How can I know if a certain failure during operation (explosion, leakage, etc) is caused by wrong stress analysis?? (wrong input, wrong analysis, etc) PLs help.

Follow up question: I have heard of an incident, specifically the pipe is destroyed... We are not in a site yet, some say the reasons are probably wrong choice of material, some say wrong stress analysis (which during analysis, wrong input may had happened and been overlooked during checking), some say abnormal condition not anticipated. So how can I know if this happening is caused by wrong stress analysis??? Please give me some points or investigative approaches for this...

The art and science of forensic analysis is best not undertaken by a person who has no experience or background in it.

I suggest you...

A) Buy some textbooks on failure analysis check Amazon.com try using as author Helmut Thielsch (may he rest in peace).

B)Retain the services of a recognized expert in this field and try to collaborate with them.

And finally and most importantly do not reach any predisposed conclusions.

John, that last statement speaks volumes.

Logically speaking, before asking any questions, a failure point would need to be determined (what/where), then how it failed, and finally, why it failed. If my thinking is correct, then the question EIDT is asking should be irrelevant until the what, where, and how is answered. Especially not until you've at least seen the site.

Sounds like an interesting, yet sad, assignment. I hope there were minimal casualties. Or better yet, none.

Just my thoughts.

It actually happened somewhere, and I tried to research who the design company is, so I could have an information on how to deal with that (just in case it happened on our part). Consider me as really new in my profession, could you please give some typical (detailed) picture of that physical defects, what it looks like, what it should look like (??) (suppose looking back through calculations, and reviewing, before I could come up with a decision, could take a lot of time)on how I could really know that a certain failure is caused by wrong analysis...

Get out while you still can. God help us.

Do you have your reference books on order yet? Have you looked into the retention of an expert in this field?

Hello,

The reason for investigating a structural failure is to prevent it from happening the same way again. The benefits to that do not need to be explained. To have any other goal for such an investigation is beneath the dignity (if not the ethics) of ENGINEERING.

It takes many, many years of experience to become a competent investigator of structural failures. This on top of a very solid technical education. You must be trained in such arts/sciences as failure modes and effects analysis, fracture analysis (just looking at the fracture surface will tell an EXPERT a lot), and reading of crystal balls. If you approach the failure with the notion that you will pin it on a single cause you are likely doomed to failure. There are no "short-cuts" to failure investigation and there are not "short-cuts" to becoming a competent failure investigator.


Regards, John.

Remember you need the right type of expert. Sometimes being too much of an expert in only one area is not necessarily a good thing. An expert also needs to be seen to earn his money. I came across a multiple common mode failure. The expert reason was fatigue caused by poor welding. IMO, while technically correct, the real but non expert reason was a very poorly thought out design (being polite) which was impossible to fabricate. The end result of the "expert opinion" was that the real reason for failure was alive and well in design manuals and still being intstalled.

The fabricator often gets blamed for lousy design details... most people don't even consider weldments and their design important enough to detail!