My 316L stainless steel weld is magnetic!!

Date 02-14-2007 12:53

Actually, there is no "problem". 316L, due to the composition of the chemistry, is likely to be magnetic to some degree. Just because some of the welds are magnetic and others are not, is a bit misleading. Usually, the 316L contains sufficient ferrite (which is magnetic) so that depending on the strength of the magnetic, will register as magnetism. I would venture to say that if you used a sensitive enough magnet you would have all the welds register magnetism. The 316L molten weld pool is austenitic and as it cools to solidification, some ferrite will form around the grain boundaries. The cooling down rate will greatly determine the amount of ferrite present. Just because you say that only 30 joints are magnetic is probably registering magnetism only because of the strength of your magnet. Again, the magnetism of the welds are in direct proportion to the amount of ferrite present, and this is not unusual. I once worked with the U.S. Navy (Henry, your Navy) and they were experiencing some magnetism in some of their underwater welds, which they could not allow. They ended up having to use a pure austenitic welding material (E310-17) so no magnetism would be in the weld. Ferrite is magnetic and austenite is not. So, if your application is to allow absolutely no magnetism, you can use a 316L LF (low ferrite) which contains virtually no ferrite or you will have to go with a pure austenitic filler metal.

Chuck

By Farshid Date 02-14-2007 18:58 Edited 02-14-2007 19:29

Dear Mr. Chuck,
Thank you for your reply.
The case is not to have a none-magnetic weld. The case is doubting to use carbon steel filler metal.
You are right about ferrite and austenite. I know it is principal of ferrite number checking device. However, is around 7% ferrite in 316L high enough to hold the coin size and shape magnet?
And I am agree about rate of cooling which influence delta ferrite rate. As in duplex stainless steels Nitrogen or some extra Nickel will be added to filler to prevent extra ferrite. But in a same position and welding process and wall thickness, our pipes I mean, cooling rate can not vary a lot to influence percentage of delta ferrite. It seams, in our case some welds have a lot of delta ferrite to take the magnetic tool and vice versa. What is your idea about this delta ferrite percentage variation?

By js55

Date 02-14-2007 20:02

Unless I misread your post, I'm not sure you have a problem. You state that essentially you ended up 'sorta' using a magnet as a PMI. You state concern for welds produced with CS, yet also state that you have magnetism on welds you know to be perfomed with SS, including a SAW weld in which the magnetism changes through the bead. Unless FN is a specification issue I do not understand what your concern is.
If you wish to know specifically why the magnetism variance, my thought is that only photomic's will solve it. And not even then maybe.

By ssbn727 (*****)

Date 02-15-2007 02:54 Edited 02-15-2007 03:13

I have to agree with everything mentioned, and would like to commend Chuck once again with his explanation as to why some area's are magnetic and why some are not...

I also concurr with Chuck's explanation regarding the use of E310-17 as opposed to let's say E316L however, It's not my Navy Chuck!;) It's OUR NAVY!!!;) Second to NONE - Still!!! Seriously though - from an economical & strategic perspective, I can think of three different types of problems that come to mind with respect to an excessive amount of ferrite content must be taken into account when it comes to designing with stainless steels for use on both surface & submersible naval vessels... One is Corrosion... Two is embrittlement at high temperatures... Three is the fact that the USN would have to perform alot, and I mean alot more "degaussing" than they already do in order to lower the electromagnetic signatures on both their Surface Ships & Submarines for a variety of reasons for which I will not get into because, I do'nt want to "let the horse out of the barn" so to speak! There may exist other problems but with the lousy weather outside, I can only think of three.

js55 also makes some very good points especially when it concerns the potential for variances in the FN's in the welds... Farshid, if you really want to dig more into the issue then get an Elcometer 111-7F ferrite indicator which is accurate to + or - 10% of the reading or (YES! the readings are in ferrite numbers in accordance to WRC & IIW recommendations) + or - 1FN, so long as it's properly calibrated unlike the "Coin" shaped type magnet that Farshid probably used...

