https://cdn.mtdcnc.global/cnc/wp-content/uploads/2021/01/30224736/Image-2-640x360.jpg
    Measurement

    Made to measure

    • By Editor
    • January 30, 2021
    • 8 minute read

    “You can only make as well as you can measure” or a variation thereof is a phrase familiar to most people who work in industrial metrology. Attributed to the pioneering 19th-century engineer and inventor Joseph Whitworth at the height of the first industrial revolution, the phrase is as true now as it was in 1840 when his ‘End Measurements’ invention demonstrated precision to ‘a millionth of an inch’ or 25 nanometres in today’s money! Metrology is a fundamental consideration that is often overlooked in manufacturing,  as Dr Christian Young explains.


    B
    emusingly metrology may often be overlooked, yet it is an aspect of modern manufacturing that is essential for creating the high precision parts that go into everything from iPhones to jet engines. Despite this, tell someone you are a metrologist and the most likely response you will get is a question about what the weather will be like tomorrow. This may all be about to change though, with metrology at the heart of Industry 4.0, the fourth industrial revolution.

    According to the National Physical Laboratory (NPL), 95% of companies test final products meet quality standards and it’s not uncommon for companies to have a Quality Assurance (QA) department; this is where you will find most metrologists and metrology engineers. Even in smaller companies without a QA department, measurements are usually taken as a separate end of line process, often by the engineer or technician who is running the machine. On the surface, this makes sense. You take a finished or semi-finished part and check a series of key dimensions against a drawing to ensure conformity. Acceptable parts move onto the next stage while faulty parts are either scrapped or returned for re-work. The same NPL study found that these product verification activities typically account for 10 to 20% of finished product costs.

    Most engineers and technicians are familiar with basic metrology instruments such as callipers and micrometers and these may very well be found on the shop floor next to a machine. They provide a quick and easy means for an operator to check single dimensions on a part periodically (just make sure they’re in calibration!). Basic equipment such as this is useful, but can only provide information about simple features like point-to-point distances. Move away from the shop floor and into a well equipped QA dept and you will find more sophisticated equipment such as CMMs or stylus profilometers, which can be used to examine more in-depth features such as overall form or surface finish. These systems all rely on physical contact with the workpiece to carry out measurements and are well established, for example, the first CMMs were developed in the 1960s and 70s.

    More recent developments have focussed on non-contact systems that typically use optical techniques and the properties of light to carry out measurements. Non-contact optical techniques such as interferometry, photogrammetry and laser measurement offer several advantages over traditional contact measurement. The first, and most obvious, the advantage is measurement speed. As the equipment doesn’t need to physically contact the part, it is often possible to measure large areas rapidly, either by scanning a single point of light over a part or by using vision systems to observe a whole surface in a single measurement.

    Another benefit is the potential for greater accuracy and resolution in certain applications. Devices such as interferometers are capable of sub-nanometer resolution, though there is usually a trade-off between speed and accuracy. Non-contact systems are also exceptionally well suited to the metrology challenges presented by new technologies such as additive manufacturing. For example, the ability of X-ray Computed Tomography (XCT) systems to image and measure internal features, without the need to section a part, is particularly valuable.

    So far, the systems discussed are the type that would generally be confined to the controlled environment of the QA area. However, there is currently a strong push towards using the benefits of non-contact systems to develop embedded metrology solutions and make product verification an integral aspect of the manufacturing process. Incorporating novel sensor networks (sensornets) into the manufacturing environment has the potential to eliminate QA as a separate process. This would allow companies to move away from limited measurement of a small number of key features or measurement of only a sample of a production run and would instead allow continuous monitoring of the part in the process.

    This can significantly reduce or even eliminate time-consuming and costly rework. Similarly, continuous monitoring of the whole production environment allows process variables to be monitored and compensated for improving consistency and reducing downtime. The ability of this vast array of measurement data to control and optimise processes in real-time, presents an unprecedented opportunity to improve productivity both in terms of throughput and product quality.

    Of course, as is often the case, so many potential benefits also bring a host of significant challenges which must be overcome before a system can be viable in a real-world manufacturing environment.  Firstly, it is rarely as simple as taking an existing instrument and bolting it to a machining centre. Instruments must be miniaturised to ensure they can be packaged and positioned in such a way that does not impede the manufacturing process. This miniaturisation is not just an engineering challenge. Often the fundamental physics behind light-based measurement means that new technologies, based on cutting edge science, must be developed.

