, could it be the reason for lack of fusion?

I do realised that with the increased CO2 content, the slag tends to accumulate at the weld toe.

I'm also experimenting open root welding with Miller's Pipepro RMD-Pro, the welding speed seems slow compared to manual tig.

By Stephan (***)

Date 06-01-2007 13:51

cudaxtreme,

first of all thanks a lot for allowing me to use one of the pictures.

I guess I will take that one, John Wright has used for explaining the subject of matter.

Without having further details it's a bit ventured to making some assumptions on the failure causes.

But on the very first view I would suppose the welding-power in relation to wall-thickness and welding-speed has to be adjusted - as mentioned - only under the consideration of not knowing further details.

I would guess it is to be seen as a "pro" you're not weaving (under using 0.2m/min. welding speed), since I suppose in case of weaving the weld-pool's volume would be much more critical by being increased in its size, and thus it should be additionally harder to control as without weaving.

Normally a rising CO2-content in Argon (I guess your shielding-gas is Argon based) is improving the depth of fusion, but on the other hand some drawbacks - as mentioned by you and fbrieden - as increased metallurgical pool- and droplet-reactions (oxidizing alloyed elements Si + Mn] = generating slag) are to be observed with a rising CO2-content, as well.

I guess - as mostly - there might be more than one reason for producing lack of fusion and thus, the failure of the side bended specimen (torch-positioning angles, welding-performance, welding-speed...).

But I guess it might be the relationship of the currently used welding-power vs. welding speed and the wall-thickness of the coupon which has to be reconsidered.

This is my humble opinion but I am 100% sure that you will quickly achieve better results, or at least... replies coming from the other appreciated fellows!

Regards,
Stephan

P.S. By the way, the question coming from Jeff with regard on "why pulsing carbon steel" is a pretty good one.

By fbrieden (***)

Date 06-01-2007 13:04

I've encountered similar indications that were caused by silicate entrapment due to improper interpass cleaning techniques. RT revealed these indications initially and mechanical testing confirmed them visually.

By js55

Date 06-01-2007 13:38

While I can't pull up the pictures to see for myself, and therfore cannot see the spcific application, I would certainly take Joe and John's conclusions for the defects.
Fusion is after all a consistent problem with pulsing applications. TWI has produced voluminous literature on this very subject.
If I may ask, why are you pulsing carbon steel?

By Joseph P. Kane (****)

Date 06-01-2007 15:08

I agree that silicon slag entrapment is the problem. I once watched a good UT INspector come in to a shop and tell the welder that he had 42 layers in this joint. They were welding but not cleaning betwen each pass. In most cases, as they were welding each pass, you could see the silicon ponds. I told the company that they would have to stop between passes and clean off the welds, but the welder said he didn't have to clean them off, they would boil off with the next pass.

When the UT Inspector rejected all the weldments, The company nearly went out of business. A month later, something similar happened on some MCAW welds, and the company had not even completed re-welding the joints on the rejected weldments. The company folded.

Often the silicon ponding is not so evident. Sometimes it is just a slightly darker colored spot on the weld, and if you don't know what it is before hand, you won't realize what it is until you chip it and see it crack or fly off.

By MBSims

Date 06-01-2007 15:11

I agree with the others that this is a lack of fusion defect. In understanding why, I'm wondering about your "GMAW-P" parameters. The amperage sound a bit low for pulsed spray. What are the pulse settings (peak, background, pulse width, etc.)? What shielding gas are you using? What wire size?

By cudaxtreme (*)

Date 06-01-2007 15:45

The only electrical variables that can be adjusted by the welding operators are the voltage and WFS. The rest are machine regulated. I'm using the pulse-pro program, 80%Ar 20%CO2, 1.2mm wire. Torch angle is set at 90degrees to work piece, 1G position.

The weird thing is, I'm not getting any defects if the welding is continuous and would get defects like this when we stop for interpass cleaning.

The reason why I chose pulse? The batch of welding operators I got do not understand english and since pulse can weld thinner material and have less spatter. I thought it would be easier when they do not have to meddle with other settings other than the voltage and WFS.

By Stephan (***)

Date 06-01-2007 17:30 Edited 06-02-2007 08:35

cudaxtreme,

I agree with Marty.

By knowing now some more details I would confirm again my predication.

The parameters you're using are - from my point of view - much too low to secure a sufficiently depth of fusion.

For using a 1.2 mm wire electrode in diameter and welding a plate of ~ 10 (?) mm thickness I would recommend to increase the current significantly followed by an additional increase in welding speed. I wonder about the ratio of amperage (low), voltage (high) and welding speed (low). From my point of view and under consideration of the pretty low welding speed the arc might burn upon a weld-pool whose volume is "buffering" the effective arc's thermal energy and thus avoiding a safe fusion in the depth.

Manganese oxides and Silicates are products of partly intricate thermochemical reactions whose measurable amounts are strongly influenced by temperature, time and the different existing partial pressures of the gaseous constituents within the arc plasma.

I.e. the higher the temperature, the amount of free reactive oxygen and the longer the time the molten metal can react with the oxygen, the higher the amount of manganese- and silicon-based slags which have - of course - to be removed before welding the next layer. By the way, it is interesting to consider that the specific oxidation within the arc-atmosphere is approx. 8 times the amount of oxidation within the liquid weld-pool...

