Smart questions
Smart answers
Smart people
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Member Login




Remember Me
Forgot Password?
Join Us!

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips now!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

Join Eng-Tips
*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Donate Today!

Do you enjoy these
technical forums?
Donate Today! Click Here

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.
Jobs from Indeed

Link To This Forum!

Partner Button
Add Stickiness To Your Site By Linking To This Professionally Managed Technical Forum.
Just copy and paste the
code below into your site.

qcrobert (Industrial) (OP)
25 Jan 11 12:56
Am testing for leaks using fluorescent dye penetrant (spray can application) on exterior of 3/16" thk. 304L matl 10' dia x 18' L vessel and using black light to reveal discontinuities from interior.
 
In the past the Magnaflux Zyglo penetrant would reveal even minor, minute cracks within 4 hrs.

I checked the vessel skin and it reads 48-50 deg F.

Tried warming up suspected areas (T-joints intersections) to approx 80 deg F and performed PT, same results, does not reveal discontinuities till dwell times of 2-5 days!

Appreciate any advice or suggestions from those with experience using fluorescent dye penetrant....

Have changed batch number for PT and still same results.
 
Would using the white developer work with fluorescent dye penentrant?  If so, would it reveal cracks faster?

Have Paint Dept wanting vessel for glass bead blasting but cannot release it till I am assured of integrity.
 
blacksmith37 (Materials)
25 Jan 11 13:08
As I read your question; Cracks are acceptable on the outer surface, but through-wall cracks are rejected ? I would suggest defining some size of OD crack as unacceptable.
I never tried white developer with fluorescent penetrant, but I doubt it will work well.
qcrobert (Industrial) (OP)
25 Jan 11 13:54
No cracks are acceptable regardless of size.

The cracks are thru wall cracks in that the dye penetrant is applied from the outside of all weld seams and a black light will reveal if the fluorescent dye has penetrated to the interior side.  This is a form of leak detection.
Helpful Member!  metengr (Materials)
25 Jan 11 14:11
I believe you may be barking up the wrong tree. The time factor involved with crack detection is based on the type of discontinuities. In other words, there could be various through-wall defects that are of such a geometry where penetrant (liquid or fluorescent) requires time to wick its way through a possibly tortuous path.

Have you looked at an elevated pressure test or using helium leak detection methods to check for leaks first?
 
qcrobert (Industrial) (OP)
25 Jan 11 17:44
I agree that there are many types of discontinuities involved in crack detection.

The ones I normally find are porosity or linear indications (cracks) from incomplete fusion when the welder did not back grind into sound metal. Btw these are PT with red penetrant & white developer.

Pressure testing cannot be performed as some of these vessels are open to atmosphere and even in a closed vessel, the cost of helium would be prohibited.

From what I have learned from others and may be my problem is we are not cleaning the heat tint and debris from the GMAW process before applying the fluorescent dye penetrant to the exterior side.  We have to be careful as not to produce a cosmetically unattractive appearance to the finished surface as to Client's specs.  The inside is thoroughly machine wire wheel brushed to remove all heat tint oxidation.

Further research is in agreement with Blacksmith as the while developer is not normally used with Fluorescent green dye but heck I may try anything at this point.

I have over 30 yrs experience in welding inspection and hold an AWS CWI and SNT-TC-1A Level II in MT & PT but have never experienced such lengthy dwell times as these especially in 3/16" thk matl.

Looking forward to further comments from members and thanks to all for your time here.
 
Kiwi2671 (Structural)
25 Jan 11 19:20
gcrobert,
The mixing of fluorescent and colour contrast developers is not permitted as per ASME V.
Can you not perform fluorescent PT from both interior and exterior surfaces ?
If it is a through wall crack the indication should be visible on both surfaces.
As metengr has stated the geometry of a particular through wall crack could take days to penetrate.
Probably, more importantly how / why are your welders getting through wall cracks ?
Regards,
Kiwi
 
metengr (Materials)
25 Jan 11 20:02
What about Ultrasonic testing versus surface NDT???? I would consider phased array TOFD coupled with a surface examination and be done with it.
vanmorrison (Nuclear)
25 Jan 11 21:40
Hello qcrobert,

You have said that it is taking 2-5 days for bleed out, could it be possible that the penetrant might be drying on you?  

Is this an inspection requirement to a particular Code or standard requiring an approved technique or specification?

There may be a penetrant that is more sensitive to very tight indications.  Your penetrant supplier may be able to recommend a type more suitable to your application.  You could also ask if there were any changes to the formulations that might account for the longer bleed out times.

Since your tests are taking 2-5 days I would also consider volumetric if possible as Mentgar has suggested. It might be more efficient for you.  The fabrication process is a concern if through wall cracking is common and anticipated.   

BTW, good for you and not buckling to production pressures!

Regards,
Van
 
unclesyd (Materials)
25 Jan 11 23:25
Right off your dwell time is too long.  the longest I've used is about 45 min. using water washable Zyglo on the majority of our equipment.  Looking for gross defects all you need to do
is wash with water, dry and read.  You can  use a developer anytime you need higher sensitivity.   
One tip is to use a sanding disk or flapper wheel instead of a wire brush.   This is especially good method in cool whether.  You might have to check small sections.  I use this method using red dye in vessels that have been in service, by sanding or grinding the area of interest and quickly spraying developer and looking for wetted areas.

Her is the Magnaflux site where you can check penetrant vs the required developer.  Magnaflux did have some training aids available.  

http://www.magnaflux.com/ProductOverview/Penetrant/tabid/96/Default.aspx
 
eyec (Industrial)
26 Jan 11 7:36
seems to me that the structural weld integrity you seem to be shooting for needs to be addressed not with longer dwell times but with first removing any linear indications on the external side weld and repeating PT until you do not see any linear indications then do the same for the internal weld surfaces.

i have never heard of PT practices with dwell times you are using.
EdStainless (Materials)
26 Jan 11 14:05
First there is no such thing as 'no indications', there has to be an acceptance criteria.  If someone says "none" then pick the correct penetrant (viscosity and drying time" and an appropriate reading time and that is your test.
Waiting longer is a different test.  Stick with the one that is specified.

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

stanweld (Materials)
26 Jan 11 14:50
Have you tried a light oil or fuel oil? They are commonly used for leak detection in tanks and duct work. Most are also detectable under UV light.   

unclesyd (Materials)
26 Jan 11 15:43
stanweld
You are dating yourself when it comes o kerosene and whiting. The article says it lasted until 1920, but my it went into the 1940"s.  You can use just about anything depending what your are looking for.
My favorite is Sok or Kroll penetrating oil and chalk line chalk. Another very good penetrate is Oil of Wintergreen.

I have the original handbook by Doane and Betz that was used in the fifties to teach P.I.  

http://www.absoluteastronomy.com/topics/Nondestructive_testing

http://www.shamrockspecialties.net/SOK.htm

http://www.kanolabs.com/google/
Duwe6 (Industrial)
26 Jan 11 16:25
You REALLY need to switch to Bubble Testing in a tank of water.  A test gas with 5% Hydrogen or 5-10% Helium will be fast and accurate.  The hydrogen mix is fairly cheap, and nonflamable at that low a percentage.  The problem with Through-Testing with PT is that there are defects that will not bleed through in weeks, say nothing of days.  Shrink porosity in castings is one that cannot be Through-tested with dye.  The actual surface path will be enormous, and you will never be able to apply enough dye from one side to fully wet all the surface of the shrink cavity [tree-shaped] and appear on the far side of the part.

For items that cannot be made to hold 10-20 psi/1-1.5BAR, you can use a vacuum box and soap solution.

If the problem is in the welds, check for drafty conditions blowing away the shield gas during welding, cleanliness of surfaces to be welded, and proper weld filler material.  If these parameters are OK, hire some new welders.  304 stainless is very easy to weld.
qcrobert (Industrial) (OP)
26 Jan 11 17:28
I agree that the problem is in the welds and this I assure you is being addressed.

10'dia x 18'L is fabricated in a closed shop, 304L matl is new, shiny & clean and GMAW with .035" ER308L wire & Tri Mix gas.

This is an open recycle vessel of complex shape (inner 36" pipe, baffles betw pipe & tank ID, slopped floor, etc) at 297 ft of weld that would be difficult to vacuum box test.

We have fabricated these in the past and PT would reveal "thru-wall" cracks or porosity (1-3 indications per tank) within 1-3 hrs.  Penetrant would remain on tank for a up to a week before progressing to Paint (clean-up) Dept depending upon their work load.  During that time I would discover no other indications so this is out of the ordinary.

The testing is not a client requirement but an assurance test by me.

No other leaks have been discovered on this vessel so it is being released.

In the future, will focus on more thorough cleaning of weld seams both interior & exterior along with application of warmer temps to metal.

