MTTF
two ways!
….and l´Hopital… MTTF= integral(R(t)) !!
Serial configuration
INDEPENCE
Compressor: complex machine!
Key?
Reduction to components:
Tool:reliability block diagram
Low and High-Level-Redundancy
Rlow >= Rhigh, Rlow more effictive
k out of n redundancy
MTTF:
or with R(x) = e^-lamda*x
Parellel Configurations
Complex Configurations: Enumeration
Here: 2^5 possible configurations
BFS (breath first search)
add start- and endpoint-number (first and last number)
list: connection of every component to others, exept for destination!!
Minimal path and Cut sets
cuts: cut for which all components must fail to cause systeme failure
path: ..system functioning
a tighter upper bound
minmal path (series) in parallel - complete opposite as tighter lower bound
System bounds
Crude estimate:
lower bound: all components in series
upper bound: alle components in parallel
often not useful, because bounds are too wide
common-mode-failures
add additional component with reliability R´ in series
common powersource, common environmental conditions
A tighter lower bound!
minimal cuts (paralel) in series
Survival Signature
Contains different types of components
m_x represents number of components of type x
lx = number of working components (of m_x); FORMULA NECCESARRY?
Standby Systems
Cold standby
Hot standby
Cold standby: Standby component is not subject to failure until switched on
Hot standby: standyby component may fail while it is in standy mode
Availability
… is quantifying the ability of a system to fulfil its objective when considering maintenance
Maintenance procedures
Average Description auf availability
Corrective: replacement or repair of failed unit (item is allowd to fail)
Preventive: regular inspections (and eventual repair)
Conditioned: repair in case degradation is detected
Unrevealed failue
Testing or preventive maintenance
And repair!
Unattended Component (no repair)
… Functions until failure, no repair action taken / allowed
——
Continuously Monitured Component (1/2)
-repair immediately after failure
-duration of repair: g(t) and G(t)
-time to failure: R(t)=e^-lamda*t
-probability of component being up at t: p(t)
Continuously Monitured Component (2/2)
Component under Periodic Test and Maintenance (simplyfied)
components operating in standbymode must be periodically tested
components are unattended between tests
failure is revealed only when tested
availability p(t) and reliability R(t) are equivalent
simpyfied average unavailability - component under periodictest and maintenance
preventive and corrective maintenance (1/2)
MTBF (meantime between failure)
average availability (maximize!)
tau: operation period, tau_r=preventive maintenance, tau_e: corrective maintenance, t: time to failure
preventive and corrective maintenance (2/2)
Preventive and Corrective Maintenance considering costs
TERCC (total expected replacement cist per cyle)
and total expected replacement cost per unit
Last changeda year ago