PRISM Forum - Message Replies


Topic: PRISM Questions and Answers
Topic Posted by: SRC (src_forum@alionscience.com )
Organization: SRC
Date Posted: Wed Jan 12 8:33:33 US/Eastern 2000
Topic Description: Welcome to the PRISM forum! Please feel free to post your questions and comments about the PRISM assessment software here.

Back to message list About this forum

Original Message:

Posted by: Bart de Boer (bart.deboer@ultracontrols.aero )
Organization:Ultra Electronics - Controls
Date posted: Wed May 9 11:40:54 US/Eastern 2007
Subject: Calendar FR vs Operating FR
Message:
RE: PRISM Field and Predicted Comparison David Dylis reply to “Failure per Million Hours” Tue Dec 10 13:23:10 US/Eastern 2002 My interpretation of the example: Failure rates in PRISM are presented in failures per million calendar hours and not in failures per million operating hours which adds to the confusion when a user is only interested in failure rates solely due to operation. I will answer the question with an example. If an item has a total failure rate of 1 failure/million calendar hours, the operational failure rate is 0.99 failures/million operating hours and the non-operating failure rate is .01 failures/million non-operating hours. If it is operating at a 5% duty cycle then we obtain: 1 Failures/Million Calendar Hours = (.05) x 0.99 Failures/Million Operating Hours + (0.95) x 0.01 Failures/Million Non-operating Hours Now converting Failures/Million Operating Hours (.05 Operating/Calendar) x 0.99 Failures/Million Operating Hours = 1 Failures/Million Calendar Hours - (0.95 Non-operating/Calendar) x 0.01 Failures/Million Non-operating Hours or Operational Failure Rate at a 5% duty cycle = 20.010101 Failures/Million Operating Hours. The formula then is: F/MOH = ({op F/MCH + non-op F/MCH}-[non-op duty*non-op F/MCH])/(duty*op F/MCH) Similarly, for a 50% duty cycle the failure rate would be: 1.91919 Was a 0.1 non-op F/MCH accidentally used within [ ] brackets?, The PRISM predicted failure rate is based on calendar time. The duty cycle in PRISM identifies the amount of time that an item operates (e.g., 5% of the time) during the calendar period (e.g., 1 million calendar hours). Therefore, in a given calendar period, items that operate with a higher duty cycle are actually operating a greater amount of time then those at a lower duty cycle. When a conversion from calendar to operating time is made and there is no or minimal difference between two predicted failure rates with varying duty cycles, then the item with the lower duty cycle will have a higher predicted failure rate since the predicted value is based on a lower amount of operational hours. It seems then that the PRISM FR is constant over calendar time, irrespective of how it’s used. i.e. with op F/MCH much greater than non-op F/MCH, failures will be compressed into the operational time. Thus if a unit has a 100% duty cycle (always on) it will, over 114 years (1 million calendar hours), exhibit the same number of failures as a similar unit with a 5% duty cycle (on for a total of 5.7 years) over the same 114 year period, with almost all of the failures occurring in that 5.7 year operating time (i.e. failing 20x more often per OH than the always on unit).


Reply:

Subject: Calendar FR vs Operating FR
Reply Posted by: David Dylis (ddylis@alionscience.com )
Organization: Alion SRC
Date Posted: Thu May 10 14:33:00 US/Eastern 2007
Message:
The example which you identied was not based on actual PRISM predictions. The same calendar failure rate was provided for varying duty cycles to show how a user could determine the operating failure rate from a given predicted calendar failure rate. PRISM does allow the user to select varying duty cycles when calculating calendar failure rates. Calendar failure rates will change based on the duty cycle selected. In addition to cycling effects, electronic devices can fail as a result of time, moisture and temperature. For a failure to occur, a contaminant or weak spot in the device is required. This contaminant or weak spot will eventually cause a device to degrade to failure over time. For instance, when an integrated circuit is operating at a 100% duty cycle, moisture is driven out of non-hermetic style packages due to self-heating. Even in hermetic packages mono-layers of moisture may be present on the surface of active components in a non-operational device. Contaminants typically require moisture, temperature and time to result in failure. Weak spots and fatigue related failures are typically the result of temperature cycling and number of temperature/power on/off cycles. In-rush currents from power cycling can weaken and/or cause marginal devices to fail. The greater the amount of time a system/device is fielded, then the greater number of temperature cycles as well as overall time that can cause failure of the device or system.


Back to message list About this forum

Reply to this message