Everyone's input has helped and will be utilized in the future.  I appreciate your time & input concerning this matter.
qcrobert (Industrial) (OP)
26 Jan 11 18:05
Kiwi2671 (Structural)

Not that I doubt you, but could you direct me to the paragraph that states that it is not permissible to mix  fluorescent dye penetrant with color contrast developer in Sec V?

According to the directions on Magnaflux Spotcheck SKD-S2 Developer, "For use with visible or fluorescent penetrant materials, after penetrant removal."

Thanks again,
QCRobert
Kiwi2671 (Structural)
26 Jan 11 19:21

gcrobert,
My apologies. I mis-interpreted the clause below. It relates to the mixing of penetrants, not penetrants and developers.

T-654 Technique Restrictions
Fluorescent penetrant examination shall not follow a
color contrast penetrant examination. Intermixing of penetrant materials from different families or different manufacturers is not permitted.
Regards,
Kiwi
unclesyd (Materials)
27 Jan 11 20:17
There is absolutely no reason to have leaks in a stainless tank to leak. It can be inspected as it's being built.  All our stainless steel tanks are 100% penetration where they are welded one side then back gouged ground and the tankers do the PT as the tank is welded out.  All our SS tanks are 100% RT, no defects allowed.  We might discuss an indication but we win every time.

The tanks are all 304L 0.015 max C.  The welding process is SMAW with the welders having a pedigree..
The tanks are all over 30' diameter and a height of 30"40",except for two 2,000,000 gallon tanks for 65% HNO3. Anywhere between 3 to 5 yeas we go in and re weld the inside pass on all welds due to intergranular corrosion.

A snippet of  information on rewelding these tanks you can fuse the affected areas with a TIG rig.  This fused metal last as long as new weld metal.
r6155 (Mechanical)
28 Jan 11 7:15
Liquid penetrant testing is not a leak testing.
Please, see ASME V and definitions in ASTM E1316.
Regards

r6155
qcrobert (Industrial) (OP)
28 Jan 11 10:36
Unclesyd,
PYP design is speced by Client, most are CJP but these two recycle tanks were not.

I have recommended all tanks to be CJP but I do not control the Client's Engineering Dept.

I myself perform the majority of repairs and I do use GTAW process.

r6155,
I agree that PT is not a leak test as defined in V.  Perhaps I should use the wording, fluorescent dye penetrant is applied to one side of the welded seam and after dwell time the opposite side is inspected with an ultraviolet light to reveal possible "thru wall" discontinuities.

Thanks for your correction.
qcrobert (Industrial) (OP)
28 Jan 11 10:43
Whoops,
Instead of PYP I meant PJP, Partial Joint Penetration.
Duwe6 (Industrial)
28 Jan 11 13:36
Again:  time for new/better welders.  Thru-wall lack-of-penetration and incomplete fusion on a regular basis is reasonable cause to cancel the welder's certifications.  It is totally unacceptable workmanship.

  These tanks are a disaster waiting to happen.  These flaws will not get better over time.  As the tanks are pressure-cycled and temperature-cycled, this type of flaw tends to propagate.  Usually, in stainless this results in leakage, not a catastrophic failure.  However, the liability for a catastrophic failure is HUGE.  It will start at 7-figures $$ and go up depending on quantity of damage and how many people got maimed and/or killed.

  Please hire a knowledgable Welding Engineer to consult on how to bring your welds up to marginally acceptable.  You are flirting with disaster.
Kiwi2671 (Structural)
28 Jan 11 19:27
qcrobert,
I am trying to get my head around how you can possibly have PJP on 3/16" wall thickness.
Are the welders starting with a square butt joint with a slight bevel preparation (to assist in seeing where they are going)or are they using a square butt joint and just running the grinder down the join to form a small U prep ?
Either way a single run on 3/16" plate should have depth of penetration at least half the plate thickness - repeat for the second run from the backside and you should have full penetration with minimal (if any) backgrinding.
The point I am trying to make is irrespective of whether the client has stipulated CJP or PJP, on plate that thin it will cost no more in time or money (or increase in distortion) to make the weld CJP.
As for the welders, it sounds like a bit of retraining might be in order.
Regards,
Kiwi
unclesyd (Materials)
29 Jan 11 17:47
One thing you can do until you get your welder in line is to have someone to follow the backside with a gas lens. need good ventilation and Do Not use N2.  
vanmorrison (Nuclear)
29 Jan 11 18:32
Hello,

Ed Stainless,  I have seen specs that state "no indications" as the acceptance criteria... notta, zippo, no bleed out what so ever.

Regards,
Van  

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!

Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close