Ultrasonic Inspection of Inconel 625 Clad Vessel 1

[James01]

I have a Carbon steel vessel that has a Inconel 625 overlay. The Ultrasonic operator has reported wall loss in the vessel and the remaining thickness is that of the carbon steel section and my thinking is that there is an area of disbonding and that this should be easily determined by a good UT operator as if it is disbonding it should act like a lamellar type defect. But is it possible to use shear waves to verify if there is corrosion in the inconel?

Thanks in advance

 

[GenB]

I think that ut will give you the tickness overall and knowing the clad t, you end up w/the remains in c/s.
laminar is rare but not imposible, you are looking for discontinuance in the metal, good ut can pickitup.
ER

 

[quadswift]

It is unusual to have lost all the inconel to corrosion / erosion. There are usually some areas more prone to it than others.If you have lost it then I would think the medium will be through the c.steel much quicker.

Has he done scans with normal probes in several areas? If so has he received the same results in different areas?

Are there any original NDT reports / certs available or previous in-service inspection reports?

You could try using angle probes, say in the region of a nozzle and see if you can pick up the the internal corner of the branch. You may get a 'ripple' at the interface between the inconel and C. steel (assuming that there is anything left) you alredy of course know the thickness of carbon steel, you then could determine the total thickness.
Regards,
quadswift

 

[James01]

Quadswift
We had UT conducted over most of the underside of the vessel and this was the only area exhibiting these signals. The last internal visual was done in 1998 and I was very surprised to hear that there was possible erosion / corrosion, even though it is produced water that is in the vessel.

I will try what you suggest with the nozzles.

Thanks
Jim

 

[unclesyd]

Is this a clad plate or weld overlay?
How thickness is the respective layers?
What type UT is he using?
Is the Inconel for purity or corrosion problems in the vessel?

 

[James01]

unclesyd

Its weld overlay, C/s is 25mm & Inconel is 3mm. The tech is using conventional A-scan with an EPOCH III flaw detector, 5Meg 0deg & 4meg 10mm angle probes. The Inconel is for corrosion protection.

Jim

 

[PVRV]

1.0 Change your UT operator 4meg 10mm probe
should clearly indicate the disbondment.

*If so check your NDT records during weld overlay.

2.0 For verification purpose, 2meg (8 by 9)mm probe
should be used.
______________________________________________________

I would appreciate your feedback on the following items:

a) What is the service temperature.
b) What is the operating pressure.
c) I am assuming that the effluent (Produced) water has
high concentrations of:
1. H2S
2. Salinity

d) What material us being utilized for the pipping system.
e) Was Hastelloy C276 considered during design stage.

Cheers

 

[syamps]

I want to check internal clad thickness or the thinning of Base meaterial in Knuckle portion of a pressure vessel dishend.

I know i can use the UT Technique(T . meter) Find the total thickness and deduct the basematerial thickness and arrive at Clad thicknees.

Is there any other instrument that gives directly the clad thickness on dishend.

 

[bmoorthy]

Ye it does exist refer UCL of ASME Sec VIII Div 1, it addresses this issue.

Good Luck
B.Moorthy
moorthykar@yahoo.com

 

[James01]

PVRV

The operating temp is 35 degC, occasional excursions up to 65degC.

The operating pressure is 5barg.

The fluid is highly saline and has no significant H2S, the pipework is ST/ST.

I dont know what other materials were considered at design stage.

Cheers
Jim

 

[RNicastro]

James,

There is an instrument on the market which I personally have used on Clad Digesters and could read the thickness of the steel and the clad overlay separately. The unit is the TMI-150 B-Scan. It was produced and sold by Rohrback Cosasco Systems. Many Testing companies have this scope but only a few good technicians left who can do this inspection properly. The TMI-150 gives a typical B-Scan presentation and if set up right will give you the answer you need. I am no longer a contract inspector so I get no bonus for leading you to them just the satisfaction that you will get your answer.

 

[James01]

We opened the vessel during our currently ongoing shutdown and have found no wall loss inside, all overlay is intact.

I have had a different UT operator scan the area and he has come up with inclusions at or close to the bonding region but was getting a good back wall echo from the overlay. Therefore the original UT tech will not be invited back.

Thanks
Jim

 

[ndeguy]

Sorry to arrive to a thread so late but this is for those like me who come across this thread at a later date.

The original operator could not have received proper training in testing bonded materials of different acoustic impedances. And possibly the second technician was mistaken in his interpretation of the signals he was receiving as inclusions.

Acoustic impedance equals the product of the material's density and the wave velocity in that material. where the wave velocity is a function of other material properties, namely the Young's modulus and Poisson's ratio.

Equations give the ratio of transmitted to reflected energy at an acoustic boundary. Thus at a water/steel interface 12% of the incident energy is reflected and 88% transmitted. Different material interface combinations will preoduce different transmitted to reflected ratios.
With a 0 degree compression wave probe transmitted through one material to be incident at 90 degrees to the interface between the two materials there will always be some sound energy from this interface returned to the probe even with good bonding. With a cladding material with poor sound transmission properties then the echo received from the surface of the cladding when testing from the opposite substrate surface will be of very low amplitude. Therefore even with a high transmitted to reflected ratio so much gain may be necessary to obtain any response from the cladding surface that a relatively high amplitude response may be obtained from the acoutsic impedance material interface. Easily misinterpreted as a disbond or inclusions. In case of doubt I would suggest having a test piece made by the same process into which is introduced an intentional disbond of known location and dimensions. this amy be used as a reference for the inspection of the parts to be tested. Or hire a tech who has evidence of training in this type of work.

 

[JR97]

Another late comer to the thread.

I am a UT Level III. Our shop has made many Inconel Clad vessels. I wouldn't be too hard on your first inspector. While I agree with the formulas and numbers the NDEGUY quotes, those formulas tend to break down because they do nto consider the anisotropic nature of Inconel. If you run the numbers, you should get the same reflection (discounting beam attenuation) whether you test from the Inconel or from the Carbon side. In real world - it isn't even close. If you try to get an interface signal from the Inconel side, you won't get anything back. If you try from the carbon side you will get a big signal.

If you are still listening James01, when the original inspection was performed, was the unit filled with water? If it was, the inspector was set-up. The majority of the little bit of sound that made it into the Inconel probably coupled straight into the water (again not so per the formulas - but definitely so in the real world) and he didn't get a strong enough signal to read, or the signal was weak and could not be resolved from the clad interface signal.

Of course, maybe he didn't know what he was doing.

 

[James01]

JR97

The tech had reported a local area of thinning ie 3.0mm less than the rest of the vessel, that is what lead us to believe that there was either an area of disbondment or we had lost the linning in that location.

The vessel was full of water during the original inspection and also the re-inspection we had done was during normal operation with the vessel full.

The original operator has been asked for his comments through his company but has failed to respond and has subsequently left to work overseas.

 

[ndeguy]

Interesting posts James 01 and JR 97, especially as they are straight from the shop.

When I tested the components they were flat plates in the workshop, so I will be particularly wary now if asked to scan a full clad vessel.

i am just wondering now after JR97's comments why the second inspection showed a strong backwall from the Inconel overlay when it was full of product. I can appreciate that the Inconel grain structure may cause scatter but perhaps the second tech chose a lower frequency probe than the first which reduced that. Thus more sound energy arrived at the Inconel/water interface and sufficent was reflected to enable identification of the backwall echo. Just goes to show what judicious equipment selection can do. Hats off to the second guy.