Reliability engineering

[noisy]

So... What does a reliability engineer do then? I've seen this job title on loads of different vacancies, but I haven't a clue what would be involved.

Any help gratefully recieved.

sarcastic answers may or may not be taken with good humour

[phd]

Not an expert - hopefully Genghis the Engineer might chip in at some point - but Reliability Engineers usually work within the Technical Services or Quality Assurance Dept. of a JAR 145 maintenance organisation and collect and analyse a lot of data both from the Aircraft Flight Data Recorders and from other sources such as condition reports on worn or failed components in order to monitor the continuing reliability of aircraft systems and components. If trend analysis shows that things are wearing or failing sooner than expected, or perhaps are actually lasting longer than expected the maintenance schedule for an aircraft can be amended to optimise the maintenance intervals. This can both help to prevent critical failure modes developing within a system unexpectedly, which could lead to a serious accident, and can save the airline money if unnecessary maintenance can be minimised.

[Lu Zuckerman]

It depends on which end of the stick you are looking at.

What PHD has described is the function of a reliability engineer working for an airline. Although some of these same responsibilities rest with the reliability department of the manufacturer of aircraft but they only apply after the aircraft has entered into service. I will come back to this when I have more time in order to clarify the responsibilities of the Reliability engineer relative to the design of the aircraft or whatever system he/she is working on.

OK, I'm back.

The initial responsibility of the Reliability engineer is to make an assessment of the reliability of a component, a sub system, a system, and ultimately the top level whether it is an aircraft or a ship as the procedures are the same. To do this the R engineer will model the system and apply failure rates to each element of the component, sub system or the system starting from the bottom up. Depending on specification requirements he/she will work down to the smallest piece parts like a nut or a bolt or a diode. Once the reliability level has been established the R engineer will then perform a Failure Mode Effects and Criticality (FMECA) analysis for the same component, sub system or the system.

The FMECA establishes the relationship to a failure of a given part to its’ effect on the component, the sub system or the system and ultimately to the top level (aircraft or ship or whatever).

The two analyses are related as the failure rates derived during the reliability modeling are used to show the failure rates for each element of the FMECA. The FMECA feeds into the Safety Hazard analyses but that is another story.

The design spec issued by the procuring authority will have the required reliability level for the top level to include Mission Reliability, Operational Reliability and any other levels that are critical to the operation of the top level element.

Here is where the system goes to hell in a handbasket. Once the top level reliability has been established the chief reliability engineer will break the top level system into its’ systems and subsystems down to the components within the subsystem. He/she will determine those systems that are most critical and apportion the reliability for those elements and anything that is left over the lower level systems and components will get what is left over. In other words the reliabilities for all of the sub elements of the top system must add up to the reliability for the top-level element.

Those components that are sub contracted to other vendors will have a reliability requirement that is derived during the apportionment and the vendor will apportion the reliability levels for the piece parts that make up the component.

Now the plot thickens. Where do the reliability figures (failure rates) come from. They can come from known failure rates reflecting components previously manufactured by the vendor, they can come from known databases that contain questionable figures (but these are government documents) so they are acceptable and so on down the line to a best guess (engineering judgement). The reliability engineer will pick the best numbers that will allow him/her to meet the reliability requirements and those numbers taken from the government database are allowed to be manipulated to convert numbers from one unrelated application to the component being analyzed. In other words the figures that are used to define the reliability in most cases have no relationship to the parts used in the component. The top number for the component will in all cases meet the reliability number specified in the design spec. This number is then provided to the airframe manufacturer who in turn plugs it into his/her calculation and the subsystems and systems are made up of components that meet the design spec but have no relationship to the true reliability of the top level.

Then, to top it off in military specs the contractor must show reliability growth. To do this the system reliabilities are recalculated using better numbers.

Reliability engineers in many cases are mathematically driven and have little background in engineering design or in operations to include maintenance. They view the subsystems , systems and the aircraft as a mathematical entity with the mathematics being used to show compliance with the spec.

How does that make you guys feel out there when you strap yourself in?

Oh yeah, Im a Reliability, Maintainability and Systems Safety Engineer but I have a strong background in maintenance and I have an A&P license.

[noisy]

Thanks for the answers guys, sound like a really fun job

At least I have some inkling of what is going on now.