Manufacturing in the aerospace sector has undergone a profound transformation in the last decades towards ever higher automation levels. It began in the scope of single parts manufacturing, and has continued spreading to other areas such as assembly operations and functional testing.
In fact, a high level of automation has already been achieved in single part manufacturing technologies (automatic lay-up operations for composite parts, numerically controlled machining for metallic parts, etc.), enabling significant cost reductions and an increase in quality in the early steps of the supply chain.
In contrast, the weight of manual activities is still high in assembly operations, which account for between 25% and 75% of total manufacturing costs in the aerospace industry. The opportunity for savings is therefore significant, if improvements are made to these operations, something for which automation has demonstrated that it is a consolidated option. This article will explore some of the benefits associated to the automation of assembly processes in the aerospace industry, and some of the obstacles facing the team in charge of this mission. We will focus on the evolution of the drilling and riveting of structural joints, which are indeed the ‘seams’ of an aircraft. Certainly these operations require such exquisite attention to detail, that the challenge posed by their automation could be compared to that of developing robotics for haute couture.
Arguments that lead to considering the automation of a process fundamentally fall in two categories: cost reduction and the increase of safety at work. The cost reduction must be endorsed by a detailed feasibility plan, including detailed analyses of the product, the assembly process to be automated, and the potential savings. A machine does not suffer from physical fatigue, and can therefore develop repetitive tasks at a steadier rate than a person, and with constant results. This reduces the number of failures and the costs due to repairs, and also increases the quality of the operation. In addition, the automation allows relieving the highly specialized staff in the aerospace industry from repetitive manual tasks.
There is a double benefit to employing workers in tasks of higher added value: on one hand, the quality of the final product increases, and on the other hand, this fosters professional development towards other activities of greater responsibility and added value. Finally, assembly operations may lead to unfavourable ergonomics, particularly when working in small areas, applying repetitive efforts, or handling certain toxic products. The automation of these operations increases the safety of workers and is one of the most powerful arguments when preparing a feasibility study for an automated application.
The automation of an assembly operation requires a significant financial investment, which justification must be carefully analysed to avoid prejudices based on false beliefs. For example, the typically low cadences of aerospace programs are not necessarily a hurdle.
The decision to automate should be based on the number of times that the operation is repeated, whether it is performed once on every one of a great number of equal products, or it is performed many times on a single product. On the other hand, the assembly is one of the last links of the complex production chain of an airframe, and it accumulates all the defects coming from the previous steps: manufacture of elementary parts, formation of subassemblies, on-time delivery, delivery conditions, etc.
This has traditionally forced to include additional tasks during the assembly which are very difficult to automate, such as inspections or small adjustments. In this sense, and strictly speaking, the only flexible assembly is the manual one. An automated assembly line needs a robust supply chain with a number of limited and typified failures. It is necessary to trade in some flexibility in favour of higher repeatability and standardization. Another traditional disadvantage of automated lines is the high cost associated to start-up (Figure 1) and maintenance (Figure 2), although this has been partly mitigated with the arrival of electronic control technologies. An automated line is very sensible to unexpected defects, and the start-up phase usually reveals errors undetected in previous steps of the production chain.
In addition, the fine-tuning phase requires the support of highly specialized workers and engineers, with a high cost associated. And finally, one of the most difficult obstacles to save when facing the automation of a process is the resistance to change of their very own users. On one hand, managers are not always aware of the recurrent costs of not automating, and on the other hand, it is difficult to change the common perception in the workshop that the machine is an enemy that destroys jobs, instead of an ally that will allow dedicating the people to safer and more valuable activities.
In airframe assemblies, drilling and riveting is a repetitive operation that consumes great amount of resources. Grids or joint lines that constitute these seams typically consist of tens of thousands of holes and rivets. In this type of structural joints, positioning and form tolerances for the holes are respectively in the order of magnitude of tenths and thousandths of millimetres.
The surfaces to be joined must fulfil very demanding requirements in terms of finishing and cleanliness. Traditionally, joining of elements such as skins and fairings to the internal aircraft structure formed by spars and ribs has been performed by manually drilling thousands of holes in each element (Figure 3.a). This is a highly repetitive operation, complicated in some cases by the strenght and stiffness of the aerospace materials. These are two of the operations that have benefited the most from automation strategies in the assembly lines.
The progressive automation of these operations stems from an understanding and detailed analysis of the whole process. Each step has been taken only after the previous one was considered sufficiently mature and controlled.
A first step in the automation of drilling and riveting operations was the development of semiautomatic drilling machines, equipped with pneumatic and hydraulic modules to aid the rotation speed and feed movement of the drill bit.
These machines – still in use today – are secured to heavy and bulky drilling templates, hence relieving the worker from the strenuous tasks of holding in position and pushing the tool. Instead, the operator only has to move the machine from one position in the drilling grid to another, and press a button (Figure 3.b). Once this technology was established successfully, the next step was the introduction of the first fully automatic drilling machines, capable of positioning the tool and conducting the whole drilling operation. Robots of various morphologies have been used in aerospace assembly factories – gantry, parallel kinematics, or linear arm –controlled by pneumatic logic at the beginning, and later by electronic controls.
Especially when drilling modern materials, these robots had to be structurally very rigid in order to achieve sufficient accuracy. This translated into hefty structures and robust guides on which to move their drilling units (Figure 3.c). The next step in the automation was equipping the drilling units with additional modules so as to completely rivet the joint. A fully-automatic drilling and riveting unit must include means for deburring the drilled hole, measuring the length of the rivet to be installed, applying sealant, inserting the bolt, and closing the stack-up.
All these functionalities significantly increase the size and weight of a drilling and riveting unit. The only way to guarantee tolerances and accuracy is to increase the structural stiffness of the whole system even further, which means more robust and heavy elements such as guides and transmissions, and expensive special foundations. New challenges In recent years, advances in fields such as artificial vision techniques, pattern recognition, materials and clamping systems, have opened the way for a more versatile type of drilling and riveting system. Little by little, the huge machines designed ad hoc and dedicated to a single product are giving way to flexible cells, using smaller off-the-shelf robots that do not need special foundations, and can be easily reconfigured for use on different products. Making an exercise of imagination, assembly stations in the future will have human operators and robots of different shapes operating together in the same area, collaborating and interacting – prototypes already exist of arachnid walking robots, snake-shaped crawlers, or robots articulated like a human arm. These future tendencies are shared by specialists in meetings like the one organized in March by the CDTI and the CATEC in Seville. Aerorobot 2010 reviewed the state-of-the-art of robotics applied to the aerospace industry, focusing on the role that Andalusian aerospace has played in its evolution. Andalusia is today at the forefront of the automation of aerospace assemblies, being home to the Airbus Centre of Excellence for the automation of the assembly of Horizontal Tail Planes at European level, and to one of the most advanced Final Assembly Lines in Europe. This success is based on the know-how developed in a number of aerospace programs industrialised in the region through time. The Mercure, SAAB 2000 and Dornier 728, the light and medium military transports from the old CASA, or the first Airbus programmes taught some of the lessons applied today to the A380 and A400M programmes. And the future is already here with the A350 and the new Airbus developments.
Rubén Carvajal Vázquez, and Manuel Heredia Ortiz. Authors of http://aergenium.es, first web dedicated since 2008 to the aerospace industry in Andalusia, Spain.
If you want to know more about the topic of this article, visit the web http://sabermas.aergenium.es, where you can find more information, support material, photographs, videos and links of interest.