Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic Welding Industry / Technical Discussions / Laminations
- - By dasimonds (**) Date 01-19-2002 12:36
I be interested to hear about any experiences people have had with laminations.

Are laminations an acceptable defect to leave?

Dale Simonds

Parent - - By - Date 01-19-2002 13:22
It generally depends on the direction of stress as to whether it is an acceptable defect or not. If you have a stress that acts through the thickness of the material, then it would usually not be acceptable. Typical examples here would be a set-on nozzle on a pressure vessel, or a T joint.

If the stress is parallel with the direction of lamination, the lamination does not actually reduce the strength of the material.

Another instance where laminations are a problem, is when the vessel (assuming it is a vessel) is in hydrogen duty. The laminations are excellent places for the H to collect and can actually cause "bubbles" to form in the material.

At the end of the day nobody really wants laminations in their materials, but you will only actively look for them when they will be a problem. As such, much plate work actually contains laminations, but because it is not detremental in many applications, nobody looks for it to any great extent.

The material specifications do usually have acceptance criteria for laminations.

Regards
Niekie Jooste
Parent - - By dasimonds (**) Date 01-19-2002 14:44
The reason I ask is that I discovered a lamination a couple of days ago while I was cleaning up some spatter on a pad eye, part of a lifting beam. It's not going to see more than 4 tons of weight at any one time.
The lamination is located less than an inch from the hole that accepts the shackle.

The engineer says leave it, but he's prone to shortcutting procedures for time saving. An example would be preheat. At -10 degrees F, I personally don't think preheat is something that should be omitted from welding procedures, no matter the use of the component.

I don't know what the relevant codes say about laminations, but my experience in ths shipyard was they never left a lamination, if they knew of it's existence.

What bothers me the most is that you never really know for sure where and how far the lamination travels, unless you did some UT, or you grind it out.

We had a case a couple of years ago with some heavy wall, 1.5", plain carbon steel steam pipe. It was an extruded material. If you weren't careful while running your hand down the outside of the pipe, you might impale your hand on slivers sticking from the OD. It was decided not to do anything about them. I did have to grind one of them out, as it was located at the weld joint, and appeared as a crack-like indication projecting from the weld into the pipe wall. It was roughly 2" in length, and traveled 3/8" into the pipe on an angle. With the RT requirements, I felt it needed to be removed. I never noticed it until I had completed the first stringer bead on the cap.

I wonder what occurs to the laminations when one applies some cyclic thermal stresses, as you might encounter with start up and shut down.

Dale Simonds
Parent - By - Date 01-19-2002 15:16
In something like a lifting beam, you will need to establish the direction of the stresses. If the lamination is parallel to the direction of stress and it is in tension, then you do not have a major problem.

If the lamination direction is perpendicular to the stress direction (In other words, the lamination is reducing the load carrying cross section.) then it is a problem and needs to be sized and, most probably, removed.

Another issue is when the material with the lamination is in compression when the lamination lies parallel to the stress direction. In this instance, the buckling load of the beam will be reduced.

The above arguments will also apply for cyclic loading.

Hope this helps
Niekie Jooste
Parent - - By pipewelder_1999 (****) Date 01-19-2002 14:04
I witnessed a 30,000 lb fan housing fall when after a lifting lug delaminated the plate it was attached to. The plate was 5/16" or 3/8" and the lug was 1". Someone decided that the 1" lug needed a 1" fillet weld instead of the as detailed 3/8" and the lug came off with about 1/3 the thickness of the plate.

It never was concluded that the plate was laminated or the excess weld caused failure in the through thickness direction. I think the latter.


Good Day

Gerald Austin
Parent - - By DGXL (***) Date 01-19-2002 17:14
From a structural standpoint, laminations are typically unacceptable. The UBC/IBC requires any material 1-1/2" or greater in thickness to be examined for laminations. I have a number of photos with failures that resulted from delaminations. I also have some macroetched specimens with the same discontinuity. I use these for training purposes regarding BM discontinuities. These have been discovered in a variety of shapes (wide flange, plate, tubulars) and thicknesses (including plate as thin as 1/2"). I've seen enitre sheets of imported 1/4" steel come apart during the flame cutting operation.

The presence of a lamination itself is not the concern, it's when that discontinuity propogates and delaminates is when it becomes problematic.

Planar type discontinuities do not hold well under stress unless it is compressive where the direction of loading will not be detrimental.
Parent - - By - Date 01-21-2002 18:22
Theoretically ALL plate material contain laminations. This is so because mostly laminations are "rolled flat" MnS inclusions. All standard steel plates contain these.

The issue is therefore the allowable extent of the laminations. Most of these laminations are below the limmit that can be detected by UT testing and as such it is assumed that the plate contains no laminations. As stated earlier, the material specification generally states the extent of testing, and the acceptance criteria. Very seldom is this testing a 100% test over the entire surface of the plate.

Obviously, when the design is such that it places a stress in a direction where the potential existance of a lamination could become a problem, additional testing by the fabricator is required.

Regards
Niekie
Parent - By DGXL (***) Date 01-22-2002 01:09
Niekie,
I respectfully disagree with several statements of your post.

1. If "ALL" plate material contain laminations, then these would be considered by designers with that criteria in mind. They (designers) do not include ALL of these in their calculations. Voids (planar or otherwise) at a microscopic or molecular level are not considered discontinuities. Any solid material has space between every atom. Most solids are nothing but space. In the aerospace industry many high stress components do not permit any type of planar discontinuity, period. This keeps the airplane airborne.

I did look up the definition of a lamination in several references including the Metals Black Book (Ferrous Metals), AWS B1.10 & B1.11, etc., none of which define the (minimum) physical size of this type of discontinuity and when it becomes termed as a lamination.

2. Other types of discontinuities that may be voids (porosity, shrink, oxides, etc.) are also commonly the source of laminations (and the most likely from what I've been told by metallurgist).

3. There are several labs here in CA that perform C-scan 24 hours a day on entire sheets of various materials for laminations (100% testing). We regularly test entire plates (1-1/2" and greater) prior to and after completion of welding to insure there has not been any delamination or through thickness shrinkage. This practice is applied regularly to cyclically (dynamic) structures as well. Much of the ultrasonic "noise" encountered during this testing is the structure of the material itself and includes voids, inclusions and whatever else happens to get rolled into this material.

I do agree that some discontinuities are acceptable to an extent (when determined by the referencing documents' acceptance criteria or the Engineer).
Parent - By R. Johnson (**) Date 01-24-2002 19:52
For plates 1/2" and over in thickness ASTM A435 is sometimes specified for the ultrasonic examination of plannar type of defects. One interesting facet about this standard is the acceptance standards.
"Any discontinuity indication causing a total loss of back reflection which cannot be contained with a circle, the diameter of which is 3 inches or one half the plate thickness, whichever is greater, is unacceptable."
This means that you may have some localized flaws that may cause real problems when stressed through the plate thickness.
In the past when critical point type loads are applied to the plate surface, we have ultrasonically tested the plate to insure that there is no subsurface flaw present that could lead to a failure.
Up Topic Welding Industry / Technical Discussions / Laminations

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill