| All items will fail at some time. However,
maintenance professionals must aim to ensure that the failure does not
come before the designed time, or if a failure occurs, that it is not catastrophic
and that the unit can be fixed and put back into operation as quickly and
economically as possible. |
| Maintenance plays its role in the reliability
game under the "set of conditions" portion of the definition, where the
terms of maintenance are spelled out to get the predicted mean time between
failures rate (MTBF). For example, the owner/user of a machine must change
air filters every 500 operating hours, drain and flush gearboxes every
6 months or when oil is contaminated with water or product, and replace
clutch pads every 1000 hours of operation, plus perform a host of other
tasks specified by the manufacturer. |
| The first step |
| If specified preventive maintenance (PM) tasks
are not executed in a thoroughly disciplined manner, the basic design conditions
have been violated. The results of these omissions will cause failures
in subassemblies, and the full useful life of the machine will not be achieved.
This situation leads to the first step toward achieving the goal of reliable
machines: Perform thorough and disciplined preventive maintenance tasks
in accordance with the manufacturer's specifications. Because the manufacturer's
design engineers understand the reliability designed into the machine,
they generally know how the machine should be maintained. Required tasks
are spelled out in operations and maintenance manuals supplied by the manufacturer
and are often part of warranty provisions. |
| Along with doing PM tasks, the maintenance
department must keep good PM records to make a claim on the warranty. The
PM tasks must be done by trained technicians who are empowered and will
look for additional problems without being directed and monitored through
every step. Good maintenance technicians can be developed through the use
of good management tools: standards, training, good pay, rewards, appreciation,
and management interest and concern. These practices increase the chances
to achieve high reliability. |
| Good technicians use all their senses to diagnose
potential problems and then, on their own initiative, take the necessary
action before failures can occur and cause downtime. |
| For example, a mechanic assigned to repair
a small metering gauge found debris in the metering vanes and cleaned it
out. He could have stopped right there because he had accomplished his
assigned task, but he went further. He traced the supply line and found
two strainers that had not been serviced properly. He commented on the
work order that the strainers should be serviced more frequently by trained
personnel. He had a sense of reliability and understood the "system process."
(There was no chance of this debris entering a finished product.) |
| The second step |
| To recap, preventive maintenance satisfies
warranty provisions and provides valuable insight into potential problems
that could cause downtime. This principle leads to the next step: keeping
records of PM-generated, failure, or project work. All machines fail or
have minor problems corrected by conscientious workers. Their labors must
be documented, usually into some sort of computerized maintenance management
system (CMMS). At this point reliability concerns are often neglected.
The RIRO role is forgotten--reliability in, reliability out--to plagiarize
the old adage: garbage in, garbage out (GIGO). |
| For the information in maintenance history
files to be used for reliability purposes, it must be put into the system
in a manner that permits extraction and analysis. Our approach to this
situation is to use the concept of reliability oriented analytic description
(ROAD). Simply put, get on the ROAD if you want to achieve high machine
reliability. It sounds corny, but it works, and that is the bottom line. |
| The ROAD description contains basic information
for making valid conclusions involving machine reliability. All technicians
do not have to know the detailed mathematics of reliability engineering,
but if they prepare good, sound descriptions, the data will be available
to personnel who have those skills. |
| The description consists of several interrelated
parts, written in standard English and in the following prescribed order
and format. Each part has its own formula. |
-
What was done? Use the past tense of an active verb to tell
what was done to an object in a detailed technical sense. Examples: "Replaced
two outer bearings on the infeed conveyor." "Rebuilt #15 snift valve."
"Inspected docks in the warehouse area for safety compliance." The What
formula: verb in past tense plus a specific object.
|
-
Why was it done? Explain the reason for the work. State why
this particular object failed or describe the failure mode. Examples: "(Lack
of lubrication caused races to wear.)" "(Side support was broken by fork
lift.)" "(Pump impeller was worn from excessive cavitation.)" Using parentheses
around the cause is a useful convention to help identify the statement
as a "cause" when the problem is analyzed in the future.
|
| Further, most CMMSs permit the classification
of work by Cause Code; a cause code should be used in addition to the full
cause statement. This code will assist in locating and collating work by
cause. However, cause codes are not a substitute for the full cause statement
because codes are too narrow and restrictive to provide the necessary insight
into what is actually happening to the basic machine. |
| Cause codes alone do not solve the analysis
problem; no one person can reconstruct what happened from memory alone
just by using the codes. Therefore, referring to a failure analysis report
is appropriate and helpful. In addition, key words can be inserted in the
descriptions to help in classification, future recovery, and report preparation.
Some possibilities are "OSHA," "Fire," "Safety," "ISO 9000," or just "ISO,"
and "TQM." Searching the description files for key words will produce reports
of special significance to the insurance agent, the safety manager, or
the risk management analyst. But the RIRO formula applies: you have to
put it in to get it out. The Why formula: (Cause) in parentheses. |
-
When was the work done? The standard repair work order form
provides the date the work was done and the work time. This information
is fine for work-time accounting, but it does not satisfy the reliability
engineer. He needs to know when the machine failed in terms of operating
hours. For example, "The coupling failed after 9053 hours of operation."
For this type of data to be collected, machines that need reliability analysis
must be equipped with hour meters or counters. These meters are relatively
inexpensive and can easily be mounted on most machines.
|
| For example, (OH 0040049) reports that an
event occurred at 40,049 hours of operation. A conscientious effort to
collect and report this information will enable the reliability engineer
to calculate an accurate MTBF rate for the equipment. This result can then
be used to determine realistic reliability evaluations based on concrete,
measured data. In turn, these data can be used to verify and justify warranty
claims with the manufacturer. The data also can be used for setting specifications
for new machines and negotiating price and warranty provisions. The When
formula: (operating hours, OH) in parentheses. |
-
Where was the work done? On the line? In the maintenance
shop? By an outside vendor? Example: "Re-installed gearbox rebuilt by XYZ
Vendor on the incline conveyor." This type of data is recorded by exception
and is used only for out-of-normal cases. Most work will be done on the
line by your own personnel, but in 6 months it could be useful to know
that a particular job was done by a certain vendor away from the plant.
Again, RIRO applies. The Where formula: an exception item; use when needed.
|
-
Who did the work? This key element is often overlooked in
reliability analyses. The information is usually available on the work
order but is seldom entered into the historical database as part of the
work description. Reliable maintenance workers are associated with reliable
machines, and vice versa. Documenting this key information produces personnel
insights often overlooked in the maintenance business. A simple way is
to record the initials of the workers on the job: "(PRS)," "(WWB)," "(RFP)."
The Who formula: (initials of technicians doing work) in parentheses.
|
-
How was the work done? Any special events or problems encountered
in the process of completing the work should be recorded. A full detailed
report is not necessary, just a note showing that a particular job involved
a special situation. Examples: "Job delayed by power outage in power plant
failure." "Parts truck had an accident while delivering needed sprockets
and chain." The How formula: an exception item; use when considered appropriate
and useful in the future.
|
| It is not a simple matter to get maintenance
workers to change their habits and adopt a new and very structured format
for writing descriptions. They must be convinced that the effort will produce
reports that will help them do a better job. One approach is to remind
them that they are, in effect, writing history for some future research
analyst. Their perspective can be: Will this description make sense to
someone else 12 months from now? 24 months? If not, they should do it over
again until it does. Getting good technical descriptions is tough work
and requires training, coaching, and patience. (See accompanying section
for some examples.) |
| Disciplined PM by trained, dedicated workers
can put a plant on the path to reliable machines and equipment, the foundation
of higher production rates. A systematic work-done program, with work recorded
in the ROAD format, will provide the essential data for reliability analysis,
regardless of the CMMS used. In turn, this analysis can provide even higher
levels of reliability and machine uptime. In short, RIRO rules in any of
these systems. The next level of maintenance, predictive maintenance using
special instruments, can be enhanced with a good solid PM program backed
with descriptions in the ROAD format. Get on the ROAD to higher productivity,
safety, and profit. |
| EXAMPLES OF USEFUL ROAD
ENTRIES |
-
Rebuilt #15 filler valve in shop and installed in Line 3 filler. (O-ring
seals failed; it appears that the material just disintegrated) (OH 012232)
(CBB)
|
-
(OSHA) Installed new protective guards on V-belt drive. (Old Plexiglas
guards cracked and discolored) (OH 23371) (ABC)
|
-
Cleaned and tightened all electric terminals to controller. Applied conductive
antiseize compound. Reset breaker. (All connectors were loose and corroded;
some showed hot on thermo-gun) (OH 01233) (PAW) Requested task be put on
PM schedule to check semiannually.
|
-
(ISO 9000) Recalibrated seamer to manufacturer's specs. (Variations too
large) (OH 2003) (GJ)
|
-
(EPA) Recleaned all drains to ensure free flow of water. (First cleaning
did not get dirt in traps) (AFF) Requested task be put on PM schedule for
monthly inspection and cleaning.
|
-
(Fire) Inspected spare parts bin for fire safety. Ordered new fire-proof
container for rags. (AMP)
|
-
Replaced D-50 sprockets and Acme chains on the Jiffy Packer. (Worn out
in use) (OH 55020) Replaced with noncorrosive, sealed O-ring chain. (CMM)
New chain costs twice as much as old chain, but expected long life and
reliability suggest that expense will be worth it.
|
