Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic Welding Industry / Inspection & Qualification / PMI Grade 91
- - By js55 (*****) Date 11-10-2008 21:26
I'm runnin PMI"s on Grade 91 filler metal and gettin low Cr(~6.5% to 7.75% - 8% min). The low filler is 55# coils of SAW 3/32" wire. We also checked SMAW and GTAW and the readings were in compliance. We checked the unit against the cal block and its good.
I've been doing business with this filler supplier company for years, and they've been doing businees with their supplier even more years. Not saying people don't make mistakes but it isn't likely. They have never missed a single chemistry with any order I have ever placed.
Also, the Mo and Nb are dead nuts on. This reduction in Cr is almost 20% below min with some readings. And this is the poser. It just doesn't seem likely to hit Mo and Nb like that and miss Cr that much (7.8/7.9 maybe). Actual MTR's also have the Cr in compliance and the Mo and Nb on the MTR's matches pretty close my PMI readings. I believe I have a problem with my testing procedure, but what?
The wire's clean.
The machine calibrated.
Any suggestions?
Parent - By MBSims (****) Date 11-11-2008 05:54
I would have an independent chemistry run on a weld sample and on the bare wire to determine if you have a Cr measurement error in the PMI instrument or a problem with the MTR's.  Perhaps the PMI instrument has trouble measuring chromium in Cr-carbide structures vs bulk Cr content?  A stretch, but worth looking in to.  If the weld sample lab chemistry is in agreement with PMI, you must be losing Cr across the arc, possibly as an oxide.  Not that familiar with SAW, but perhaps a flux issue or voltage too high?
Parent - - By Jim Hughes (***) Date 11-11-2008 13:29
js55,
are you testing in the as welded or bulk wire? I can't tell from your post.

Jim
Parent - - By js55 (*****) Date 11-11-2008 14:46
Sorry bout that guys.
I'm testing the filler metal as it sits on the coil.
I'm fixin to flapper wheel the wire and test it again.
And then I'll run a button and test it again if need be. But the dilution from the base metal will influence the results unless I run several layers.
I don't want to have to ship this stuff back.
Parent - - By Stephan (***) Date 11-11-2008 15:57
Jeff,

please allow me to ask.

How are "PMI's" with filler materials being executed?

Wouldn't it be more reasonable to analyze the molten down pure weld metal or is it in general strictly prohibited to use a filler which is lying e.g. as a wire below the allowed alloying element's levels to be molten down and being then analyzed?

Thanks in advance for pointing me in the right direction.

Best regards,
Stephan
Parent - - By js55 (*****) Date 11-11-2008 18:56
Stephen,
We are using PMI as a receiving verfication method using a Niton XRF analyzer.
Any more complicated or time consuming methods would be my last choice just prior to shipping it back.
Parent - By Stephan (***) Date 11-12-2008 14:56
Jeff,

thanks for the response!

Best regards,
Stephan
Parent - - By Jim Hughes (***) Date 11-11-2008 16:22
js55,
What's the wire designation? What about the flux?  What about alloying elements in the flux? Are you introducing a flux cored wire? If there is alloying being added through the flux the best way to PMI would be in the as welded condition. It's been 10+ years since I had to deal with SAW so I hope some of this is still relevant.

Jim
Parent - - By js55 (*****) Date 11-11-2008 18:58
Jim,
Its a solid EB9 wire.
SFA 5.23 has chem ranges for the wire (EB9) and the resulting deposit (B9). We haven't even gotten to the deposit stage yet since I have not approved the material.
Parent - - By MBSims (****) Date 11-12-2008 06:06
The specified Cr range in SFA 5.23 is 8.0-10.0 for bare wire.  I have a hard time trusting the results from a PMI tester over lab methods.  We have several of the Niton analyzers and sometimes see some strange numbers due to surface contaminants, low battery level or minimal surface area of base metal in the window.  I would send a bare wire sample for lab analysis to confirm the PMI results unless the supplier is willing to swap it out (in which case someone else will get it and may not know if the Cr is too low).  Send a wire sample FedEx to Chicago Spectro and they will melt a button and give you results fast.
Parent - By js55 (*****) Date 11-12-2008 13:16
Marty,
Thanks for the info. I'm always looking for good lab alternatives. We received an alternative shipment  of wire in and tested it and it was in compliance. Though I agree with you I am always a little suspicious of XRF machines. though for general QC purposes they are quite reliable. So now I have good SAW wire, good SMAW rod, good GTAW wire, and real doubts about the first SAW shipment.
Parent - - By Stephan (***) Date 11-12-2008 15:12
Marty,

