1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
Dismiss Notice
You must be a registered member in order to post messages and view/download attached files in this forum.
Click here to register.

Reliability Prediction by knowing component counts and processes involved

Discussion in 'Reliability Analysis' started by JMReyes, Feb 11, 2016.

  1. JMReyes

    JMReyes New Member

    Joined:
    Feb 11, 2016
    Messages:
    3
    Likes Received:
    1
    Trophy Points:
    2
    Location:
    PHILIPPINES
    Hi, I'm new to this forum and I'm looking for insights into what Reliability Prediction Model is available out there with regards to Electronics Assemblies and systems with known components and processes.
    For example, if I have a 4-layer PCB with 80% surface mount devices and 20% discrete components that will go thru reflow soldering, wave soldering, conformal coating, potting encapsulation, etc... Is there a prediction model that is out there? or there is really no shortcut and I have to build the assembly and test it and then do all the reliability demonstration tests to come up with the reliability numbers?
     
  2. Miner

    Miner Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    576
    Likes Received:
    492
    Trophy Points:
    62
    Location:
    Greater Milwaukee USA
    There are newer alternatives to the outdated Mil-HDBK-217, such as Telcordia SR322, FIDES, Reliasoft's Lambda Predict and several more, but none provide good estimates of the actual product life. They are best used to either identify the components most likely to fail, or to compare the relative reliability of alternative designs. They work well for these, but not for estimating the actual life. Given your scenario, I would investigate using a Physics of Failure approach. The CALCE consortium through the University of Maryland would be a great resource in this. CALCE has several software packages (e.g., SARA) to help with this. There are several others also available such as CADMP-II and DFRSolution's Sherlock.
     
    JMReyes and Bev D like this.
  3. Bev D

    Bev D Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    605
    Likes Received:
    663
    Trophy Points:
    92
    Location:
    Maine
    CALCE consortium is making some good attempts. but as Miner indicated reliability prediction models typically don't work well for predicting actual failure rates. a great article on this is Why the Traditional Reliability Prediction Models do not Work. Also available at the CALCE web page. The difficulty is that all models are based on assumptions and theories; the actual situations are often ignored and certainly the effect of 'process' is often ignored. Process includes everything form solder processes to the semiconductor fabrication process that can introduce far more weaknesses in the system than the 'cold' design recognizes. many models often work only at characteristic nominals and not various combinations of component distribution across tolerance ranges. There are of course some good things the models can tell you that can point out weaknesses or help understand a first budget for failure or warranty and so they shouldn't' necessarily be abandoned altogether. or as George Box said "all models are wrong, some are useful".
     
    JMReyes likes this.
  4. Miner

    Miner Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    576
    Likes Received:
    492
    Trophy Points:
    62
    Location:
    Greater Milwaukee USA
    Ravi Khare, JMReyes and Candi1024 like this.
  5. Bev D

    Bev D Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    605
    Likes Received:
    663
    Trophy Points:
    92
    Location:
    Maine
    Thank you! I needed a good laugh today!
     
  6. JMReyes

    JMReyes New Member

    Joined:
    Feb 11, 2016
    Messages:
    3
    Likes Received:
    1
    Trophy Points:
    2
    Location:
    PHILIPPINES
    Miner, thanks. I will need this 'alternative methods' once i'm too bored reading the references you shared earlier :)
     
    Bev D likes this.
  7. JMReyes

    JMReyes New Member

    Joined:
    Feb 11, 2016
    Messages:
    3
    Likes Received:
    1
    Trophy Points:
    2
    Location:
    PHILIPPINES
    Bev and Miner, Thank you for all the references you cited. I'll be browsing the internet for these materials, download them, and read through them.
     
  8. charanjit singh

    charanjit singh Member

    Joined:
    Jan 21, 2016
    Messages:
    32
    Likes Received:
    10
    Trophy Points:
    7
    JMReyes, hope you have solved your problem by now; or you still need some more inputs
     
  9. Mike wizowski

    Mike wizowski New Member

    Joined:
    Sep 22, 2017
    Messages:
    4
    Likes Received:
    2
    Trophy Points:
    1
    I would think you wouldn't want to go through the pain of build the assembly and test it and then do all the reliability demonstration tests to come up with the reliability numbers all yourself as an individual. Wouldn't you be better off reaching out to a custom who could draft that for you and give you a running quotes and cut through that time and headache. Granted you'll be dealing with ballparks but that seems like what you're going after. The think manufacturing resource planning is the best bets of preserving your capital.
     
  10. Miner

    Miner Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    576
    Likes Received:
    492
    Trophy Points:
    62
    Location:
    Greater Milwaukee USA
    Hello Mike and welcome to QFO. Are you associated with the company to which you provided a link?
     
  11. Mike wizowski

    Mike wizowski New Member

    Joined:
    Sep 22, 2017
    Messages:
    4
    Likes Received:
    2
    Trophy Points:
    1
    Hey there. I am not. I have been a customer before when I used to dabble more heavily in manufacturing. But they treated me well so i'm not shy about passing them along. Hope I didn't flag you sorry about that.