Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic Welding Industry / Inspection & Qualification / UT data taken during on-stream vs shutdown - the difference?
- - By khaiis Date 01-14-2015 11:06
Good day all.. i'm from oil & gas company.
I've an issue to resolve with regard to ut data.
Is there any difference between ut data taken during on-stream vs shutdown?
For both equipment operated at ambient & above 65 deg celcius.. is there any paper / study done for this?
Parent - By Superflux (****) Date 01-14-2015 13:04
Calibrate your equipment at 65°C on the EXACT material and all should be good.
Parent - By ctacker (****) Date 01-15-2015 02:38
The sound velocity varies with the temperature of the test object. That can cause appreciable errors in the measurements if the UT machine was calibrated with a reference block of one temperature and the tested object are a different temperature. Some of the newer flaw detectors have the correction factor built in, or it's possible Henry can find some printed tables online.
A general note is, as the temperature of the test object increases, the sound velocity slightly decreases.
.
Parent - By thirdeye (***) Date 01-23-2015 21:09
I have witnessed UT screen displays of cracks in a shaft change in amplitude (higher) when the shaft is under load, and then watched the amplitude fall when the load is removed.  And as mentioned heat can have an effect as well. Years ago doing plant thickness baseline surveys, it was common to leave step-wedge calibration blocks on various piping runs for the course of the project so they would be the same temperature as the pipe itself. We also used magnetic thermometers to record the temperature of the pipe for the report.
Parent - By 46.00 (****) Date 01-24-2015 15:16
yes
- - By 803056 (*****) Date 01-15-2015 01:19
You bring up an interesting subject. Recently I finished a project where I had to test a valve  body with a history of developing fatigue cracks after several months of service. Sure enough, I tested the valve and there was a crack right where the mechanic told me the cracks typically develop. The valve was under pressure and the crack was clearly defines, a nice signal on the display. Life was fine.

The contractor pulled the valve out. It was no easy task; it weighted a couple of tons and had to be moved laterally and they up and out. Long story short, they have me retest it once it was pulled out and the new one installed. No crack! Crank up the gain, no crack!, cranked up the gain to over 100 db, there amongst the back ground you could just make out the signal, but you would miss it if you didn't know where to find it.

An engineer suggested heating the valve with a torch. Differential heating, damn, there was the crack, clear as day, proud as a new feather. Just goes to show you, your never too old to learn something new.

Best regards - Al
Parent - - By welderbrent (*****) Date 01-15-2015 01:25
So, was it leaking or could it have still been usable? 

If this is an ongoing problem, does anyone know what is causing the cracking? 

I presume that is not a cheap valve.

He Is In Control, Have a Great Day,  Brent
Parent - - By 803056 (*****) Date 01-15-2015 01:45
The idea is to detect the crack during a scheduled shut down rather than experience an unscheduled shut down at the tune of 1 1/4 million dollars a day of loss revenue. The valve operates at 6500 psi. When it fails, all hell breaks loose.

This was the first time anyone had suggested using UT to check for cracks. It was the mechanic that bought up the subject with, "Can you check a valve for cracks with that machine of yours?" referring to the UT machine. The valve body was 29 inches thick.

We suggested a redesign to eliminate sharp radii and sharp corners to reduce the probability fatigue. The engineering firm that I partner with did the redesign and they are fabricating a new valve so it will be ready when this one (the new one with an old design) fails.

I guess the point is, when components are in service, cracks tend to open and are more easily detected. However, the crack might be much more difficult to find once the pressure is reduced, the temperature settles to ambient, etc. The results can be very different!

Best regards - Al
Parent - - By welderbrent (*****) Date 01-15-2015 11:57
Looks like there was an employee really thinking about what was going on.  Everybody counts for a system of 'team work' that makes a difference.  Wish they all thought like that. 

I assumed the main point of your post in this thread, just had a couple of questions for my own information and curiosity. 

Sounds like a good fix has been come up with as well.  Good work all around.

He Is In Control, Have a Great Day,  Brent
Parent - By 803056 (*****) Date 01-15-2015 13:58 Edited 01-15-2015 15:39
You are on target. The mechanic was watching me scan the various components and started asking questions about the valve body and whether the UT was applicable for that type of application. Without asking permission of the higher-ups, he asked me to take a look at the valve during a lull in the schedule. He took me to the bowels of the plant and we tested it. He was really impressed at what we saw on the display when I showed him the crack, mapped out the extent (area) and could tell him exactly how deep it was. He took the information to his supervisor and told him, "Al says the valves only got a couple of weeks to go before it lets go. We have to change it now, while we have the chance!"

OK, so he was a little excited and spilled the beans a little premature, but he had 90% of the information correct.

After some discussion and assurance from me that it was indeed cracked and yes, I agreed that based on the time it had been installed and the apparent growth rate, failure should occur in about 4 weeks time. I took the supervisor down into the hole with me, repeated the test, at which time they agreed it was better to spend the 60 man-hours to pull the valve, than to wait and have another unscheduled shut down.

This company has a progressive management style where they encourage their people to take ownership of the equipment they operate and maintain. I have to go back in about a month to do more inspections to detect any potential problems. They were really happy with the results of the testing I was doing for them. We found several components with early to midlife fatigue cracks. We were able to predict the remaining life and avoid the costly unscheduled shut downs. That is where experience and working together with some really great mechanical engineers can make a huge difference in the financial well being of the client.

Best regards - Al
Parent - - By CWI555 (*****) Date 01-26-2015 16:01
Al,

The changing conditions imply an on off condition. The thickness you report would be prone to shrinkage of the CA/CB type. If that shrinkage were close enough to the surface, it could feasibly produce the results mentioned. From a UT standpoint, I have witnessed CA take on crack characteristics under service, but nothing when sitting. As you aptly noted, the sharp radii/corners would be the likely failure points. However, the root cause of the stress may be the other direction. It may be the sharp corners interacting with what would otherwise be acceptable shrink/stresses by the valve body manufacture if the valve surfaces were devoid of sharp radii and corners as you have mentioned.

I assume cast as if it were milled with that T it would be phenomenally expensive to the point of nearly/actually prohibitively so.

My .02
Parent - By 803056 (*****) Date 01-27-2015 16:49
The valve body was actually a forging.

Al
Up Topic Welding Industry / Inspection & Qualification / UT data taken during on-stream vs shutdown - the difference?

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill