Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic Welding Industry / Technical Discussions / Can we detect presence of CS filler in SS304L GTAW welding?
- - By ramakant_v (*) Date 06-28-2007 15:17
Hello all, I am facing a problem of detection of use carbon steel filler wire in S304L welding. The following is the details of the problem:

1. Welding process: GTAW
2. Parent metal: SS304L seamless pipes (1" Sch40 to 2" Sch 40)
3. Usual filler wire: ER308L

Problem: Accidentally similar looking filler wire which is used for carbon steel was applied for making root weld run of long and complicated pipe  line. However, subsequent weld runs were made by using correct grade of filler wires i.e. ER308L. Everybody will agree SS304L material is basically used to combat severe corrosive environment and definitely use of carbon steel filler wire a big blunder. Can somebody suggest any definite NDT method for detection of CS in these welds. The constraint is that there are very large number of joints involved and there is a mix up of the CS and SS based joints. I have tried ferrite meter but ferrite measurement does not give any definite change from the expected values. Can somebody throw some light on the issue.

regards
RK
Parent - - By hogan (****) Date 06-28-2007 15:28
if you use copper sulfate you will see the different metals (a spray bottle worked the best for me). one will be gold in color and one silver.
Parent - By jon20013 (*****) Date 06-28-2007 15:50
Nitric acid solution.  One part nitric added to 3 parts water is how you woul ddo an etch but not sure if this would work in your case?
Parent - By ramakant_v (*) Date 06-28-2007 17:20 Edited 06-28-2007 17:24
thanks for the reply. Definitely copper sulphate will work and tell about the presence of CS. But i cannot access the pipe line from inside and i cannot cut the pipe line to see the micro structure or etch structure. more over applying Copper sulphate on outside will probably show no result because is is totally covered by SS308L filler wire. But definitely I will give it a try.
Parent - By Stephan (***) Date 06-28-2007 17:01
RK,

good hints coming from hogan and jon20013 - as ever.

Nonetheless I guess it's hard in the field of NDT finding somewhat, bringing reliable assessments in regard to your topic.

I would - if ever - recommend to try Ultrasonic Testing due to Ferritic and Austenitic Materials have different physical properties in regard to their sound propagation (as far as I know based on their different microstructures).

But as an expert has told me once, it is tough and needs lots of experience to test only austenitic welding-joints. Thus I suppose it's a power tougher to Ultrasonic Testing joints you have spoken of. Additionally there is surely a varying amount of dilution between the materials to detect, complicating the matter certainly.

However meanwhile there are - as I could find out - some improved technologies been developed. In particular for welding-joints on dissimilar materials.

See also:

http://www.ndt.net/article/v04n09/erhard/erhard.htm

http://www.ndt.net/article/v05n08/elbern/elbern.htm

Perhaps this helps a bit but I am sure there are real NDT-experts in the forum to "throw some light on the issue"...

Kind regards,
Stephan
Parent - - By MBSims (****) Date 06-28-2007 20:13 Edited 06-28-2007 20:37
A test used for detecting embedded iron on austenitic stainless steels is to wet the surface with water and wait 24 hours to see if rust stains appear.  A faster method is to use a copper sulfate test.  The iron contamination areas will appear copper colored.  You could pour this solution into the pipe spools, slosh it around, then flush with demin water after allowing it to work for a few minutes.  Then the weld roots would have to be visually inspected with a boroscope or video probe to identify which welds are affected.  ASTM A380 describes the copper sulfate test and appropriate pickling solutions.  Of course pickling would not be a benefit for welds with carbon steel root passes.  These would have to be cut out, machined back to remove the weld and fusion zone, then rewelded.  On pipe this size, you may be better off to just start over with new material.

http://www.thefabricator.com/TubePipeProduction/TubePipeProduction_Article.cfm?ID=888

http://www.geinspectiontechnologies.com/en/products/rvi/vp/xlg3/overview.html
Parent - - By js55 (*****) Date 06-28-2007 20:22
I think Stephen may be offering the best course of action. We may need the UT guys chiming in on this one. If you can't access the ID the only way I can see to possibly detect the CS is UT.
But, by the time you consider penetration of the root pass with SS hot pass or filler , reducing the pure CS to perhaps a thickness of 1/16" or less, can you read a defraction from austenitic to ferritc that close to the ID wall?
Parent - - By MBSims (****) Date 06-28-2007 20:33
I think it would be very difficult to UT the 1" to 2" schedule 40 pipe, even to do a full volume exam.  Focusing on the ID surface will require focused array and/or a cal block that represents the problem being detected.
Parent - By hogan (****) Date 06-28-2007 20:43
very dificult with UT. even with an expeienced UT tech. i don't think it would worth the effort, due to acuracy of results.
Parent - - By CWI555 (*****) Date 06-28-2007 20:57
This would be very difficult. Phased array UT maybe. I do recall something with Eddy current for materials evaluation. It's been a long time since I've chased a butterfly, So any current experts out there please chime in. I believe it was either in welding journal, materials evaluation, or inspection trends magazine did a piece on this not long back.
Parent - By js55 (*****) Date 06-28-2007 21:21
ramakant,
There's the consensus. Unfortunately if you really believe its possible they were welded with CS it sounds like cut outs.
Parent - - By CWI555 (*****) Date 06-29-2007 01:09 Edited 06-29-2007 01:25
Maybe with eddy current and a measure of impedance therein.
http://www.aws.org/itrends/2006/01/011/
Parent - - By Shane Feder (****) Date 07-03-2007 02:06
Greetings everyone,
A couple of suggestions that may help;
1  Would a magnet work ? The carbon steel root run will give a stronger pull on the magnet than the fully stainless weld.
2  Positive Material Identification with a portable analyzer ?
3  As the WT is only 3.38 and 3.91 mm grind a small area of the cap off on the top of the weld and test with copper sulphate, the root run would have just about come out flush so you may be able to expose any carbon to the surface.
Hope that helps,
Shane
Parent - By js55 (*****) Date 07-03-2007 13:22
I think Shane has hit on something here. Why worry about accessing the root. PMI the weld cap. With a wall thickness in that range the dilution effect from a CS root will still be effecting the chemsitry of the cap. No way an 18-8 balance is recovered in just two or three passes. Just ask the overlay guys. I'm gussing the Ni and Cr show low.
Parent - - By CWI555 (*****) Date 07-03-2007 17:15
After a little research, I think the Eddy current method could provide the right answers within a reasonable error factor.
A Diffraction PMI would not pick up the ligher elements such as carbon. Spark emissions may, but I believe the Eddy current would be the way to go. If it hits an area that has been welded in any part with carbon in a stainless weld, it should speak up loud and clear.
Any areas in doubt, perform test as per Shanes comments for copper sulfate.

My two cents worth,
Gerald
Parent - - By js55 (*****) Date 07-03-2007 18:39
Gerald,
Eddy current may be the way to go. I dont know enough to say one way or the other. But I don't believe carbon matters. Or at least won't be the definitive element. Dilution from the base metal and subsequent SS filler will most likely leave you with a carbon content still in the SS L grade range. Telling you nothing.
However, because neither Ni or Cr are present in carbon steel in anything but trace levels dilution should drop them to below the SS chemical range.
Also, if I may ask, upon further thought, will the copper sulphate work. Isn't it designed to detect Fe oxides. SS has a high Fe content itself. Grinding into the the root pass will not uncover Fe oxide, but Fe in solid solution. If copper sulphate worked on Fe content in solid solution of a weld metal it would reveal it in SS too. Wouldn't it?
Parent - - By CWI555 (*****) Date 07-03-2007 23:29 Edited 07-04-2007 00:41
A380
"7.2.5.3 Copper Sulfate Test--This method is recommended
for the detection of metallic iron or iron oxide on the surface of
austenitic 200 and 300 Series,"

The test is designed for free iron/iron oxide. A stainless weld welded with the correct filler such as found in SFA-5.9 will not exibit the same reaction.
Making a determination as to if this will work would be very simple. weld two small samples. one correctly, one mixing the filler. Proceed with A380 method
for copper sulfate. It can be proofed one way or another.

