Assistance to begin a Reliability Program

[MrUser]

I need some advice from some of the reliability professionals in the room. I work for a small company as a Mechanical Engineer and have been asked to begin a reliability program. My company makes tools and components for the construction industry. Frankly I am not sure where to begin. I was asked to do this to help improve our current products and help validate new upcoming designs. Of course this in in addition to what I normally do each day. Anyone know of a computer program or book to assist a non-statistician with confidence levels and sample sizes? I need to start to build the foundation to a reliability program.
HELP!

 

[IRstuff]

You can download some mil specs such as MIL-STD-338:

http://assist1.daps.dla.mil/quicksearch/basic_profile.cfm?ident_number=54022

But the bottomline questions are:
> What is the reliability of your product in its environment?
> What are the various failure modes?
> How can you accelerate the failures to get data in a meaningfult timeframe?
> What can you do to change the outcome?

I'm guessing that your issues are wear and breakage? Anything else?




TTFN

 

[MrUser]

Thanks for the quick response IRstuff! The reliability of our current product varies greatly depending on the person using the tools. Being in a construction atmosphere, the tools are sometimes severly beat. You are correct with our failures being wear and breakage. Accelerating the failures can be done internal by continuous cycles. We have troubles duplicating the harsh environment. Improving reliability is the easy part. The tests need to justify the cost to redesign and improve the components.

Right now we pick a number out of the air and test X components to a predetermined number of cycles. But choosing the number of components to test is anything but scientific. There is no confidence that the number of samples chosen for the test will be truely representative.

I am anything but a stats guy. I must have spent too much time in the CAD lab back in college. LOL! Im hoping someone automated the stat calcs in the form of an easy to use computer program.

 

[m777182]

There is a lot of computer programs but mostly all of them suppose a certain level of statistics.
I would suggest you to start with a simple introductory text. I keep a tinny book by Hitoshi Kume, Statistical Methods for Quality Improvement.It will guide you on a very friendly way through data collection, cause end effect diagrams to histograms, plotting scxater diagrams further to quality control. No violence, easy, examples, no theoretical high math. ISBN 4-906224-34-2 C0034.A sort of Mozart's "Kleine Nachtmusik".
m777182