Transfering of PQR weldpass average or single values to WPS 2

[kjell123898]

We have performed a PQR of a 45mm plate in S355 material and weld position 3G. Welding method is FCAW. The number of single welding layer is 50 pcs achieved during the PQR welding.
Please may someone state which of the calculation methods are most appropriate when observed heat input values to be transferred to the applicable WPS.
Question: Shall we use an average heat input value based on all 50 passes (+/- 10% min/max range)? or shall be rather use the minimum and maximum heat input value from the single pass to determine the heat input range in the WPS?
This matter is the subject of extensive discussions and I feel it will be like that one strives for a wps with greater validity than technical advisable....

Kjell Gr°nvold

 

[gtaw]

When one qualifies a WPS using either a manual or semi-automatic process, there is bound to be some variation in travel speed and the resulting heat input. Codes are not explicit with regards to how the contractor is to handle these minor variations. I tabulate the welding parameters during the qualification process by recording the parameters, including preheat, interpass temperature, and the thickness of each layer. The data is processed with the Statisitcal Analysis tool that is an add-on found in Excel. I use the median values plus one standard deviation as my ranges. I haven't had a problem because my ranges are typically within the ranges permitted by the code.

Best regards - Al

 

[DekDee]

kjell,
From the 50 runs take the lowest heat input and highest heat input and that is your "PQR range".
Then, dependant on what your applicable code allows you can apply a +/- to those figures and that will give you your acceptable range for your WPS,
Cheers,
DD

 

[gtaw]

One reason I use the statistical analysis tool is it provides a better weighted average of the heat input and welding parameters. I guess it is a case of what one feels comfortable with.

Best regards - Al

 

[DVWE]

kjell,

How much variation do you have between the maximum heat input value, and the minimum heat input value?

Also, to what code are you qualifying this WPS?

 

[kjell123898]

hi
the minimum heat input value is 0,57 kJ/mm (measured of only one pass of 53 passes) and max was 1,55 (measured of only one pass of 53 passes)
let me underline that this PQR welding is carried out by one of our subcontracor and I'm disagree in this way to establish the final welding range on
The weld qualification code is ISO 15614-1
Kjell Gr°nvold

 

[racookpe1978]

I'm wondering why 50 passes were needed on a sample joint only 45 mm thick base metal.

That's many, many more then I've ever seen for that thin a production weld - especially with FCAW, which lays down a lot of weld metal quickly.
Even GTAW would need many fewer passes. And your strength tests will reflect that number of passes and the heat changes.

 

[gtaw]

You may have a point racookpe. A typical 25 mm plate requires about 40 beads using low hydrogen electrodes using SMAW. However, the number of beads required using small diameter electrode and a stringer bead technique is not unreasonable for a groove weld that is 45 mm thick plate.

I would be hesitant to accept a PQR based on the welding parameters of two or three "cherry picked" beads or even 3 or four beads checked at random as the basis of heat input for either a manual or semi-automatic process. Either method will have considerable variations in travel speed as the welder fills the groove. I believe Kjell's posted values vary by 300% using the lowest value as the base point. Those individual weld beads do not necessarily represent typical values, statistically they could be outliers. That's why I record the parameters of each weld bead as they are deposited. The statistical analysis provides a more realistic representation of the welding conditions used by the welder and a better representation of the results. The approach may be conservative and results in tighter control of the parameters listed by the WPS and tighter than that permitted by the applicable welding standard, but in the end it provides more consistent results. I like consistency.

I believe a test witness should be present while the test assembly is being welded to record the parameters used for each bead deposited. Some will argue there is no requirement for someone to witness every weld bead deposited. No argument from me. Most codes have no such stipulation. I had a project where I was required to audit the welding documentation. I noted there was no witness recorded by the PQR. The welding engineer in charge said there is no requirement that a test witness be present. I said. "I agree, but who recorded the welding parameters listed by the PQR?"