The concern is that some carbon steel filler has found it's way into the weld. The nickel content in SA 240 304 is 8-10.5 and the chromium is 18-20, for the weld material according to SFA 5-9 standard 308 is chromium 19.5-22.0 and 9.0-11.0 nickel. It is not full proof as it's dependant on specific dilution percentages. If it's gross it will ring out loud. However, the carbon content for standard 304 .08 and depending on the grade down to .03, an excess of that would be a smoking gun in this case.

Standard Eddy current is a product of electromagnetic induction. There are other methods that derive their basics in the same such as RFT and flux leakage.
Given that, the Hysteresis Loop for stainless weld vs a carbon weld will be dramatically different, and since eddy current works from electromagnetic induction it should be a simple matter to put together a couple of standards that will test this problem with reasonable accuracy without cutting into the weld to begin with.
Now if the thickness were to great, you may not get a reading from the center, so it will likely be limited to .300" wall at most depending on the input current.
I think it will work for the intended need.
My votes for a cross breed of eddy current with copper sulfate for anything any doubt as per Shanes comments. Once many years ago I was certified Level II ET but it's been a while since I've chased a butterfly. So if there is a level III ET out there please speak up.
I do know that material identification is listed as one of the uses of eddy current.
Parent - - By ramakant_v (*) Date 07-05-2007 07:01
Hello all,

I must appreciate the kind of interest and valuable suggestions I got from all of you. Thanks for that. Yes, i have tried to use magnet and this is surely a good solution, the pull is definitely more in case of joints with CS at root. But all this depends on the experience of the person and this "feel" of pull is really operator dependent. Moreover, in case the weld contains more delta ferrite, there is a greater attraction towards the magnet due to its presence and there is a doubt again against the genuine welds. So the use of magnet is definitely a positive solution but it leaves little bit of doubt.

Copper Sulphate Test:
I tried this test on my many samples but it did not work. I think dilution effect is not very prominent and CS material remains probably up to root or slightly up to the next weld layer. More over, as there are large number of joints (of the order of few thousand), grinding all the joints and capping it again with weld after CuSo4 testing is really a troublesome task.

I would like to share some more information with you, i have tried eddy current testing method with frequency set at lower end. And i am getting some positive indication. My tests are still on and I will discuss and share with you all after the tests are over.

Can anybody suggest the effect on Ferrite Number on the crust of the welds which has CS at root of it?

Regards,
RK
Parent - - By hogan (****) Date 07-05-2007 14:54
this is a bit out of the box, but it seems that the magnet is working (somewhat poorly), and you said you have some samples. i'm guessing but, one with a cs root and one without. could you wrap a welding lead around the pipe, at a set distance and check the weld with a Gauss meter? the cs should produce a measurably higher field strength. It would take a bit to refine the procedure, but it might work, maybe
Parent - By CWI555 (*****) Date 07-06-2007 04:10
What you suggest is in effect a flux leakage test. The resulting longitudinal magnetic field would set up a flux leakage condition at any flaw assuming a ferromagnetic material.
SS is paramagnetic and CS is ferromagnetic. What you've suggested should work with some modification. Since SS is paramagetic, when the current is shut off, the SS part due to it's low permeability and low retentivity would drop off to near nothing on that part. Any area containing CS should retain a residual field which may be detectable via the Gauss meter assuming a high enough initial current is used. I can't say as I've ever heard of what you said before, but I believe the theory is sound enough. I may add that a hall effect tesla/gauss meter may be needed for sensitivity reasons.
Parent - - By billvanderhoof (****) Date 07-06-2007 01:36
This is a pure guess but if you take an ordinary compass (the north finding kind not the circle drawing kind) and place it on the joint if the joint is magnetic and not currently magnetized the needle should align with the joint.  The compass must be level or the needle will bind and of course if the joint is already alligned north and south no info will ensue.  If the joints are north/south alligned it may be possible to cancel the earths magnetic field with another magnet somewhere nearby.  I leave the experimentation to you and wish you good fortune.
Bill
Parent - - By ramakant_v (*) Date 07-08-2007 05:06
Hello all,

I would like to update all of you. I have cut many test coupons and after etching by NITAL Solution, in some of the wels coupons, wherever penetration is  little less, it seems that CS has completely dissolved in SS matrix. However, at places where weld penetration is little more (i.e. it is above ID surface of the pipes) CS is visible. How can it happen. I am still confused. But magnetic properties of  the weld material with CS at root is definitely greatly enhanced.

My eddy current tests are still on I will update after a test trial on my actual welds.

Regards,
RAMAKANT
Parent - By Kix (****) Date 07-10-2007 15:05
Well, 1'' AND 2'' sch 40 can be done in 2 passes easy, but some guys may need to run 3 so the carbon layer can't be that deep under the cap pass unless your boys are trying to hide some hi lo from the RT guys by putting a gorrila cap on there.  So i'd say if all else fails for you to grind into each weld a just a little bit and put some sort of acid in the grinded area then wipe it off and see if you get some corrosion or rust.  The pure stainless should have no rust and the diluted stainless should show signs of corrosion.  Use your test welded samples to see how far in your have to grind and if you can get any sort of result.  Have your welder on hand to be touching up the grinded area and if you come up on a cut out he allready is there set up to weld.  Just a thought, but i'm sure you will come up with something before this would have to be done.
Parent - - By CWI555 (*****) Date 07-13-2007 04:25
Will be very interested to hear the results.
Parent - - By ramakant_v (*) Date 07-18-2007 18:03
Hello All,
Sorry for a little late response. I was out to site for one week to check the suspected joints by eddy current method. I made many mock up test pieces with varying pipe sizes (8 NB to 50 NB Sch 40) and tested on Machine. The operating frequency was set to 1000 Hz only so that eddies can be generated up to 4-5 mm. The results were really sharp, crisp and very distinguishable. We used all the test pieces and the machine gave immediate signals and we were able to see the presence of even slightest amount of CS at root. The repeatability of the test was also very very good. I checked many joints at site and found some indications and now testing of other joints is going on. SO PROBLEM IS REALLY SOLVED. I WOULD LIKE TO EXTEND MY SINCERELY THANKS to all of you.

I would like to share some more tests i carried out. The cross section of the pipes were etched by:

1. Oxalic acid
2. Nital Solution
3. Aqua Regia

In etching test by all above, specially for smaller pipes up to 25 NB Sch 40, the micro structure was really confusing. To my surprise there was no presence of any CS at any point of the cross section. No presence of pearlite structure i.e. CS WAS COMPLETELY DISSOLVED IN SS. This even happened at point where weld penetration was on excess side..!! But definitely magnetic property at root was more than that present at outside of the weld.

For 50 NB Sch 40 pipe, at root where penetration was slightly excess, up to 1 mm clear CS structure was observed in Nital Solution.

To find out what disappearance of  to CS in SS weld, I tried to check composition of the weld metal at different distances from root. As the thickness of the pipes is very less, the conventional methods were of no use because they will not give composition of any particular point. I used EDS method  and it really proved that there is a complete dissolution of CS in SS. 1 mm above root % of Ni and Cr was really half of the filler wire. As we moved 2 mm from the root, the composition of the weld was that of filler wire.

I tried these also:

1. Magnet: it really works but is "feel" dependent.
2. CuSO4: because of complete dissolution of CS in MS, it did not work. I tried this at root side also, and surprisingly, no instant golden copper like color appeared. Slight golden color appeared only after 5-6 minutes where as it should have appeared in seconds in case of pure CS. On top side of the weld, there was no color change at all. So applying CuSo4 on out side of the weld was not useful.

So friends all sorts of information i have shared with you. Any other idea is surely welcome. My on job tests are still going on. As huge number of joints are involved, it will surely take very long time to check all of them.

With thanks and regards
Ramakant
Parent - By CWI555 (*****) Date 07-19-2007 22:28
Glad to hear it worked out for you.

Regards,
Gerald
Up Topic Welding Industry / Technical Discussions / Can we detect presence of CS filler in SS304L GTAW welding?

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill