Failure Mode Effect And Criticality Analysis For Risk Analysis (Design) And Maintenance Planning (Exploitation)

Transcription

1 Failure Mode Effect And Criticality Analysis For Risk Analysis (Design) And Maintenance Planning (Exploitation) J Lair & JL Chevalier CSTB Sustainable Development Department Environment Durability France Summary: The ISO (ISO, 1998) set of standards Service life planning is developing an integrated design framework for buildings and constructed assets, incorporating technical, economic and environmental aspects. In this context, CSTB has developed a risk analysis and maintenance planning tool, based on the use of the Failure Modes and Effects Analysis (FMEA). It includes: first, a structural analysis which allows us to identify morphology, topology and physico-chemical constitution of the product and its components, second, a functional analysis leading to the identification of the various functions ensured by the product and its elements, as well as a description of its environment, These first two steps lead to the behaviour modelling of the product. finally, an FMEA in order to identify failure modes (exhaustive search for the behaviours, degradations and failures of elements), their causes and effects, taking into account the potential problems and errors which could occur during the process. This step leads to the identification of degradation and failure. We thus can build the failure modes (or failure scenarios). The methodology proposed simplifies each step of the approach, and provides us with a graphical representation of the product behaviour. Keywords. Durability, FMEA, failure modes, risk analysis, maintenance 1 INTRODUCTION 1.1 Context Non-quality or poor quality problems in the building domain (representing 10% of the French building turnover, i.e. 6.5 to 7.5 billion Euros according to Le Brigand (1998)), the part of maintenance and operating stages in the cost of a building (60 to 70% of the overall cost (Perret 1995)), as well as the new aim Sustainable Development 1 (natural resources preservation, energy saving, etc), led us to work out methods and tools for durability assessment. 1 A development which answers the needs of the present generation without compromising the ability of future generations to answer theirs (Charlot-Valdieu and al 1999). 9DBMC-2002 Paper 032 Page 1

2 1.2 Objective In this paper, we present the durability assessment methodology (with reference to (Lair 1999)), and then propose additional developments and applications to the sketched methodology. In particular, we include the tool in the quality approach, during the design stage (risk analysis help for design), as well as for the management of existing buildings (maintenance planning diagnosis / maintenance / repair procedures). 2 METHODOLOGY 2.1 General FMEA has been used since the sixties in the aeronautics and car industry and is thus an efficient tool for the safety and risk analysis of these systems (Leroy 1992). It is intended to check the ability of a system and its parts to meet user s needs and either is generally used during the design stage in order to target and examine weak points, before mass production (Modarres 1993). It allows us to identify the potential future behaviours (success or failure) of construction products. It leads to an exhaustive list of all failure modes, i.e. all problems that might occur: at each stage in the construction process (design, manufacturing, installation, transportation, storage, etc), in service (influence on the product s performance, degradation or failure of one of the product s constituent materials/components constitutive material/element). In the following paragraphs, each step in the methodology will be presented and illustrated with basic examples. 2.2 Structural analysis Principle (Figure 1) The principle of this study is to built an accurate description of the product: by dividing it into elements, and by identifying the constituent materials. Product Elements Materials Figure 1: Structural analysis This first analysis allows us to describe the structure of the product being studied. The following are identified (Baloche and al. 1999): morphology (geometrical shape, dimensions, etc), topology of relations with other objects, physico-chemical composition of its constituent elements and their own description Example: Roofing system The following graph is a representation of a flat roof system (left part of Figure 2), and its corresponding structural diagram (right part of Figure 2). Outdoor and indoor environments have to be defined. They are composed of all potential climatic and use that correspond to the environment in which the product will be in service (Figure 3). 9DBMC-2002 Paper 032 Page 2

3 Outside 6 - Added heavy protection 4 - Roofing membrane 5 - Screen (independence) 3 - Insulation panels 2 - Vapour barrier 1 - RC slab Inside Figure 2: Section of a flat roof system and structural representation Outdoor 2.3 Indoor Temperature (high or low) Temperature Rain, snow, hail, ice Vapour Sun Indoor pollutants Wind Pollutants, weed-killer Vegetation, moss, lichen, etc Shocks and loads (persons) Figure 3: Climatic and use A list of common climatic is defined in order to help the user (generic ). Some specific have to be defined according to the use (chemical products in some industries, etc). 2.4 Functional analysis Principle A product meets users needs. The functional analysis of the needs expresses the needs in terms of functions. We also have to identify the functions of each element (i.e. the role of each element in the overall behaviour of the product) Example As an example, a flat roof system should ensure two major functions: thermal insulation (against temperature gradient), watertightness (role of roofing membrane against rain, snow and hail). In addition to these functions (user s needs), we have to consider the function of each element: the independence screen (element number 5) of a roofing system limits the strains in the roofing membrane, due to the movements of insulation panels (especially thermal dilatation), resistance of all elements to the environment (temperature, sun, rain, etc). 2.5 Behaviour analysis Modelling principle A product is defined as an order of elements and materials that ensure a set of functions. 9DBMC-2002 Paper 032 Page 3

4 In the building domain, The product fulfils a function could be expressed as The building product transforms climatic (between input and output). It acts as a filter between two environments, filtering heat flows between outdoor and indoor environments (thermal insulation), stopping water from outdoor (watertightness of a roofing system), etc. But, the same climatic can have an impact on its constituent elements and could involve: modification of the materials properties, degradation and even failure, etc Environment 1 Environment 2 Product Climatic Set of functions Order of elements Modified climatic Figure 4: Product representation We can express the behaviour of elements on the same diagram, showing on the one hand that they fulfil a function, on the other hand that they are likely to be modified and degraded by climatic and use Behavioural graph In this context, once we have identified the structure and the functions of the product, the next step consists in modelling its behaviour. The approach adopted is opposite to the structural analysis: from the behaviour of the materials, we deduce the behaviour of the elements and then the behaviour of the product itself (Figure 5). Behaviour of the product Behaviour of the elements Behaviour of the materials Figure 5: Behaviour analysis For that purpose, a behavioural graph is plotted. The product is composed of n elements, placed in series or in parallel. each block being an element, and each link being a relation between blocks (physical contact, flow transfer, etc). Figure 6 is an example of a very simple graph (only series relations). It ensures a barrier role between outdoor and indoor environment, as a composite wall: OUTDOOR Climatic Element 1 Function Modified climatic Element i Function Modified climatic Element n Function Modified climatic INDOOR Material Material Material Figure 6: Product behaviour Special cases In order to have an accurate representation of the product, additional information is required: cavity (as the cavity between the internal and external glass of a double glazing unit or the cavity in a pipe), interface, either simple contact (added heavy protection on roofing membrane) or fixed interface (as the interface between seal and glazing). They are both represented as elements: in the first case to take into account thermal and acoustic characteristics, as well as flows (air, water, etc) and deposit (condensation, dust, etc), 9DBMC-2002 Paper 032 Page 4

5 in the second case, mainly to take into account problems of compatibility/incompatibility of materials, gluing properties, watertightness and airtightness Nominal behaviour (t = 0) From this graph, we can deduce the nominal behaviour of the product, i.e. we know the response of a product (and its element) to a given set of climatic and use. At t = 0 (without any degradation and considering that the product was correctly installed/implemented), we can identify the initial state of stresses. It is summarised in the following table: outdoor go from the top to the bottom, indoor go from the bottom to the top. OUTDOOR High or low Temp. Rain Snow Hail Ice Sun (UV) Wind Pollut., weed- Vegetation (roots) Moss, lichen, Shocks Loads 6 Added protection x x x x x x x x x x x x 4 Roofing membrane x x x x x x x x x 5 Independ. screen x x 3 Insulation panels x x x 2 Vapour barrier x x x x 1 RC slab x x x x INDOOR Temperature Vapour Indoor pollutants We thus have information on: the various flows in the product (e.g. thermal flow), Figure 7: Environmental stresses Initial conditions the key elements (protecting elements: role of heavy added protection for roofing membrane against UV, shocks, etc), 2.6 Degradation and failure analysis Failure modes (t > 0) Given the nominal functioning of the product, we start working on degradation and failure study. We study the behaviour of the product if the behaviour of a material or element deviates from its normal behaviour. The principle is an iterative study. Step 1: Degradation of elements due to climatic or use We first analyse the influence of initial environmental stresses on each element. For instance, we have to study the behaviour of the roofing membrane towards high and low temperatures, water (rain, snow, hail, etc), ice, various pollutants, vegetation (including moss, lichens, etc) and shocks. The methodology used allows us to take into account not only these single but also the combination of these : combined from the same environment, either concomitant (water and low temperature from outdoor environment freezing/thawing cycles) or successive (sun and high temperature followed by rain thermal shock), 9DBMC-2002 Paper 032 Page 5

6 combined from different environments (high temperature from outdoor and low temperature from indoor temperature gradient in the element). Knowing the element (and its constituent material) as well as the potential stress, we identify the potential behaviours: wind can result aggregates being blown away (from heavy added protection), water on roofing membrane can cause the removal of solvents, high temperatures can involve thermal ageing (hardening), low temperatures can involve embrittling, and make tear easier under load stresses. Step 2: Structural or environmental modifications (degradation or failure identification) The behaviour or the degradation of the first step leads to the modification of the environment or the structure. As examples to degradation and failure, the following was observed: under temperature stress, the roofing membrane gets old. This ageing causes hardening. Although roofing membrane is degraded, it still fulfils its main function: watertightness, moreover it will not be able to withstand mechanical stresses as well as initially. if the heavy added protection has blown away from some parts of the roofing system (failure of the protection of the roofing membrane function), this roofing membrane is no more protected against UV and temperature. This comes down to a new cross in Figure 7 (column Sun(UV), line Roofing membrane). The initial state is updated and becomes a State 2 condition for which we have to study the effect of UV on roofing membrane. Step 3: Degradation of elements due to updated climatic or use Given the modification of structural or environmental conditions (step 2), we now study again the behaviour under new environmental conditions (action of environmental on elements due to structural modifications, mainly loss of protection): photochemical ageing of roofing membrane due to the UV stresses, Again we identify the modification of both structure and environment (step 2), then step 3, and so on For instance, once the heavy added protection has failed since it was not thick enough to protect the roofing membrane, the latter is stressed by UV and temperature (step 2). It will fail by cracking or tear, and then will not fulfil its watertightness function any more. Rain, pollutants, will go through the roofing membrane and reach the insulation panels. OUTDOOR High or low Rain Snow Hail Ice Sun (UV) Wind Pollut., weed- Vegetation Moss, lichen, etc Shocks Loads 6 Added protection x x x x x x x x x x x x 4 Roofing membrane x x x x x x x x x x 5 Independ. screen x x 3 Insulation panels x x x 2 Vapour barrier x x x x 1 RC slab x x x x INDOOR Temperature Vapour Indoor pollutants Figure 8: Environmental stresses State 2 Note: We also take into account construction process problems, i.e. problems occurring during the following stages: design, transportation, storage, installation, use, etc. 9DBMC-2002 Paper 032 Page 6

7 We use the 5M rule to identify the various potential defaults, negligence or errors concerning: Material (quality, incompatibility, surface quality, etc) Manpower (surface cleanness, dimensions, etc), Middle, i.e. environment (temperature, humidity, etc), Means (non-adapted tools, etc), Method (time limit, etc). They intervene as structural or environmental modifications in step 2. The failure of a reinforced concrete element could be due not only to the impact of environmental impact (CO 2, ) but also to vibration, temperature, curing, position of reinforced bars, etc. 3 RESULTS AND LIMITS 3.1 Results We thus obtain: information about the Nominal behaviour of the product in a given environment in order to document reference service life (Sjöström and al., 2001), information on the degradation and failures, expressed as a list of degradation and failures (FMEA table) or a failure tree (with scenarios). In this table are listed for each element, the modes, causes and effects. In column Causes, we distinguish errors and problems occurring during the process (cause type 0), and the different levels of degradation or failure: cause type 1 (initial environmental stresses), cause type 2 (occurring after the degradation or failure of an element), and so on Function Element Modes Causes Direct effects Indirect effects Watertightness Roofing membrane Cracking, tear 1 -Load and wear Cracked membrane Water penetration (outdoor humidity) 1 - Dimensional variation of support (bending) (to insulation) 1 - Adherent ice (compressive stresses) Local lifting 1 Wind (air flow) Slipping 1 - Vapour pressure 1 - Temperature - Bitumen slipping Loss of watertightness Water penetration 1 - Temperature - Excessive quantity of bitumen performance (to insulation) 1 - Overload (slope and friction) Loss of continuity Stress on edge Perforation 0 - Punctual load (installation) 0 - Punctual load (maintenance) 0 - Punctual load (use) Pierced membrane Water penetration (to insulation) 1 - Vegetation Vapour membrane (if roofing membrane fails) Cracking, tear or perforation Refer to the item Watertightness (indoor humidity) Watertightness (Indoor humidity) Vapour membrane Perforation Cracking, tear Thermal insulation Insulation panels Water absorption 0 - Problem during installation (concrete surface finishing) 1 - Dimensional variations of support (RC slab) 1 - Dimensional variations of insulation panels (swelling) 0 Problem during installation (concrete surface finishing) 1 Dimensional variations of support (RC slab) 1 Dimensional variations of insulation panels (swelling) Loss of watertightness performance Loss of watertightness performance 2 Roofing membrane failure Loss of thermal characteristics Water penetration (to insulation) Water penetration (to insulation) 2 Vapour barrier failure Swelling Stress of roofing membrane Figure 9: FMEA table of a roofing system (Extract) 3.2 Limits At the moment, the main limit of the methodology is at this moment the binary reasoning, i.e.: we do not have any indication of time before degradation or failure, i.e. progressive degradation (just consider that they will occur anyway), we do not measure the intensity of the phenomena (just consider that they will occur anyway), 9DBMC-2002 Paper 032 Page 7

8 we do not take into account geometrical aspect as partial degradation (10% of the surface is degraded for example). This limit leads to an exhaustive list of degradation and failures, with obviously some improbable and unrealistic failure modes. Two solutions are currently under study in order to refine the results (select the most relevant modes in terms of frequency or gravity and delete the others): developing an expert method for criticality analysis (Probability x Gravity of effects), introducing qualitative reasoning in order to take into account progressive degradation. 4 INTERESTS AND PERSPECTIVES This tool could be used at the different stages in the product life: Design and installation (risk analysis and maintenance planning), Use (maintenance planning or failure diagnosis). The following table (Figure 10) gives the results and the objectives of this tool. Construction process stage Results Objectives Design Installation Use 1 - Risk analysis Identification of weak points of the product (characterised by a level of criticality) 2 - Quality management 3 - Preventive maintenance 4 Conditional maintenance 5 - Corrective maintenance Identification of critical operations (Installation, use) or critical elements leading frequently to degradation and failure. - Forecasting of potential behaviours in time. - Assessment of the criticality of possible consequences. - Identification of symptoms, warning signs of failure (condition assessment, diagnosis of degradation) - Forecasting of future behaviours (given actual state). - Assessment of the criticality of possible consequences. Identification of failure causes from the observed failure (diagnosis of failure). Figure 10: Different uses of FMEA tool 4.1 Risk analysis This tool is intended to help industrial partners, to contribute to innovation. Improve quality and reliability from design stage by: - identifying problems, - giving priorities. Improve the construction procedures (transport, storage, setting up) and use by: - identifying problems, - giving a check-list of key steps (scale of priority). Improve proactive maintenance procedures by: - identifying problems, - giving priorities, - proposing solutions to maintenance. Improve reactive maintenance procedures by: - identifying problems, - giving priorities, - proposing solutions to maintenance or repair. Improve corrective procedures by: - explaining failures, - proposing solutions to repair. A successful experience with an industrialist was conducted on an innovative cladding system. Two potential failure modes were detected and solved. 9DBMC-2002 Paper 032 Page 8

9 Even if the failure modes are uncertain, it is worth pointing out the fact that the product could fail in use. We then analyse its behaviour more accurately using a traditional procedure (artificial testing, natural ageing, etc). Based on the FMEA study, we know that : Such materials will be stressed by UV By means of traditional methods for durability characterisation, we try to know whether the product is designed to withstand such stresses? 4.2 Quality management FMEA is used as a tool allowing experience and knowledge to be capitalised (as well as the production of information) concerning the frequent defects, errors, negligence occurring during the construction process, expressed in checklists. For the following three maintenance strategies, other considerations have to be taken into account. For example, the degradation could be accepted because of high maintenance costs. 4.3 Preventive maintenance FMEA produces information concerning the key elements, i.e. elements on which the good working of the product depends (top elements in the failure tree). This justifies the updating of existing lists of maintenance operations. Example: Product/element Operation Heavy added protection in a flat roofing system Limited thickness No vegetation Draining pipes in a wooden window Preventing water accumulation (rotting of wood pieces) Figure 11: Preventive maintenance 4.4 Conditional maintenance FMEA is particularly useful for degradation and failure modes with symptoms (warning signs). With a condition assessment of the building and the products (surveying only the critical elements, guided by FMEA results), we identify symptoms, search for causes and propose relevant cure (maintenance or repair solutions). We are thus able to prevent costly and/or hazardous failures by identifying warning signs. Example: Symptoms Cause Solution Water accumulation on a roofing system Obstruction of draining pipes because of bad maintenance Degradation (cracking) of seals in a window Climatic Figure 12: Conditional maintenance Cleaning of the draining pipes Change of seals 4.5 Corrective maintenance (repair) FMEA is used to explain a degradation and a failure, i.e. to identify the top event and the different events leading to failure, in order to propose relevant repair solutions instead of temporary solutions (searching top event). Example: Failure Scenario / Top event Solution Air permeability of a window 1 - Defect in hinges adjustment 2 - Wear of seals 3 - Air permeability Hinges adjustment Figure 13: Corrective maintenance 9DBMC-2002 Paper 032 Page 9

10 5 CONCLUSIONS: FMEA A TOOL FOR QUALITY IN CONSTRUCTION Based on FMEA, the tool developed in CSTB gives elements to assess the durability of construction products. On the one hand, we can improve the reliability and quality of innovative products. With a risk analysis from the design stage, weak points are identified. We can define relevant preventive actions (risk analysis and quality procedure). On the other hand, in a refurbishment/diagnosis context, the FMEA tool allows the in-service follow-up of existing products and supplies information for IMR procedures (Inspection/Maintenance/Repair) such as quality control and maintenance planning (conditional and corrective maintenance), failure diagnosis (capability of detecting degradation, prediction of the future degradation or failures, identification and selection of relevant maintenance or repair solutions). It plays an important part in capitalising on experiments and knowledge within building management. It is a real advantage for architects, owners, project managers, insurers, users... to: Have a guaranteed quality approach, Improve and reveal the quality of products and practices, Optimise design costs, Communicate reliable information between actors of construction, Improve products and practices quality, Optimise setting up and maintenance costs, Decrease environmental impacts at each stage in the construction process, Avoid errors, defects and omissions. In a context where non-quality or poor-quality costs are being added to the general wish to decrease the global cost of construction, FMECA and management tools should have a determinant role. 6 REFERENCES 1. Baloche, C. Chevalier, J.L. and al. (1999). Traité de Physique du Bâtiment - Connaissances de bases. Paris. Ed.: CSTB. 832 pages. 2. Charlot-Valdieu, C. and Outrequin, P. (1999) La ville et le développement durable. Cahiers du CSTB (Livraison 397 Cahier 3106) 3. ISO TC59/SC14 (1998) Buildings and constructed assets Service life planning. ISO/DIS Lair, J. Chevalier, J.L. Rilling, J. (2001). Operational methods for implementing durability in service planning frameworks. CIB World Building Congress, April 2001, Wellington, New-Zealand. 5. Lair, J. Le Téno, J.F. Boissier, D. (1999). Durability assessment of building systems. Durability of Building Materials & Components 8, Vancouver, Ed : NRC - CNRC. 6. Le Brigand, S. (1998) , photographie de la sinistralité. In CSTB Magazine n 119 (Nov. 98): pp Leroy, A. and Signoret, J. (1992). Le risque technologique. Paris. Ed.: Presses Universitaires de France. Que sais-je? n pages. 8. Modarres, M. (1993). What every engineer should know about reliability and risk analysis. New York. Ed.: Dekker. 350 pages. 9. Perret, J. (1995). Guide de la Maintenance des Bâtiments. Paris. Ed.: Le Moniteur. 431 pages. 10. Sjöström, C. Jernberg, P. Frohnsdorff, G. (2001). Interoperational methods for implementing durability in service planning frameworks. CIB World Building Congress, April 2001, Wellington, New-Zealand. 9DBMC-2002 Paper 032 Page 10