"The welder recorded the information while he welded the coupon." was the reply.

I said, "He must be an excellent welder. I would like to meet him."

"That's not a problem. He's in the shop now. He qualifies all of our WPSs because he is our best welder." replied the engineer.

We went out to the shop floor and located the welder. The engineer introduced me to the welder, hand shakes all around, and we got down to business. I showed the PQR to the welder and asked, "Did you record these welding parameters when you welded the test coupon?"

His response was exactly what I expected, "Buddy, you don't know a damn thing about welding do you? You've probably never welded in your life!"

"Why would you say that?" I replied.

"Because anyone that knows anything about welding knows I cannot concentrate on laying a good weld bead and watch the meters on the machine at the same time." was he response to my question. He continued, "How would I keep track of my welding speed?" You could hear in his real thoughts in his voice, he thought I was a complete idiot. How could I be so stupid and know so little about welding?

The welding engineer's face turned dead white. You could see in his eyes he was thinking, "Shut up, stop talking, you're killing me. Pull the knives out of my back!" He knew the jig was up. All the WPSs, the PQRs, and any welder tested to those procedures were voided by the welder's last statement. All the product welded to date was noncompliant with the code and the contract. The mule dung had hit the fan and no one was spared. All the welding procedures and the welders had to be requalified and all the completed and semi completed product were scrapped. A costly miscalculation by the engineer to say the least.



Best regards - Al

 

[DVWE]

kjell,

I am not familiar with the ISO code you listed.

I will say however, that the only way to establish a minimum and a maximum heat input range is to have recorded EVERY weld pass, not just 2 out of 53. Then, you could either establish the range based on that, or the way gtaw has suggested above.

 

[DekDee]

Agree totally with Al & DVWE.

What is interesting is the major differences in heat inputs.
Without knowing full details I can only surmise.
Run 1 may be an open root ?
Runs 47 to 53 may be capping runs where the heat is turned down a little bit.
However, Runs 2 to 46 should be exactly the same.
All runs should be recorded on the PQR as noted above.
However, I am a little more lenient and would accept this on the PQR if all parameters were unchanged between multiple runs.
Run 1 ...........
Runs 2 to 46 ..................
Runs 47 to 53 ...............

I (and this is only a personal opinion) would accept 0.57 kj/mm (+/- whatever the code allows) for the root run (if that is what it was).
I (and this is only a personal opinion) would accept 1.55 kj/mm (+/- whatever the code allows) for all remaining runs.

Cheers,
DD

 

[kjell123898]

hi all in the loop,
yes, I can understand that all of you wonder about why 53 passes are needed in this plate connection where the wt is only 45mm but done is done. Please find copy of WPS and PQR sheet where the valuse are listed for your info attached.
I will request them to calculate an average HI value based on these single values instead of min/max and rather use this average value as subject for final calculation where -/+ 25% is applicable for max./min. heat input tolerances.
Thanks for all valuable feedback from all of you and an expecially thanks to you "gtaw" for a real and entirely appropriate/correct description of the real welding world many of us experience daily in our work,,,again thanks a lot,,,, I will use your description at all appropriate occasions in the future.
thanks for your time

Kjell Gr°nvold

 

[SJones]

Sometimes it pays to have something like this on the job to cut the arguing down a little (it won't be eliminated):

http://parteklab.com/arc-welding-data-logging-and-monitoring-system/

Since there are both Charpy and hardness considerations with ISO 15614-1, there's not too much scope for wandering too far with heat input (more strictly, arc energy) either up or down. Naturally, the welding fabricator tends to want the "Dek Dee initial approach (subsequently modified to tighten things):" largest + 10%, smallest - 10% and apply the results anywhere in the weld; the client wants the "gtaw version (enhanced)." Something in between has to be negotiated.

Steve Jones
Corrosion Management Consultant

http://www.linkedin.com/pub/8/83b/b04

All answers are personal opinions only and are in no way connected with any employer.

