Getting Started with Windchill Risk and Reliability > Windchill Prediction > Understanding Prediction Calculations
Understanding Prediction Calculations
Windchill Prediction calculates a number of metrics, including failure rate, MTBF, reliability, availability, and MTTR, which are briefly described below. For more information, see the help.
Calculation
Description
Failure Rate
In a reliability prediction analysis, failure rate is a value that represents the number of failures likely to occur over a time period, usually defined in units of FPMH (failure per million hours) or FITS (failures per billion hours). For example, if a unit has a failure rate of 2 FPMH, that unit is likely to fail twice in a million hours. Failure rates are predictive values computed using equations based on statistical field data analysis.
MTBF
The MTBF (mean time between failures) represents the number of hours a unit operates between failures. For non-repairable units, which cannot be repaired once a failure occurs, the MTBF is equivalent to MTTF (mean time to first failure). Therefore, the terms MTBF and MTTF are sometimes used interchangeably, although doing this is not theoretically accurate. In Windchill Prediction, the term MTBF is used.
Reliability
The probability that a system remains operational until a specified time. Reliability is a time-based probability value, so it is always a metric between 0 and 1. A reliability of 0 means that a system is infinitely unreliable, or never functioning. A reliability value of 1 indicates that a system is infinitely reliable, or always operating.
Availability
The probability that a system is operating properly at a specific time point. Availability, like reliability, is a time-based probability metric between 0 and 1. Availability is a function of both reliability (how quickly a system fails) and of maintainability (how quickly the system can be repaired). The common term "five nines" refers to a system that has an availability of 0.99999, which means that it is operational 99.999% of the time.
MTTR
MTTR (mean time to repair) is the average time it takes to repair a failed unit and return it to an operational state, based on the average repair time for its replaceable components. The MTTR is calculated by rolling up the average repair times for lower-level components to subassembly and system repair values. If Windchill Maintainability is used in conjunction with Windchill Prediction, the computed MTTR values of components and subassemblies can be obtained from the Maintainability module. Units for MTTR values are typically hours or minutes. MTTR is also referred to as MCT (mean corrective time).
Windchill Prediction can also compute the above measures using mission phases. To perform mission calculations, you define a mission profile, which describes the phases of a mission your system experiences. For example, the mission of an airplane might consist of three phases: take-off, flight, and landing. Each one of these phases occurs in different environments and conditions. You can also define a phase as dormant or active to factor in the lower failure rates experienced during dormant phases.
The Prediction module performs mission-based calculations by factoring in the environment, temperature, state (dormant or active), and percentage of time in each phase. You can view the results for the entire mission as well as for individual phases.
Windchill Prediction supports derating analysis as part of prediction calculations. Derating analyses allow you to assess which components in your system are overstressed. You can use the results of derating analyses to ensure your system components are operating within acceptable limits.
The Prediction module also supports a variety of reliability allocation methodologies. When using these methods, you allot MTBF values across the subassemblies and/or components of your system, and then, based on specified weighting factors, calculate the projected failure rate of the components and/or subassemblies in the system.
Reliability allocations are useful when you have reliability goals that must be met. For example, if contractual requirements dictate that your overall system MTBF must be at least 10,000 hours, you can split up that 10,000-hour goal across the various items in your system based on a weighting system of your choosing. Perhaps you assign one highly complex subsystem an MTBF of 5,000 hours, and split up the remaining 5,000 hours over all other subsystems. In this way, you can apportion MTBF goals to ensure that your overall objectives are met.