    A second physical challenge is that the manufacturing environment is usually about as far removed from the sterile confines of the QA lab as it is possible to get. Temperature variations, stability, vibration and physical contaminants such as coolant and swarf can all hinder accurate measurement. Developing robust, rugged sensors that can operate in these environments is essential while adding additional sensors that can quantify the external variables are necessary to compensate for their effects.

    With such substantial physical challenges to overcome, it is easy to overlook other, less tangible problems. Building networks of sensors which are continuously measuring a wide variety of workpiece characteristics and production variables generates vast amounts of data that must be curated, evaluated and utilised effectively if it is to be of any value. This is beyond the capabilities of even the most exceptional human operator or existing automation systems. To solve these challenges, we must draw upon fields such as mathematics and data science to develop a means of handling data and making it machine-readable.

    Sophisticated Artificial Intelligence (AI) systems are then necessary to use these data sets to make real-time decisions. Developing these AI systems is not only a theoretical but also a practical challenge. Real-world data sets from a wide variety of manufacturing environments are essential for teaching these AI systems how to interpret and use the data to make decisions. However, accessing these data sets is a difficult task and researchers are dependent on the relationships they can build with industry.

    It is not uncommon for the current generation of advanced CNC machining centres to incorporate sensors and process monitoring that can be used to achieve some degree of process control and optimisation. These systems are relatively basic and still rely heavily on the skill and experience of the operator to achieve good results. As true embedded metrology solutions become more commonplace and sensornets are incorporated across the manufacturing environment, fully autonomous control will become increasingly possible. Accompanying the development of these autonomous systems is the capability to create true digital twin models, and the ability to accurately simulate the manufacturing process is fundamental to the concept of agile manufacture; optimally choosing the right machine and process within a set of given performance constraints.

    Increased capability in terms of product quality and production throughput is not the only benefit that can be achieved through better metrology practice. Reduced scrappage, more efficient processes and less waste material are all key components of sustainable manufacture.

    All manufacturing companies stand to benefit from increased metrology capability. From small companies manufacturing low volume/high-cost parts, right through to mass manufacture of low-cost consumer goods, the benefits are obvious. Any company in the process of considering upgrading their manufacturing capability would be overlooking a critical factor if they did not also consider the potential productivity benefits which can be made through the use of better measurement tools and techniques. Taking a proactive approach to incorporating metrology into all aspects of production, through design, manufacture and verification, is likely to yield a significant long term competitive advantage. The UK Metrology Research Roadmap, recently published by the Future Metrology Hub, highlights many ways in which industry and academia can work together to develop the next generation of measurement technology.

    The Future Metrology Hub – Author

    The EPSRC Future Metrology Hub, led by Professor Dame Jane Jiang, is a consortium of universities with a world leading track record in metrology research. The project brings together experts from a wide variety of fields such as instrumentation, data science, metrology applications and manufacturing technology.

    https://cdn.mtdcnc.global/cnc/wp-content/uploads/2021/01/30224736/Image-2-640x360.jpg

    Made to measure

    “You can only make as well as you can measure” or a variation thereof is a phrase familiar to most people who work in industrial metrology. Attributed to the pioneering 19th-century engineer and inventor Joseph Whitworth at the height of the first industrial revolution, the phrase is as true now as it was in 1840 when his ‘End Measurements’ invention demonstrated precision to ‘a millionth of an inch’ or 25 nanometres in today’s money! Metrology is a fundamental consideration that is often overlooked in manufacturing,  as Dr Christian Young explains.


    B
    emusingly metrology may often be overlooked, yet it is an aspect of modern manufacturing that is essential for creating the high precision parts that go into everything from iPhones to jet engines. Despite this, tell someone you are a metrologist and the most likely response you will get is a question about what the weather will be like tomorrow. This may all be about to change though, with metrology at the heart of Industry 4.0, the fourth industrial revolution.

    According to the National Physical Laboratory (NPL), 95% of companies test final products meet quality standards and it’s not uncommon for companies to have a Quality Assurance (QA) department; this is where you will find most metrologists and metrology engineers. Even in smaller companies without a QA department, measurements are usually taken as a separate end of line process, often by the engineer or technician who is running the machine. On the surface, this makes sense. You take a finished or semi-finished part and check a series of key dimensions against a drawing to ensure conformity. Acceptable parts move onto the next stage while faulty parts are either scrapped or returned for re-work. The same NPL study found that these product verification activities typically account for 10 to 20% of finished product costs.

