14.05 FMEA

Categories: Procurement

Introduction
Failure Mode and Effects Analysis (FMEA) was introduced into the aerospace and automotive industries in the 1960's and 1970's.  Design FMEA is a framework in which decisions can be made having full regard for the outcome of those decisions in terms of:

• Failures that might occur
• Risk of failure occuring
• Consequence of failure

Manufacturing FMEA is also practised in some manufacturing industries.  This is a way of considering how a product might be wrongly assembled in terms of:

• What wrong assembly might occur
• Risk of incorrect assembly
• Consequence of wrong assembly

This information is used to redesign products and their components so that incorrect assembly is either impossible or readily detected.

Aspects of FMEA are covered in BS5760: Part 5:  'Guide to failure modes and criticality analysis' and BS5750: Part 0: 'Guide to quality management &  quality system elements'.

FMEA has been used in the construction industry on a few high risk civil engineering projects.  However, greater emphasis is being placed on risk management in the building industry and FMEA can offer a way forward for the construction and cladding industries.
 

Definitions
FMEA uses the following definitions when discussing the consequences of decision making:

Failure occurs when a component or assembly does not meet, or function, with design intent.

Failure mode is the manner in which a component or assembly failure occurs, or the way in which a component or assembly does not meet design criteria.

Cause of failure describes the discrepancy leading to failure.

Effect of failure is the result of a particular failure as experienced by the user.
 

FMEA procedure
There are three stages to the implementation of FMEA to a cladding project.  In order they are:

• Identify failure modes and corresponding causes and effects,
• Analyse risk by ranking failure modes according to occurence, severity and non-detection,
• Modify design or construction process to reduce risk of occurence.

Identification of failure modes
Failure modes are identified from a knowledge of previous product and construction failures.  They may be based on knowledge of specific materials and products or they may be based on known failure mechanisms such as the weathering of polymeric materials by U.V. and heat.  The failure modes may be identified at a component level but they should also be identified at an assembly level for it is at this level that unforeseen interactions between components occur.

The identification of failure modes should yield patterns of failure where component failure may initiate assembly failure.  Assembly failure may then constitute a serious failure whilst consideration of the component alone might not at first appearance be a serious failure.  For instance the inadvertant sealing of a drainage hole may manifest itself in the form of a failed glazing unit that has been sitting in a wet glazing rebate.

Identification of failure modes at a material or component level can be facilitated by the use of checklists.  A checklist for all cladding materials, components and assemblies would be extensive and it may be more appropriate to have check lists that relate to each material and its use with other materials.

At a whole wall level potential failure modes are:

• Water penetration
• Excess air infiltration
• Lack of fit
• Insufficient durability
• Condensation formation
• Programme overrun
• Structural failure (excess deflection and breakage)
• Budget overrun
• Difficult maintenance
• Inadequate thermal insulation
• Insecurity against intruders
• Inadequate acoustic insulation
• Appearance (colour matching, gloss etc.)

Assessment of risk
The risk posed by any failure depends on:

• Probability of occurence
• Severity of effects
• Risk of non-detection

It follows that attention should be given to reducing risk where there is the greatest probability of occurence combined with the greatest severity of effect and the greatest chance of non-detection.  Risk will normally be reduced by reducing the probability of occurence of a component or material failure or increasing the likelihood of detecting the fault during manufacture and construction of the wall.

Where these three factors of risk can be measured or ranked then it is possible to multiply the three factors to obtain a risk priority number (RPN).  These numbers are then used to rank the combined probability that a potential cause of failure will occur and will lead to a predictable failure.

Risk of occurence requires documented evidence of failures.

Severity effects may be ranked on a scale of 1 to 10 as follows:
 

Risk of non-detection will depend on the ease with which something can be inspected and QA systems in place.  Items of work that cannot be inspected easily such as the cleaning and priming of the substrates in a wet applied sealant joint will rank highly whilst errors that are obvious to any observer such as the colour match of different components will have a low ranking.
 

Practical FMEA
In the automotive industry FMEA is readily carried out because failures result in warranty claims and /or purchase of replacement parts.  Motor manufacturers therefore operate in a closed loop where they mass produce cars and receive feed back that enables them to modify the product or manufacturing process.

In the construction industry there is little feedback about failures and little knowledge about the risk of component or material failure. However, FMEA need not be complicated to reduce the incidence of facade failures.  A study by the CWCT showed that of the failure modes listed above testing is usually carried out for only three of them and that the occurence of those failure modes was:
 

To concentrate on reducing water penetration would be a good start to improving cladding.  The causes of water penetration for a large series of walls tested by one test house were:
 

A major cause of failure is poor workmanship.  Many walls, when tested for water penetration resistance, fail the initial test and require further work before they resist water penetration.  For a series of sixty tests the following failures were recorded:
 

Simplified FMEA
Designers can carry out an informal form of FMEA by considering the consequences of failure and risk of occurence for a particular failure mode.  This image shows a plot of occurence against severity of failure.  The failure modes of structural failure, water penetration and unacceptable finishes are shown on the chart.  The reader is invited to place other aspects of wall performance on the chart.  Walls will be improved if designers, makers and erectors pay more attention to those aspects that lie furthest from the origin of the chart.

FMEA deals with concepts of risk and ranking.  In practice many failures occur because things are done that inevitably lead to failure.  For instance using sealants that stain adjacent stone work and using gaskets with unsealed joints at the mitres.  A simple form of FMEA could be implemented in which some things are not done because they lead to failure and some things are always checked because they might lead to severe failure.

At a very simple level if two solutions to a design problem are possible and one is known ti involve less risk of failure then it should be the design solution.