8. PURCHASINGPURCHASING
• CASTINGSCASTINGS
• HEAT TREATMENT REQUIRED (shownHEAT TREATMENT REQUIRED (shown
on drawing)on drawing)
• Heat treatment was OMITTED, castingsHeat treatment was OMITTED, castings
failed to meet hardness, tensile strengthfailed to meet hardness, tensile strength
requirementsrequirements $$
• Failure analysis resulted in redesign toFailure analysis resulted in redesign to
thicken area where failures had occurredthicken area where failures had occurred
13. SUBSTITUTIONSUBSTITUTION
• Natural Gas Processing PlantNatural Gas Processing Plant
• Stainless steel and Cupronickel – sameStainless steel and Cupronickel – same
plant, same fabrication shopplant, same fabrication shop
• CuNi filler used on stainless steelCuNi filler used on stainless steel $$
• COPPER CONTAMINATION CRACKINGCOPPER CONTAMINATION CRACKING
• Ref. Welding Journal, Feb. 2010Ref. Welding Journal, Feb. 2010
14.
15. DESIGN and INSTALLATIONDESIGN and INSTALLATION
• Welded construction, two heavy plateWelded construction, two heavy plate
sections, pipe – joined by weldingsections, pipe – joined by welding
• Needed FULL penetration weldsNeeded FULL penetration welds $$
16.
17.
18.
19.
20. FATIGUEFATIGUE
• Repeated or fluctuating stressRepeated or fluctuating stress
• Stress below tensile strength of theStress below tensile strength of the
materialmaterial
• Progressive fracture from initiation point(s)Progressive fracture from initiation point(s)
– Often created by aggressive environment orOften created by aggressive environment or
sensitizing condition (corrosion, fretting, arcsensitizing condition (corrosion, fretting, arc
strikes, weld spatter,strikes, weld spatter, Lack of PenetrationLack of Penetration))
21. Trinity Metals LaboratoryTrinity Metals Laboratory
QUALITY PHILOSOPHYQUALITY PHILOSOPHY
• QUALITY is a fundamental business strategy forQUALITY is a fundamental business strategy for
Trinity Metals Laboratory.Trinity Metals Laboratory.
• Our GOAL is to be the best provider of testingOur GOAL is to be the best provider of testing
services in our market in terms of accuracy,services in our market in terms of accuracy,
dependability, and value to our customers.dependability, and value to our customers.
• We are committed to providing our customersWe are committed to providing our customers
with consistently high quality testing services,with consistently high quality testing services,
– defect freedefect free
– on timeon time
– at competitive prices.at competitive prices.
Editor's Notes
There was a game show called “Don’t Forget the Lyrics” in which a contestant had to remember the words of songs. The lyrics were projected on a screen and the song was played with accompanying singers….up to a point. The background singers stopped and the words on screen went blank…;at which time the contestant had to sing the exact following words.
Here’s our version of that game tonight: {no singing allowed!} “…what we have here is_____”
-right: FAILURE TO COMMUNICATE
When we investigate failures we need to discover what was the root cause of the failure. We need to discover this so we can stop it from happening again (subliminally: Toyota accelerator linkage). In my experience, many of the failures we are asked to investigate start with a failure to communicate some critical item of information. Sometimes the failures are small, maybe manufactured parts or components failing to meet specification. In such cases the failure analysis may be referred to as "process troubleshooting". In other cases, parts or components fail in service, with a range of negative consequences. We are lucky if field failures only result in lost time and revenue, while some occurrences result in personal injury – or worse.
Among component failures that have come to me for investigation, I find that many have resulted from lack of communication at some critical juncture. Certainly many more notorious events have been marked by lack of communication, such as Custer’s Last Stand at Little Big Horn and the Bay of Pigs invasion.
Before going further, let’s look at the importance placed on FAILURE ANALYSIS by ASM:
ASM INTERNATIONAL is the Materials Information Society. This is a page from ASMINTERNATIONAL.ORG under the members online access to Metals Handbook Desk Edition. It displays the two major headings available to members there: Engineering Aspects, and Failure Mechanisms. It notes that over 1000 cases are documented here. (It is also true that you must be a subscriber to access them.
On the next slide is a screen capture from Lonestar Chapter of EDFAS, which is an entire Society whose members are associated with failure analysis. EDFAS is an affiliated society of ASM. The high level of reliability of electronic devices we take for granted today can be traced back to the efforts of these professionals who specialize in failure analysis.
In my experience, many times failures of communication occur in areas such as (but not limited to…I’m sure you can think of additional categories…
DESIGN and MATERIAL SELECTION; PURCHASING; INSTALLATION; SUBSTITUTION; REPAIRS; and APPLICATION / OPERATION.
In many areas, once specifications are made, they don’t go away. In our lab we see many instances of outdoor construction in which the specified material is A36 steel, but with supplementary requirements for Charpy impact resistance. A36 was first specified by ASTM in 1960 as the replacement for A7. It never was intended to be impact resistant. OFTEN we find that A36 will not provide the required impact resistance, so we are asked to perform testing to re-certify a tougher [newer specification] steel to meet the requirements of A36 just because the specifications for many outdoor constructions have never been reviewed and updated. Consequently, we also occasionally see failures in which A36 was supplied but did not have adequate impact resistance.
Like in the previous example, fabricators sometimes fail to adequately consider the effects of grade, or steel production methods on the properties of the steel. In an example of components that failed during fabrication is a HYDRAULIC FITTING that started out as a preform that was plasma cut from plate of net thickness (1-1/2") for the finished part. Two intersecting holes were drilled, then tapped. Several were found on inspection to be cracked, each at or near the centerline of the plate.
Now it just so happens that the steelmaking process can result in elemental segregation. In this design and processing path, the centerline segregation was present at the smallest cross section of the part, and was sensitive to the stresses resulting from fabrication. In this instance about 20% of the parts were cracked. I am given to understand that the fabricator increased the number produced and inspected 100% before shipping. However, that is an ill-advised route, because more units are likely to fail due to installation and during subsequent operation.
One cracked part is shown here (pointer) along with some of its classmates. The crack is easily visible (pointer). Below is a portion of the cross section of that part. I took hardness readings at 7 locations: these 6 plus one near the plate surface. Location #1 is nearest the flame cut surface in the Heat Affected Zone AND LOOK AT THAT HARDNESS: HRC51! Near the plate surface only B94. Clearly the centerline segregation zone reacted to the heating of the flame cutting. Now just think how differently the results would have been if only the properties of all steels were always uniform and isotropic like it’s assumed.
Here’s the full cross section 1-1/2” square with a 1” tapped hole…and this nice crack.
The next example may have been a failure of communication in the PURCHASING arena. The customer presented a part from stock along with a drawing, and followed up with a few blurry photos that were good enough to show how the failed part looked. The part was a casting, and the drawing called out heat treatment and the required hardness and tensile strength. The requirements were presented on the drawing, but still communication failed.
Whether the required heat treatment was omitted by conscious decision to save money, or inadvertently, matters little. But cost of replacing the failed parts, plus recalling all unused parts in the field only to have them heat treated…OH! And the cost of losing customer confidence and good will…far outweigh any minor amount of savings from the omitted heat treatment.
The inset shows the part as we had sectioned it for testing. The solid end (pointer) was sound, and the hollow cylindrical cross section showed no voids (although the tab had a very noticeable void, it did not participate in the failure, which consisted of this tab being ripped out of the wall (pointer). The un-heat treated wall was insufficient to bear the applied loads. One positive result is that this part has been redesigned to strengthen the section. And heat treatment of all future parts will be assured (?)