 

[gtaw]

I have to admit that I am not familiar with the nomenclature used to specify the diameter, etc. I'm limited to American English.

The listed data appears to be complete. To verify I understand what was done; you used GMAW short circuiting transfer for the root and second layer followed by FCAW.

I would disregard the root and second layer deposited using GMAW for the simple fact that the CVN samples probably didn't include any metal from that region and it is not representative of the remaining deposited weld metal.

If GMAW using short circuiting mode is required for production, I would qualify it separately to ensure CVN samples extracted from the weld metal and HAZ is representative.

Excluding the weld layers deposited with GMAW-S should eliminate the outliers for the heat input analysis and provide more consistency in the calculated heat input.

My good friend DekDee has a leg up on working with ISO. They do differ in subtle ways when compared to AWS and ASME standards. He knows his stuff.


Best regards - Al

 

[sayee1]

Note that ISO definition of heat input is different from ASME, with ASME HI defined as arc energy by ISO. I am going by the ADME definition here same as everyone else, assuming that is the language you use. Irrespective of this I am amazed that this question gets asked so many times. A WPS is a work instruction - to be primarily followed by whom - the welder! If a welder cannot record parameters during welding, how the heck can he control it during welding? I therefore think personally that HI measurement "maybe" good for approve sure qualification, if and only if a recording device such as AMV Weldcheck is used to record instantaneous parameters - when you start going towards some of the exotic high frequency processes or fancy electrical controls such as on power wave 450, even this isn't appropriate. Of course here I am talking about manual and semi auto and not mechanised or automatic welding! Even if this is done can you link the (questionable) impact toughness associated with every pass? That is why the run out length method is better but even that is impossible for a welder yo control during welding. In all my life working with 1000s of welders, and some were best in industry, I know only one who could really control Keats input during welding, but that was only when I wanted it. I would rather qualify a par with all high heat input parameters as high as a welder can weld and one with as low heat as low as possible for all passes. I would then test the high HI PQR for toughness and the low heat input PQR for hardness. I have excel spreadsheets that I ran for over 50PQRS measuring heat inputs in different ways and quantifying WOS HI through different methods including ones discussed to check which gives most realistic ranges when applied - will search for it and share later and it is pretty self explanatory.

 

[DekDee]

Al,
Thank you for the kind words but I do not feel they are appropriate - when someone mentions ISO I run as fast as I can in the opposite direction. LOL !!

Kjell,
Please let me clear one thing up - you do not use the average heat input - you use minimum and maximum.
EN ISO 15614-1
8.4.8 Heat Input
If welding procedure tests have been performed at both a high and a low heat input level, then all intermediate heat
inputs are also qualified.

Whether your lowest and highest heat inputs are correct is a totally different story.

Now for the range of these values - up 25% for impact requirements and down 25% for hardness requirements.
If you do not require either for your application then the code gives no requirements for range of heat inputs (similar to ASME IX where heat input is only regulated for impact tested applications).

Now for your PQR - very, very strange.
1.2 mm vertical up FCAW with a 60-70 degree included angle on 45 mm plate would require approx. 15 - 20 stringer runs to fill.
This PQR has an included angle of only 50 degrees and yet it has taken 46 runs to fill (have they been grinding out more than they are putting in) ?
With manual FCAW there will always be minor differences in travel speeds but if they are all stringers (with no weaving) the differences should be minimal.
Run 12 took 199 seconds to complete (HI = 0.81 kj/mm) and Run 13 took 388 seconds to complete (HI = 1.55 kj/mm)- nearly double the time ?

On a side note - there is no record of interpass temperatures on the PQR ?

8.4.10 Interpass temperature
The upper limit of the qualification is the highest interpass temperature reached in the welding procedure test.

The contractor has stated less than 250 degrees on the PQR - what if the highest interpass temp reached on the test was only 180 degrees ?
How do you know what the upper limit is as it was not recorded ?
How do you know if the lower limit (preheat temp) was exceeded if it was not recorded ?

Good luck,
Cheers,
DD

 

[kjell123898]

hi again all,
sayee1 (Materials)states:
"I would rather qualify a par with all high heat input parameters as high as a welder can weld and one with as low heat as low as possible for all passes. I would then test the high HI PQR for toughness and the low heat input PQR for hardness"
yes I'm agree and this is in line with ISO 15614-1 requirements and probably most appropriate way to ensure the quality max / min welding range regarding to ensure the overal mechanical properties.
My current case is that I receive a number of welding procedures from a variety of welding suppliers (vendors) with different perceptions of standards and how these should be interpreted correctly in order to do subsequent welding qualifcations and mind that this in the context of very limited time constraints so I have to choose on the basis of common sense sometimes.
DekDee (Petroleum)states
"Whether your lowest and highest heat inputs are correct is a totally different story" yes,it is precisely this which is the core of the whole matter....... I'm disagree with our sub supplier (those who made PQR CEM/32/2015) because of they will go for only one single pass with the lowest energy (0,57kj/mm) represents the lowest energy range and the highest single value (1,55kj/mm)represents the highest energy range??? My concern is how can they (my subsupplier) guarantee that the charpy's are performed exactly of the pass with highest energy and how can they guarantee that the hardness is performed exactly of the weld pass with the lowest energy. Although we should be so lucky to perform mechanical testing on the welding layer (single highest or lowest) the final test result will not be representative for the actual section.
DekDee (Petroleum)states
Now for your PQR - very, very strange. Yes,agree and I will have your notices in my mind during coming discussions
thanks again for your time
BR
Kjell Grønvold
Weld Eng/NDE Level III

Kjell Gr°nvold

 

[gtaw]

So, my friend, back to ground zero.

If the fabricator can demonstrate the two layers of weld metal deposited at the highest heat input was included in the CVN specimen, I could live with the test results. If he cannot document where the specimen was extracted, and that it does include the area in question, I would reject his premise his WPS is properly qualified.

There is a presumption the "welding engineer" understands some basic concepts of welding metallurgy. If he is weak, it is his employer's concern, not yours.

Your concern is whether the WPS will produce the desired results and that it meets the requirements of the welding standard.. If not, you are not doing your job.


Best regards - Al

 

[DekDee]

I thought the highest heat input noted was either a typo or the welder must have fallen asleep halfway through the run ???

 

[SJones]

Is it not the intent behind the ISO 15614-1 wording to require the construction of two test welds: one with all runs at a low heat input (arc energy), and one with all runs at a high heat input (arc energy), thus setting the lower and higher bounds (as per sayee1)? This would be something that would have to be agreed upfront between the parties and monitored closely during test welding.

kjell's sub-supplier is attempting the typical contractor's "trick" when the mechanics of defining a WPS from the PQR have not been specified and agreed beforehand: slip a quick pass in somewhere, and a slow pass somewhere else and then we'll have the heat input cake and eat it with the biggest possible range for the whole weld so that we maximise the chances of getting through all the production welding without any rework.

Unfortunately, without a specification, or other contractual instrument, to nail it down, then there will be tears from the resultant correspondence fisticuffs.



Steve Jones
Corrosion Management Consultant

http://www.linkedin.com/pub/8/83b/b04

All answers are personal opinions only and are in no way connected with any employer.

 

[DVWE]

Is it not the intent behind the ISO 15614-1 wording to require the construction of two test welds: one with all runs at a low heat input (arc energy), and one with all runs at a high heat input (arc energy), thus setting the lower and higher bounds (as per sayee1)? This would be something that would have to be agreed upfront between the parties and monitored closely during test welding.

I would tend to agree with that, not knowing much about the code the WPS was qualified to. I only say that because ASME is similar, and that there is an ASME code interpretation that states the same.