You are on page 1of 20

AVAILABILITY AND MAINTAINABILITY

One-shot devices/equipments have to


operate/function satisfactorily at the
intended point of time.
Most products/devices are meant for
intermittent or continuous duty.
Many of these items are repairable.
Once such a product fails, a diagnosis of
failure followed by a corrective action can
put the product to its initial or pre-failure
state and the product becomes available
for use (intended function). Thus , the
repair function comes in.
Efforts to enhance availability of a product
at all times during its useful life go much
beyond repair.
In fact, we focus on
1. Monitoring /inspection at appropriate
times and opening the product up
2. Repairing the components that are going
to fail, by replenishing the required
properties and
3. Replacing the components that have
already failed.
These three activities constitute the core
of maintenance.
Like any other activity, maintenance as a
broad activity as also each of its
component activities has to be
1. Planned
2. Monitored and
3. Evaluated.
Continued availability through necessary
maintenance corresponds to
maintainability.
However, maintainability to ensure both
point-wise and interval availability is
primarily determined by design features
that take due care of ease and
effectiveness of inspection and
repair/replacement.
Usual availability analysis assumes only
two states viz. functional or up and failed
or down for any component or system at
any point or during any period of time.
Point-wise availability or operational
readiness requires that the product is in
the up-state at the intended point of time.
Availability will be quantified as a
probability that the product is in the up-
state when required.
Point-wise availability of a non-repairable
item which is required to be up at many
different points of time can be estimated
as the proportion of times the item was
really in the up-state.
Proceeding analytically, we can take
availability as a function
A (t) =1 if the item is up and
=0 if the item is down , at time t.
The availability over a period (0,T) is the
integral of A (t) over (0,T).
The limit of this as T goes to infinity is called
limiting availability.
Period availability can be interpreted as
the proportion of time the item remains up
or operational during a given period of
time.
The item could be down ,waiting for or
undergoing repair, during the remaining
portion of the given period.
When we consider an item that starts in
the up-state, remains there for a random
duration and fails, waits and then goes for
repair for random times, and then comes
back to the up state, we define availability
as under
Availability= MTTF/ (MTTR+MTTF) where
MTTF=mean time to failure and
MTTR=mean time to repair.
To increase availability, therefore, we
need to
Increase MTTF (mean length of stay in the
up-state)
Decrease MTTR (including mean time
waiting for repair)
Or do both.
We proceed further by noting that most
items in use are multi-component systems
with certain configurations.
To enhance MTTF, we can
introduce cold or warm redundancy
use devices /systems under less than
rated capacity
increase component reliabilities.
An important and sometimes challenging
task is to optimize redundancy to achieve
the highest reliability subject to cost or
weight or volume constraints.
Reduction in MTTR would obviously imply
Avoiding delays in taking up
repair/replacement processes and making
them effective. Also inspection of system
state should be simple and quick.
Maintainability is a consequence of two
functions viz. (1) system design and
(2) maintenance.
The second relates to maintenance on
failure/ break-down.
Planned maintenance leads to enhanced
availability.
To reduce maintenance delays and
enhance maintenance effectiveness,
planning of maintenance manpower and
maintenance facilities as also of spare
parts plays an important role. Spares
planning must take into account failure
rates of different parts as also their costs,
besides considerations of holding
inventories with associated costs.
Calculation of availability requires
calculation of MTTF as the inverse of
system failure rate. The latter is obtained
from system configuration and component
reliabilities. For simple systems like series
or parallel or k-out of-n, this is relatively
simple. Even for complicated systems,
system failure rate can be worked out by
the min-max theorem.
System mean time-to-failure can be also
derived from the probability distribution of
system failure time, starting with
corresponding distributions for
components. Expressions are simpler if
components function independently with
the system and there is no mechanical
sticking. This again is a consequence of
system design.
Redundancy optimization attempts started
with Ketteles algorithm and has since
proceeded through many search
algorithms genetic algorithm.
A lot of recent work on this topic
corresponds to soft computing, especially
those which use fuzzy component
reliabilities or fuzzy system configurations.
In the final analysis, availability is the
composite of reliability and maintainability.
In many cases, availability and reliability
are synonymous.
Availability is ensured through proper
design, careful use and effective
maintenance.
All these three are covered in Concurrent
Engineering.

You might also like