SA517B P11B PWHT soak time exceed 2

[faiswish]

Hi Frends,
we have new constructed pressure vessel made out of SA-517B which is P11B Group 4 shell thickness 0.370" after vessel been fabricated sent out for stress relieving as per engineers issued PWHT instruction sheet heating time suppose to be 350F/hr and soak time 30 minute at 1050-1100F. We get the vessel back after PWHT came along with PHWT chart which was heating time 150F/hr and soak at 1100F for 7 hours insted for 30 minutes?. Is that acceptable?.
Thank You!

 

[metengr]

SA 517 is a high strength quenched and tempered steel plate material. Since this is a nonconformance, you need to disposition the material as accept or reject based on PWHT treatment you reported. To disposition this material, you should have as a minimum, hardness testing done to evaluate ultimate tensile strength properties. There is also impact test requirements for SA 517 Gr B based on agreement between the purchaser and supplier for test temperature. You need to check on what was specified on the purchase order. Final disposition may come done to having a coupon simulate the PWHT and determine if the material is acceptable for mechanical properties and impact test results.

 

[faiswish]

Thank you metengr,
My openion ASME VIII div I UHT 56 set the minimum requirment for 30 min and it does not say for maximum that mean we can go for 7 hrs?. And as per CWB module 8, it never effects on grain boundaries, grain size and mechanical properties as long as not exceeding above its tempring temprature or austenite region and 1100 F is way below than austenite region. Am I righr?
thank you!

 

[metengr]

No, your interpretation is incorrect regarding PWHT time at temperature - which is one hour per inch of thickness at a minimum temperature. Extended PWHT time will affect both weld and base metal properties, which is why I said you need to determine the extent.

 

[weldstan]

You also need to check the PQR's supporting PWHT time at temperature if impact testing was required. See QW-407.2.

 

[racookpe1978]

NO!

In this case, you CANNOT just make a "judgement" or "evaluation". The heat was just about in spec, but the time at this thin a metal was much. much too long. Mentally, you are trying to "justify" the excess time as if it were 35 minutes instead of 30 minutes. Or the temperature actually peaked at 1175 for a few minutes instead of 1100 for 30 minutes.

7 hours is a loooooooooooooooong time that WILL affect crystal growth and crystal size and grain arrangement and every metal property: nozzles, welds, walls, supports, and end caps. Also, if the steel and nozzles and internals were not properly supported during the entire heat treat cycle, they have sagged and drooped because the red-hot metal has little strength against gravity's pull.

You MUST actually test the metal properties on at least three samples that have been equally heated: All three samples (test coupons) need to have representative welds and HAZ included. Be good to have nozzles in one of the samples, but probably not essential.

 

[faiswish]

We have deciede that we run PQR and PWHT for same amount of time including repair like 8hrs and see if it pass the mechanical testing, and cut off two specimens from the vessel including weld seams send out for impact and tensile testing. I will let ypu guys know either ways.

 

[EdStainless]

It is good to see someone doing the right thing.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[faiswish]

now we are cutting off specimens from the vessel as per thermocouple map most exposed area is top of the vessel that's what location we recommend to take off semples and send out for lab test; is that anywhere code recomendtion for cutt off semple locations for UHT vessel?

 

[metengr]

No, this is not addressed in the construction code. Why can't you obtain more SA 517 Grade B material, similar to that used in production, and simulate the PWHT versus using actual vessel material?

 

[faiswish]

unfortunatly production impact test coupons are not on the shelf no more they been thrown away, eventually we are requalifing new WPSs after finding out whether actual exposed material will pass the lab tests, that way is more economical rather doing WPS(s) first if WPSs pass but actual exposed matriel fails that mean we los time and money that we spend on requalifying WPSs.

 

[tbuelna]

One question I would be asking is why didn't your heat treat vendor submit an NCR and request MRB/corrective action from you before delivering the part? Based on what you describe your vendor did not perform the work as instructed, and even provided documentation confirming the fact. After you get things squared away with making a conforming replacement part for your customer, I'd have a serious talk with that heat treat vendor.

 

[faiswish]

As per vendor this is still acceptable to exceed at time teprature it does not effect on grain boundary or mechanical peroperties unless it reaches up to transformation temprature , so we will do lab testing and see if is fail than we can talk to vendor.

 

[racookpe1978]

Yeah, yeah, yeah.

Your "vender" is going to say anything. Can say anything. Might say anything.

Submit the samples. Trust the results. NOT the vender.

 

[faiswish]

we are going to send cut off specimens from the vessel I propose to do impact test for weld metal, HAZ and base metal, tensile test, I wonder do we need bend test on it as well? but on the other hand bend test will be better than ever anyway and also bend test is for to check the soundness of weld or ductility of material.
Or what other test would you recomend?
Thank you!

 

[metengr]

Impact and tensile testing are more than adequate. Ductility will be measured from tension specimens.

 

[racookpe1978]

If I understand you correctly, I agree with you that the most likely result of very long annealing times at high temperature will be "excessive" ductility and larger, very even grains through the metal. Probably very, very easy to bend.

Then again, once pressurized, the vessel would not explode. Merely "blow up" into a nice round sphere. Then it would nicely tear at the weakest point. 8<)

 

[faiswish]

i get the result after tensile test, all fractures were in the weld metal partial cup & cone type fracture, over 110 ksi but less than the required 115 ksi.

 

[racookpe1978]

Thank you for the feedback about the test results.