To resume, due to I am working very often in the field of automated welding-applications (robotics), I am also often asked: "Stephan, how fast can we weld by using solid single wire GMAW in combination with a robot?"

Well I promise, you will never hear any kind of accurate prediction in "x mm/s" from me (believe me, I know what I am talking about and I guess you know what I mean) but as a rule of thumb I say 10 mm/s in mean is a real value to calculate with for economical efficiency of the system. This means 0.6 m/min. but often more. In GMA-Tandem-Welding we are often required to adjust > 30 mm/s welding speed to making the system economically effective for the user.

However, when I am converting 0.2 m/min into mm/s, you are welding with 3.33 mm/s, which is from my standpoint definitely to less.

So once again I would recommend to increase the arc-performance + the welding speed, and - eventually - to change the torch drag angle to a slight backhand one to use the arc-pressure for repelling the molten weld-pool .

Regards,
Stephan

By cudaxtreme (*)

Date 06-03-2007 16:54

Thanks for the advice Stephan,

I will try to adjust the gun angle from perpendicular to a 10-15degrees drag angle and increase the WFS.

BTW sidetrack abit, I'm using premixed gases 80%Ar 20%CO2 for carbon steel and 99%Ar 1%O2 for stainless steel. I wonder if the gases would segregate if left alone for a long time?
Ar is lighter than CO2 thus have the tenancy to move to the top of the cylinder?

By Stephan (***)

Date 06-03-2007 17:50

cudaxtreme,

thanks for the reply.

I may hope you will succeed..!

With regard to your question my answer would be "No", there is no segregation to be expected due to the statistical behaviour of gaseous matter, in particular "Browns motion".

Since it would probably lead to far in explaining it in detail I would suggest to search the Internet for "Ideal-" or "Real-Gas".

I have found a pretty simulation I would like to recommend you to having a look on or download it, respectively:

http://ideal-gas-in-3d.nasanbat-namsrai.qarchive.org/_download2.html

My very best regards and good luck in your efforts,
Stephan

By swsweld (****)

Date 06-04-2007 03:42

I'm no expert on P-GMAW but I have used P-GTAW alot. It looks like the indications are on the toe of the weld between the root & hotpass. Is the face of root pass flat or convex? If convex, the pulsed hot pass may not be able to get proper fusion on the walls. What degree is your groove angle? Too steep of angle would make fusion more difficult. Do you have occilation capabilities with your system? We used more current to fuse the wall and less current to travel across the face. If the PipePro does not have this ability is it possible to run a split bead hot pass with angle tilted slightly towards toe?

It sounds like you don't want the welders to adjust the numbers or equipment so the last suggestion may not help.
Page 60 in the June 07 Welding Journal mentions the importance of cleaning. The disc grinder may be leaving contaminates. It may not take much to prevent the pulse action from missing a good tie in at the toe. You might use a power wire brush and/or a chemical solvent to clean after the disc or instead of the disc. I know the article was on alloys not carbon steel but if the parameters are borderline it could have some effect.

By Stephan (***)

Date 06-04-2007 06:39 Edited 06-04-2007 07:43

Tim,

good points!

In particular the convex bead forming is a crucial one.

This is also to be considered when using a backhand drag angle, which benefits the forming of a more convex seam cross section.

Regards,
Stephan

By cudaxtreme (*)

Date 06-04-2007 08:16

Here's what I have done.

Adjusted the gun angle to 10degrees backhand.
Adjusted the arc cone wider to produce a flatter weld
Increased the travel speed to make sure the arc is always at the edge of the weld pool
Increased the weaving(side wall) delay time.

The groove angle is 60degrees

By Lawrence (*****)

Date 06-04-2007 12:32

Cuda.

Are you using 0.045 solid wire? 0.035 wire does not like to pulse, even with the best new Miller power supplies. I'm surprised you feel the RMD is slower than GTAW, I just can't imagine how that could be happening. RMD shoud lay in the root with just about the exact same feed rates as regular short circiuting transfer, the only difference is that RMD is more adaptable and consistant than short circuit for operator or fitup changes in gap, land, and stickout distance.

It appears you might be trying to achieve a goal of doing your entire pipe with a single gun and single wire, which the PipePro ought to be able to do if it is as good as advertized. However, if you stick with the GTAW root or even if you don't there may be other more consistant alternatives.

Not to be a smart guy here but;

Since your working in the 1-G position you could easily chuck the $10,000 pulser, or just switch it off and put on a roll of E70T-1 0.045 FCAW wire... You could increase your travel speed about 40% and kiss your fusion problems good bye. If you choose E71T-1 you could do the same thing in all positions.

Still only WFS and Voltage to contend with. The travel speed increase and lack of spatter will more than make up for slag removel (it will just peel up anyhow.)

By cudaxtreme (*)

Date 06-04-2007 14:14

FCAW is the next thing to explore but first I must make sure the GMAW is working well.

By Joseph P. Kane (****)

Date 06-04-2007 14:31 Edited 06-04-2007 14:38

One thing I found years ago, is that the exact position before and after Twelve O'Clock was more significant than the work angle and the drag angle. It sounds like it may be the same problem I encountered training the welders, who were using the part rotator. Depending on which pass they were applying, the start point had to be adjusted to within a very narrow margin +/- 12 O'Clock.