Reliability Design of Mechanical Systems

This chapter will be discussed with the necessities of new reliability methodology such as parametric accelerated life testing (ALT) in the established product developing process. By adopting new technologies of product in compliance with customer needs from marketplace, mechanical engineer tries to design new structure and mechanism embodied with the sophisticated features. By doing so, the modern mechanical products might survive in global competition. They are often required to have higher performance and reliability for the necessary intended functions, although the product cost and developing time reduce gradually. Mechanical products therefore have faulty designs in the structure configuration. Repeated loads or overloading in the product cause structural damage and finally reduce its lifetime. A failure of products with rare possibility might suddenly arise in field. Because new product has faulty designs in the speedy development time, there is the presence of risks on product recalls at all times. The design process of company requires new assessment methodology of reliability in lifetime. It includes discovering the design defects, correcting them, and checking if the lifetime target of product is achieved.

This chapter will review the historical product recalls including natural hazard and the methodology of its reliability assessment that were developed in the last century. Based on product specifications, engineer would develop new mechanism and its structure. In marketplace product recalls frequently happen. They come from the inheritance design defects in the problematic parts and are determined by the lifetime of product. If product is subjected to repeated loads or overloading and there is faulty design, product failure suddenly arises in its lifetime. To prevent it, engineers in the previous century have developed new reliability concepts such as the bathtub curve, Weibull analysis, data analysis, and the others. For instance, the frequent derail accidents of railroad in the early of nineteen century started the research for its root cause and made the S-N Curve. The chronic failed vacuum tube in the WWII created the bathtub curve. As NASA developed for the space shuttle program in the mid-sixties, FMEA, FTA and Weibull analysis for reliability testing today have been widely used in company. Now since Integrated Circuit (IC), transistor radio and TV in the late of 1960s are introduced, Physics Of Failure (POF) become more important tools to analyze the failure mechanics in product. However, in the field of mechanical/civil system, representative POFs were still fracture and fatigue. As improperly choosing shape and material in the design process, product has faulty design—enough strength and stiffness in the final structure of product. As a solution mechanical engineer should find the problematic parts by reliability testing method and modify them before product launches in market.

To better understand the parametric ALT that is a core of this book, this chapter will briefly review the basic definitions of reliability engineering that can be used widely for reliability testing. It consists of bathtub, reliability index, fundamentals in statistics and probability theory, statistical distributions like Weibull, and experimental design. When product is designed, engineer knows if final design has problems and is satisfied with the reliability target. Engineer should fully recognize the basic concepts that reliability testing is required. From customer’s standpoint, reliability depends on the product design and can be explained as two separate concepts—product of lifetime and failure rate. Engineer for reliability theory may feel complex because it requires an extensive concepts of probability and statistics. To conduct the reliability testing of mechanical product and obtain the reasonable test data, mechanical engineer should understand the modern definitions that can be used. When product is subjected to repetitive stresses (sole factor) and there is design faulty in it, product will fail. However, there is no current methodology because product failures rarely happen in its lifetime. As an alternative, we might suggest parametric ALT in Chap. 8.

This chapter will deal with concepts of failure mechanism, design and reliability testing. Important design aspects of mechanical structures that implement the intended function and have its mechanism in the design process have a good quality—stiffness and strength for its own loading. Product requirements on stiffness, being the resistance against reversible deformation, may depend on their application. Strength, the resistance against irreversible deformation, is always required to be high through the design of product shape. If there is faulty design in the structure where the loads as an uncontrollable noise factor are applied, product will suddenly collapse in its lifetime. The failure mechanics of mechanical product is facture and fatigue—main Physics Of Failure (POF), which can be characterized by the stress (or loads) and materials on the structure. As modifying the shape and material in the structure, engineer would improve the faulty designs and increase its lifetime. These activities are called design. To improve the quality (or design) of product, (qualitative) method like FMEA, FEA, and Taguchi method will be established in the last century. On the other hand, quantitative method using reliability testing is still developing. The final goal of reliability testing is to discover the design problems and reveal if the reliability target for product is achieved.

After mechanical products are designed, we can find the structure and their mechanism. In operation the mechanical/civil systems will work interactively and is subjected to (random) loads. This chapter will discuss how to analyze the loading characteristics of that is the root cause and dominant factor of the product failure. Product has their own particular structural loads in field. Mechanical engineer is designed to have the proper shape and material of mechanical structure with enough strength and stiffness. It therefore can withstand the effects—deformation, vibration, etc.—of system loads in its lifetime. If not, a typical pattern of repeated load or overloading may cause a small portion of structural failures suddenly in its lifetime. Rare possibility such as product failure should be assessed through parametric Accelerated Life Testing in the design phase whether final design of product is properly designed. As first step, there is a modeling of the dynamics systems that can be implemented by traditional method like Newtonian or bond-graph. The time response of mechanical system for (random) dynamic loads represents from its modeling. If simplified and counted as a sinusoidal input for mechanical (or civil) structure, the rain-flow counting method and miner’s rule can enable engineer to assess the system damages—fracture and vibration—for dynamic loading. However, because there are a lot of assumptions for this analytic methodology, it is exact but complex to reproduce the product recalls due to the design failures. So we should develop the final solutions—experimental method like parametric ALT that will be discussed in Chaps. 8 and 9. Load analysis also will be helpful to figure out why the failure of problematic parts in mechanical engineering is revealed.

This chapter will demonstrate the fluid motion and mechanical vibration of product. When airplane is in taxi, take-off or landing, a variety of functions such as airframe, wings, fuselage, engine, control, etc. are required. Especially, fluid mechanics is the critical area that studies the effect of forces on fluid motion. Due to fluid motion, they might be repetitively subjected to (random) loads and make problems—fracture or vibration. To withstand their own loads in airplane, mechanical structures like wing are designed to have proper stiffness and strength. If not, airplane will abruptly cause some problems in flight. To experimentally assess the structural effects of fluid force, engineers should understand the basic concepts of fluid mechanics—viscosity, boundary layer, etc. And we also will discuss the mechanical vibration which happens when system is applied by undesired force—imbalances in the rotating parts or the movement of a tire on a gravel road. Unless isolated by proper stiffness design, system will create unwanted sound or waste energy. Ultimately mechanical system will fracture. To prevent it, engineers might find and correct the design faults of mechanical products. Therefore, parametric ALT might be a kind of engineering solutions.

This chapter will discuss the effect of structural material that is subjected to (random) loads. Fatigue fracture catastrophically occurs in product lifetime when there are stress raisers such as holes, notches, or fillets in design. To understand the structural damage, engineer might understand the basic concepts of mechanics of materials—stress, mechanics of material, deformation, slip, fracture, and fatigue. To withstand their own loads, mechanical structures are designed to have proper stiffness and strength. Requirements on stiffness, being the resistance against reversible deformation, may depend on their applications. Strength, the resistance against irreversible deformation, is always required to be high. If improperly designed, small portions of product will suddenly fracture in its lifetime. To understand the design of product—automobile, bridge, skyscrapers, and the others, engineer should figure out why the mechanical system failures will happen. Through the current reliability methodology engineer still doesn’t know whether product design overcomes (random) loads in its lifetime. For example, the failure of mechanical system like aircraft wing during a long flight can occur in short time or tens of thousands of vibration load cycles. To assess the product damage, we need new reliability methodology like parametric accelerated life testing in the design process. And we also will deal with the mechanical corrosion in the later section.

As a new methodology for reliability design of mechanical system, parametric Accelerated Life Testing (ALT) for different modules will be introduced in this chapter. It consists of parametric ALT plan, generalized life-stress failure model with a new effort concept, acceleration factor, and sample size equation. As applying the accelerated loads to the mechanical structure, the weakest parts in product will reveal. Mechanical engineer can modify the product design to have enough strength and stiffness. Engineer therefore is to confirm whether the final design of mechanical system meets the reliability target. However, there are pending questions—actual testing time and sample size—how to carry out reliability testing. If a few of samples are selected, the statistical data accuracy for reliability testing will decrease. If a sufficient quantity of parts is tested, the test cost will demand considerably. Therefore, the best solution that can decrease the testing time and the sample size is the accelerated life testing that is based on the load analysis. If so, the weakest parts in product (or module) will be exposed. Engineer can solve the problematic designs by correcting them. To save the testing time and decrease sample size, parametric ALT is a kind of solution. The accelerated factors could be found in analyzing the load conditions of real dynamics system. It also requires deriving the sample size equation with accelerated factors. Therefore, if reliability target is assigned, we can carry out reliability testing.

Final mechanical products through design process are designed to have no enough strength and stiffness in product lifetime. Even though properly designed, mechanical products have faulty designs. By parametric ALT, engineer might find and correct them. At that time it is important to have deep insights how to apply the reliability testing. In this chapter, parametric ALT and its case studies will be applicable to a variety of mechanical product—automobile, construction equipment, appliance, airplane, and the others. Parametric ALT is a kind of methodology to confirm if mechanical products in the reliability-embedded design process satisfy the reliability target. The parametric ALT has the following procedure: (1) setting reliability target, (2) determining the accelerated factor, (3) if sample size is given, the parametric ALT will be carried out to the testing time required to meet the reliability target (or its specification). By identifying the faulty designs and modify them, new product will confirm if the reliability target achieves.

This chapter will discuss the concept of system engineering that will be modern engineering. Mechanical product is developing under the principle of system engineering because it requires a lot of functions from customers. Product reliability becomes one of the product requirements in system engineering. So when mechanical system with the sophisticated technology put into plan, product reliability in the established design process should be implemented with reliability methodology like parameter ALT. If not, new product will be faced with quality problems. To settle down them, company will pay the quality costs.