Anywho, check out this interesting article on shot peening with respect to minimizing SCC in 300 grades of stainless steel: http://www.shotpeening.org/ICSP/icsp-1-54.pdf

Notes on Ferrite Numbers & Calibration from Elcometer Inspection Equipment...
"After much testing around the world, it has been decided that it is impossible to determine accurately the true ferrite content of stainless steel weld metals. Previous figures of percentage content were very dependent on their source, the measuring instrument used ands its calibration. So Ferrite Numbers were established to define ferrite quantity and are related to the ferrite quantity of an equivalently magnetic weld metal."

"The calibration of a gauge is such that the ferrite reading can be related to the seperation of the gauge's magnet from low carbon mild steel plate. The primary standard is thickness and the secondary standard is ferrite samples, and the Calibration Certificate supplied with the Elcometer 111-7F relates the gauge reading to both standards." ANSI/AWS A4.2-97, ("Standard Procedures for Calibrating Magnetic Instruments to Measure the Delta Ferrite Content of Austenitic and Duplex Austenitic-Ferritic Stainless Steel Weld Metal.") is the standard used to calibrate in order to maintain traceability to internationally recognized secondary standards.

Here's an interesting article that farshid might want to review:
http://www.egmrs.org/EJS/PDF/vo291/151.pdf

Then there is an alternative to the Elcometer 111-7F...
Here's the webpage & .pdf on the Fischer "Feritscope" MP30E which is convenient because it shows readings in both percentage of ferrite or in FN's starting from 0.1 to about 90FN:
http://www.helmut-fischer.com/ProductDetail.asp?PN=603-712
http://www.helmut-fischer.com/globalfiles/DE_Ferit_MP30E_0605_EN.pdf

Then of course like js55 mentioned, there are the "Money is no object" methods of Positive Material Identification instruments which include: 1.The "XMET 3000TX" Portable Alloy Analyzer from Metorex (Finland), Metal Master 2000, Model 898 Alloy Analyzer from NITON and Model XMET3000 from Metorex. These analyzers can analyze alloying elements in Carbon Steel, Alloy Steel or Stainless Steel, Tool Steels, Ni Alloys, Cobalt Alloys and Copper Alloys. The elements that can be analyzed are Cr, Ni, Mo, Mn, Fe, Ti, Nb, Zn, Cu, Co, W, V, Sn, Pb, Ta, etc. The identification is done on basis of major alloying elements only and grade of material is reported by our inspection engineer. Please note that this instrument will not analyze elements like C, S, P, Si, Al, B etc and hence material like 316L and 316 can not be segregated by these instruments. Here's an article in AWS "Inspection Trends" that covers X-ray Fluorescence FP-XRF analyzers: http://www.niton.com/documents/inspectiontrendsarticle0104.pdf

Here's another one that talks specifically about stainless steel PMI:
http://www.niton.com/documents/cnxstainless.pdf
Here's a "nifty" Niton XLi/XLp 800 Series alloy analyzer specifically designed for in service PMI applications: http://www.niton.com/Content/alloy_analysis/alloy_inservice.asp
Here's the "XMET 3000TX:
http://www.oxford-instruments.co.uk/wps/wcm/connect/Oxford+Instruments/Products/X-ray+Fluorescence/X-MET3000TX+in+Metal+Recycling/X-MET3000TX+in+Metal+Recycling
http://www.oxford-instruments.co.uk/wps/wcm/resources/file/eb6a410811a4b6a/X-MET3000TX_Scrap_brochure.pdf

2.The "ARC-MET 8000" MobileLab Portable Spectrometer
This portable spectrometer is ideal for PMI and can determine all elements in Carbon Steel, Alloy Steel & Stainless Steel. It can be used to analyze Ferrous and Non Ferrous materials including Nickel, Cobalt Aluminum and Copper base alloys along with Carbon. Low Alloy and Stainless Steels. Separate calibration will be provided for each matrix. Thus this is a very precise instrument and is used in lab as well as field operations. It uses Argon Gas for Ferrous samples where determination of elements like C, S, P and Si is essential to determine exact grade. It can analyze elements including Fe, C, Si, Mn, Cr, Mo, Ni, Al, Co, Cu, Nb, Ti, V, W, Pb, and Zr.

ARC-MET8000 MobileLab Optical Emission analyser is designed to identify all the key elements in metals - especially where highest accuracy and/or the analysis of light elements like C, Al, S, P, Mg, Si is needed and when sorting low alloys and aluminums. Ideal, for example, for separation of 316 H (>0.04% C) and 316 L (<0.03% C). Use one unit throughout the whole production cycle! Classification of raw material (recycling metal), QC of semimanufactured products, QC of final product.

Recognizing that carbon analysis in steels is of primary concern in many material verification functions, and e.g. in welding work, ARC-MET8000 has the unique capability of measuring carbon in both air and argon modes. ARC-MET8000 offers: Accurate results using Air or Argon measurements,
Fast grade identification and assay,Only one probe with an integrated display
Low level carbon analysis using Air burn in a few seconds,Battery operation,
long cable between main unit and probe, and convenient mobility
Unique to ARC-MET8000: the probe is the heart of the system. Ideal for ferrous and non-ferrous metals:
Low alloy steels, Stainless steels, Tool steels, Low alloy (white) cast iron, Aluminum alloys, Titanium alloys, Nickel alloys, Cobalt alloys, Copper alloys, Zinc alloys, Magnesium alloys... Other metals can also be added.

Reference: Oxford Instruments: http://www.oxinst.com/wps/wcm/connect/Oxford+Instruments/Products/Optical+Emission+Spectroscopy/ARC-MET8000/ARC-MET8000
http://www.oxinst.com/wps/wcm/resources/file/eb2dc70735e235f/ARC-MET8000_brochure_June_2005.pdf

So now that you've got a sampling of the methods currently used in determining weld metal alloy verification specifically involving the stainless steel grades, I hope you choose the method best suited for your needs... Btw, you can rent some of the equipment described above as an option to purchasing the equipment outright. If you want to know where? Do a google search!

Respectfully,
Henry

By js55

Date 02-15-2007 15:19

Henry,
Great article on shot peening to prevent SCC/IGC in 18-8's. I was fascinated by the idea of using FN or ET to determine adequate levels of peening for these applications. Shot peening is something that can be be quite practically applied. Pipe shops specializing in power piping use ID 'blast' technologies every day. Its a small matter to switch to shot and apply it as peening as opposed to surface quality criteria. Another intersting point is how this thinking has been around for >60 years. A lot of stuff established by these old boys has been set aside and often forgotten.

By Sourdough (****)

Date 02-17-2007 23:45

How's it hangin' Hank? Mines still curving to the left......

By chuck meadows (***)

Date 02-14-2007 23:18

As I stated in the e-mail on my computer, the sensitivity of your magnet might allow it to hold on the weld at 7FN. Personally, I would prefer a Sevren or Magna-Gage to check ferrite instead of a hand held magnet. All that hand held magnet is telling you is there is some magnetism. It doesn't register a % or number, if I understand you correctly. Cooling rate can definitely change even if using the same machine, welding process and wall thickness. When you first start the weld, you will have a quicker cooling rate and it will slow down as you get back to where the weld was started. I have seen many different degrees of ferrite on a single weld seam. The variance is not a lot, but it certainly is capable of varying. We won't even go into ferrites in Duplex, as that has been a topic of discussion many times. But, in your case I don't think you have a problem. But, if you suspect that a whole weld seam might been welded with carbon steel filler metal, I would hope that someone would have checked the grade of the wire before it got put on the machine. Or at least checked it after you thought you had a potential problem.

By js55

Date 02-15-2007 20:55

I just finished the article posted by Henry on Low Temp aging of 316L. Very interesting.
My opinion is that this is a classic example of metallurgical research that far outstrips practical welding engineering, except maybe to emphasize caution when evaluating ferrite volume through magnetic testing. The optical micrograph followed by more sophisticated techniques will be far more accurate. The decomposition of ferrite into magnetic alpha and alpha prime phases, and non magnetic G-phase as responsible for the discrepency between the results of magna gage and micrography evaluation was the primary conclusion, though any deleterious effects of these phases was not expanded upon. Perhaps an embrittling along the lines of sigma or chi may be conjectured. Good stuff Henry. Thanks.

By chuck meadows (***)

Date 02-15-2007 21:48

I very seriously doubt that sigma, chi, or alpha prime has any part in this. A low carbon material would have to be in the embrittlement temperature MUCH longer than any normal welding process. We have gotten off the original post, as the poster was only trying to figure out if it is normal for a 316L to be magnetic. The most accurate way to measure ferrite is the ASTM E 562 method, point counting. Here, samples will be electrolytically etched in either NaOH or KOH in such a manner that the contrast of austenite and ferrite will be optimum distinguishable at 500x magnification. In this test, a minimum of 15 fields and 36 points per field are used. But, in all honesty, I doubt if the poster was wanting to go that far as to do an expensive destructive metallurgical test on a production test just to see if the weld was magnetic. Magna-gage and Sevren gages, or any evaluation using magnetic forces to measure ferrite is strictly a prediction, and is certainly inferior to a destructive ASTM E 562 examination, which is an actual not a prediction.

By js55

Date 02-15-2007 22:01

Chuck,
Yeah, I agree. The times involved in the article were aging times up to 10000 hrs. Certainly not applicable to the posters question. I added it here since Henry had posted the article. I do however think that a caution when using magnetic evaluations for ferrite is applicable. Especially when FN is considered a ferrite percentage. Maybe I should have opened a new string.

By chuck meadows (***)

Date 02-15-2007 22:18

Unfortunately I haven't taken the time to read Henry's posted articles yet (no offense Henry), but I'm pretty aware of the aging of the austenitics, especially at 10,000 hours where some depletion of chromium from the ferrite will result in carbide growth and cause some of the ferrite to transform to austenite. But, the kinetics of this carbide growth is dependent on the exposure time and temperature. Either way, this carbide growth (carbide precipitation) will take place at the austenite-ferrite interphase boundary.

By chuck meadows (***)

Date 02-15-2007 22:06

Also, if any intermetallic phases or compounds are suspected, the ASTM E 407 metallographic examination would be done. Here, the ASTM etchant number 98 (K3 Fe(CN)4 with either KOH or NaOH) would be used. Again, a 500x magnification would be used to determine the absence of nitrides, carbides, and any other intermetaallic phase at the grain boundary. No deleterious phase shall exist, especially sigma and nitrides of aluminum, vanadium, niobium, etc. But, a maximum of 0.05% intermetallic phase or precipitates are allowed. I think the poster had a little bit of m agnetism due to the chemistry of the 316L material, which is normal. But if he wanted to go the route of expensive destructive testing to check for the amount of ferrite and the absence of intermetallic phases, then ASTM E 562 and ASTM E407 are the way to go.

By Farshid Date 02-15-2007 22:08

Actually I was thinking before, that relatively small percentage of delta ferrite in the 304 & 316 are not enough to have any influence on a simple magnet; therefore always, as a simple test, I have used a magnet to check 304 and 316 welds.
I have asked the contractor to check the welds by MPI. He does not have any MPI device and he preferred to cut out and re-weld the joints!! Because it is much less expensive for him comparing to buy a MPI device.
Thank a lot for all of incomes.

By chuck meadows (***)

Date 02-15-2007 22:25

Do you mean PMI rather than MPI? Like we've discussed before, the sensitivity of the magnet you use can certainly detect ferrite (magnetism). It's your ball game, but I think it is going to extremes by having him cut out the welds just because you detected soome magnetism in the welds. But, that is up to you. I would have changed magnets. LOL

By ssbn727 (*****)

Date 02-16-2007 02:44

Hi Farshid!

Please try to convince the contractor to either follow Chuck's suggestions, or rent some PMI equipment, or to contract someone that can do this type of work which btw, would be far less expensive and time consuming than to cut out and re-weld all of the joints based on what basically amounts to a guess especially when one compares the overall cost between the two choices.

The other choice you have is to cut out one of the questionable joints that would represent all of the joints in question, and to choose one that has relatively easy accessability in repairing the joint... Send the material that is to be cut out to a lab that has the equipment perform PMI on the samples in question.

This way one will minimize the potential amount of unecessary rework, should the samples show that the material is of the correct grade for the job in question!!! If the PMI shows the material is indeed sub-standard, then at least one has proof of it, and the appropriate parties would be backcharged for the mess but, if the material is found to be correct and yet the welds are cut out - based only on a hunch, then I think some folks would react like "Ricky Ricardo" in "I Love Lucy" when he used to say: "Honey! You got alot of essplaining to do!!!" if you get my meaning!!! I mean after all , one must COVER THEIR BEHINDS in a case like this!!! Otherwise known as "CYA". Either way I would never make a decision based on what basically amounts to a "Hunch".

Farshid, whatever the contractor decides to do in order to remedy the situation is up to them but, I would make a suggestion to the contractor that they subcontract someone to perform either ferrite% verification, or some sort of PMI beforehand in order to "CTB", so that once the ferrite %'s, or the material composition of the questionable welds are verified before re-working all of the joints, no one will look like they made an expensive guess... At least one that would probably be more expensive than the "OVERALL" expense of having to perform either Ferrite % verification via severn gauge, "elcometer" or Fischer "Feritiscope" on all of the joints in question like Chuck suggested, or to subcontract a professional to perform PMI on only one of the joints in question via their lab or on site in order to verify a "hunch"...

I would hate to be in the position of performing a complete re-work of all of the questionable joints, only to find out later that it was'nt necessary to re-work all of the joints in the first place. Anywho, best of luck!!!

Respectfully,
Henry

By Farshid Date 02-16-2007 12:41 Edited 02-16-2007 13:12

YES SORRY. I MEAN PMI TO CHECK ANALYZE OF THE METAL. Unfortunately it is not in my hand to stop the contractor. I just asked for some shots of PMI to check analyze; and if I stop him to do nothing (not PMI not cut-out), he will claim against us. Actually this story is started before my post and I have to follow the game not to be claimed. I have suggested the contractor to rent a PMI; however he preferred to cut-out because he has a sub-contractor to force him and therefore cut-out has expense for sub-contractor!

By chuck meadows (***)

Date 02-16-2007 14:10

Henry hit it head-on. All of this could have probably been prevented had proper QC documentation been followed. Farshid, no disrespect intended, please, but if proper witnessing of the PQR and WPS was followed, this would not be a point of discussion. Sorry...

By Ted.Chungli (*)

Date 02-17-2007 08:06

Farshid!

I suggest - befor cutting of joints - ask your contractor to select worst joint which is highly magnetized. only cut it and send to lab. it's so cheaper than using PMI analyzer, then decide to cut all joints!!
You do not know what's inside your weld! reason is high Ferrite number or really high Fe in your weld!

Ted

By GRoberts (***)

Date 02-21-2007 20:47

A good PMI check that is quick and inexpensive is using copper-sufate solution. Put some on stainless steel and nothing happens. Put it on carbon steel, or contaminated stainless steel, and the copper precipitates out, leaving a copper colored surface. If you can pass that test, it would gurantee you don't have carbon steel filler metal.