PDA

View Full Version : accelerated life test with multiple/combinations of stressed variables


Hebbat
March 18th, 2009, 01:32 AM
Hello
i would like to test the life of a product however there are 4 variables that can contribute to failure--> pressure, temp, current, frequency

by having 4 separate sample groups when i run the accelerated life test, i will keep 3 variables constant and stress the product with only one variable at a time. This will give me independant results wrt to the life time of the product.

What i'd also like to do is understand how the combination of variables affect the failure rate. What kind of test plan should i use for this? can i stress all variables at once and still use minitab to analyse the results?

thanks

Hebbat
March 19th, 2009, 03:42 AM
hello forum

my grand apologies. I was referred here from the isixsigma forum and i didn't realise this was unrelated to minitab.

I do not use any of this software but i am definitely open to suggestions on how to achieve these test plans that i am seeking in other ways.

Does anyone have any ideas?

cheers

Heb

David
March 19th, 2009, 01:31 PM
Hi Hebbat,

Thank you for coming by. No problem, we are glad that you stopped by nontheless. ReliaSoft does have an accelerated life testing application called ALTA (http://alta.reliasoft.com). ALTA is designed to analyze data from accelerated life tests and can anlayze data with up to eight stresses. ALTA could be used to analyze the data from your test. In addition, the Accelerated Life Testing Reference is online at http://www.weibull.com/acceltestwebcontents.htm. A free demo of ALTA can be downloaded from our Free Download Store at http://www.reliasoft.info/download/product.php?productid=9&cat=6&page=1.

In regards to how the "combination of variables affect the failure rate", I assume that you are referring to the possible interactions between the stresses. Given this, an article on determining interactions using ALTA can be found at http://www.weibull.com/hotwire/issue53/hottopics53.htm.

I hope this helps.