excellently described.

Is there actually a "minimum" or even "maximum" number of "XRF-shots" (I apologize for not knowing the correct term) per sample to even get an average value of a specific element content, or is there just "1" "shot" needed or allowed, respectively?

I mean, in worst case one could be tempted to measure just long enough to finally achieve the "correct" value.

Thanks in advance for a short reply and best regards,
Stephan
Parent - - By js55 (*****) Date 11-12-2008 19:24
Stephen,
I do not believe there is an industry standard.
It should be clearly defined in your PMI procedure, or in customer specifications.
But I wouldn't refine the procedure too closely. You could end up restricting yourself on compliant materials given the variability of the technology.
Parent - By Stephan (***) Date 11-13-2008 06:29
Once again thanks a lot Jeff!

Always learning from you...

Best,
Stephan
Parent - - By MBSims (****) Date 11-13-2008 03:20
As far as I am aware, there is no ASTM or other standard addressing XRF for PMI applications.  The instrument makers claim a degree of accuracy approaching lab instruments.  But for lack of a standard addressing calibration and performance of the measurements, I don't see how the results can be treated as anything but "informational".  One drawback is they can't read the elements with low atomic number, such as carbon, so they can't distinguish plain carbon steel from cast iron.  The number of shots should not matter if the surface is properly prepared and the window is in full contact with the surface being measured.  The longer the time held for a measurement, the more accurate the result should be.  I think Niton does offer a collimator for their instrument to reduce the window opening for individual weld beads.  We have seen errors from mill scale and surface oxides, so we usually grind to bright metal.
Parent - - By Stephan (***) Date 11-13-2008 06:34
Marty,

please see my reply to Jeff above!

Don't want to repeat myself hence, a heartfelt "Thanks" also to you for taking your precious time to shed some bright light on this.

Amongst others in particular the "(...) I don't see how the results can be treated as anything but "informational." statement was very... informational!

My best regards,
Stephan
Parent - - By 464238 (**) Date 11-14-2008 18:22
We generally weld "PMI Buttons" to insure the as welded condition. weld a Square area 1" x 1" about four or five layers in depth and grind the top layer off and then PMI. that has been the general acceptable method for pmi in our industry. Saying that I decided to do an experiment I got a tube of Bare tig wire ER80S-B8 and PMI'd just the wire and like you was not able to obtain the required Cr content. I believe that it is the coating on the wire

Readings

Mo .942
Cu 36.52
Co .395
FE 56.10
Mn .353
Cr 5.45

Using Niton XL3t Analyzer

my suggestion would be to PMI a sample in the as welded condition.
Parent - By js55 (*****) Date 11-14-2008 22:08
That's what I'm gonna do.
thanks
Parent - - By Stephan (***) Date 11-15-2008 11:33
464238,

very interesting.

That was the basic thought when I asked: "Wouldn't it be more reasonable to analyze the molten down pure weld metal...".

However, may I ask what the final Cr content of the wire you have described (Cr 5.45%) was finally, as been measured in the "as welded condition"?

Thanks and regards,
Stephan
Parent - By js55 (*****) Date 11-17-2008 13:38
Stephen,
I am now at my last option. Other than sending it back.
Up Topic Welding Industry / Inspection & Qualification / PMI Grade 91

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill