AEM 336: Reliability & Sampling Prediction & Modeling Outline • • • • 2 System Reliability Series System Reliability Parallel System Reliability Series-Parallel System Reliability Review Static Systems: Systems where failure of one component has NO effect on the probability of any other component failing Dynamic Systems: Components are dependent; failure of one component will affect the probability of failure of another component Review Series System: A complex system of independent units connected together (interrelated) such that the entire system will fail if any one the units fail. Review Parallel System: components are connected in such a way that a redundant, or standby, part can take over the function of a failed part to save the system. Review The Product Rule: if a system has n components, each with a reliability P1, P2,…,Pn, the reliability of the system (Rs) is Rs = P1 * P2 * … * Pn, where, Rs = Prob. Of system functioning as intended Pn = prob. Components functioning as intended Example: 3 components – A(.92), B(.95), & C(.96) Rs = A * B * C = (.92) * (.95) * (.96) = .839 Review Equivalent Component Reliability: Rs = Pc * Pc * Pc = (Pc)n Establishing Equivalency for non-equivalent reliabilities: A(.92), B(.95), & C(.96) 𝑃𝑐 = 𝑛 𝑃1 ∗ 𝑃2 ∗ ⋯ ∗ 𝑃𝑛 = 3 .92 ∗ .95 ∗ .96 = .9432 Review Unreliability (U): defined as 1-Reliability U = 1 - Pc for a component U = 1 - (P1 * P2 * … * Pn) Or U = 1 – (Pc)n Series Systems Review Series System Reliability using Failure Rate (λ): Rs = P1 * P2 * …* Pn Rs = e-λ1T * e-λ2T *… e-λnT Rs = e-T(λ1 + λ2 + … + λn) Where: λ = failure rate of component T = x-hour reliability of the system Review Example: Failure Rates: λ1 = .002 λ2 = .001 λ3 = .0025 λ4 = .0005 ∑ = .0060/ T = 100 Rs = e-T(λ1 + λ2 + … + λn) = e-100(.006) = .5488 or Rs = e-100(.002) * e-100(.001) * e-100(.0025) * e-100(.0005) = Rs = .8187 * .9048 * .7788 * .9512 = .5488 Calculator Tip: eˆ((-100).002) Parallel Reliability The reliability of a parallel (or redundant) system MUST be determined by 1st calculating the probability that the system or part WILL fail (unreliability). Rs = 1 – (U1 * U2 * …* Un) Where: Ux is the unreliability of a component AND Rs = 1 – (Uc)n = 1 – (1 – Pc)n Parallel Reliability Example: RA = .92; UA = 1 – PA = .08 RB = .95; UB = 1 – PB = .05 RC = .96; UC = 1 – PC = .04 Rs = 1 – (1 – Pc) = 1 – (UA * UB * UC) = = 1 – (.08 * .05 * .04) = .9998 SERIES RA*RB*RC 83.9% vs. vs. vs. PARALLEL 1-(UA*UB*UC) 99.98% Parallel vs. Series 2 vs. 3 vs. 4 components at Pc = .70 Parallel Series 2 .70 .70 = .91 2 (.70)2 = .49 3 .70 .70 .70 = .973 3 (.70)3 = .34 .70 .70 4 .70 .70 = .992 4 (.70)4 = .2401 Parallel vs. Series 1.20 0.97 1.00 0.99 0.91 0.80 Parallel 0.70 0.60 Series 0.49 0.40 0.34 0.20 0.24 0.00 1 2 3 4 Series-Parallel Systems (B) RB = .234 .9320 (A) RA = .358 UB = .766 .9520 (C) RC = .086 .9660 UC = .914 Series Part => Parallel Part => (RA)(RBC) (RBC) ***Find unreliability of B & C *** Must find this 1st! Series-Parallel Systems .107 1) RBC = 1-UBC = 1-UBUC = 1-(.766)(.914) = 1-.700 = .300 2) RA = .358 3) RS = (RA)(RBC) = (.358)(.300) = .107 High vs. Low Level Redundancy Parallel Systems Parallel Components High Level – Entire System in Parallel Series .7 .7 .7 .7 .7 .7 RS = 1 – {[1 – (.7*.7*.7)] * [1- (.7*.7*.7)]} = .5684 High vs. Low Level Redundancy Low Level – Component Level CAN BE REPLACED .7 .7 .7 .7 .7 .7 RS = [1 – (1 - .70)(1 - .70)3 = .7536 The Unreliability of each of the 3 Parallel Parts of the System Dynamic Systems Series Dynamic Systems: Calculate the failure rate of system by summing the reciprocals of the means (MTBF) of each component Example – 100-hr Reliability Subsystem MTBF Reciprocals 1 5000 1/5000 = .0004 2 6000 1/6000 = .00016 3 4500 1/4500 = .0002 4 2200 1/2200 = .00045 5 8650 1/8650 = .000115 .00116 failures/hr Dynamic Systems Subsystem MTBF Reciprocals 1 5000 1/5000 = .0004 2 6000 1/6000 = .00016 3 4500 1/4500 = .0002 4 2200 1/2200 = .00045 5 8650 1/8650 = .000115 .00116 failures/hr RS = e-λT = e-.00116(100) = .8905 Parallel Dynamic Systems 2 Types 1) Manual Switching 2) Electronic Switching Assignment 1) Exam on Chapter VI Modeling & Prediction, Tuesday, November 2 2) Assignment: Worksheet on Prediction & Modeling Due November 2