    Most engineers and technicians are familiar with basic metrology instruments such as callipers and micrometers and these may very well be found on the shop floor next to a machine. They provide a quick and easy means for an operator to check single dimensions on a part periodically (just make sure they’re in calibration!). Basic equipment such as this is useful, but can only provide information about simple features like point-to-point distances. Move away from the shop floor and into a well equipped QA dept and you will find more sophisticated equipment such as CMMs or stylus profilometers, which can be used to examine more in-depth features such as overall form or surface finish. These systems all rely on physical contact with the workpiece to carry out measurements and are well established, for example, the first CMMs were developed in the 1960s and 70s.

    More recent developments have focussed on non-contact systems that typically use optical techniques and the properties of light to carry out measurements. Non-contact optical techniques such as interferometry, photogrammetry and laser measurement offer several advantages over traditional contact measurement. The first, and most obvious, the advantage is measurement speed. As the equipment doesn’t need to physically contact the part, it is often possible to measure large areas rapidly, either by scanning a single point of light over a part or by using vision systems to observe a whole surface in a single measurement.

    Another benefit is the potential for greater accuracy and resolution in certain applications. Devices such as interferometers are capable of sub-nanometer resolution, though there is usually a trade-off between speed and accuracy. Non-contact systems are also exceptionally well suited to the metrology challenges presented by new technologies such as additive manufacturing. For example, the ability of X-ray Computed Tomography (XCT) systems to image and measure internal features, without the need to section a part, is particularly valuable.

    So far, the systems discussed are the type that would generally be confined to the controlled environment of the QA area. However, there is currently a strong push towards using the benefits of non-contact systems to develop embedded metrology solutions and make product verification an integral aspect of the manufacturing process. Incorporating novel sensor networks (sensornets) into the manufacturing environment has the potential to eliminate QA as a separate process. This would allow companies to move away from limited measurement of a small number of key features or measurement of only a sample of a production run and would instead allow continuous monitoring of the part in the process.

    This can significantly reduce or even eliminate time-consuming and costly rework. Similarly, continuous monitoring of the whole production environment allows process variables to be monitored and compensated for improving consistency and reducing downtime. The ability of this vast array of measurement data to control and optimise processes in real-time, presents an unprecedented opportunity to improve productivity both in terms of throughput and product quality.

    Of course, as is often the case, so many potential benefits also bring a host of significant challenges which must be overcome before a system can be viable in a real-world manufacturing environment.  Firstly, it is rarely as simple as taking an existing instrument and bolting it to a machining centre. Instruments must be miniaturised to ensure they can be packaged and positioned in such a way that does not impede the manufacturing process. This miniaturisation is not just an engineering challenge. Often the fundamental physics behind light-based measurement means that new technologies, based on cutting edge science, must be developed.

    A second physical challenge is that the manufacturing environment is usually about as far removed from the sterile confines of the QA lab as it is possible to get. Temperature variations, stability, vibration and physical contaminants such as coolant and swarf can all hinder accurate measurement. Developing robust, rugged sensors that can operate in these environments is essential while adding additional sensors that can quantify the external variables are necessary to compensate for their effects.

    With such substantial physical challenges to overcome, it is easy to overlook other, less tangible problems. Building networks of sensors which are continuously measuring a wide variety of workpiece characteristics and production variables generates vast amounts of data that must be curated, evaluated and utilised effectively if it is to be of any value. This is beyond the capabilities of even the most exceptional human operator or existing automation systems. To solve these challenges, we must draw upon fields such as mathematics and data science to develop a means of handling data and making it machine-readable.

    Sophisticated Artificial Intelligence (AI) systems are then necessary to use these data sets to make real-time decisions. Developing these AI systems is not only a theoretical but also a practical challenge. Real-world data sets from a wide variety of manufacturing environments are essential for teaching these AI systems how to interpret and use the data to make decisions. However, accessing these data sets is a difficult task and researchers are dependent on the relationships they can build with industry.

    It is not uncommon for the current generation of advanced CNC machining centres to incorporate sensors and process monitoring that can be used to achieve some degree of process control and optimisation. These systems are relatively basic and still rely heavily on the skill and experience of the operator to achieve good results. As true embedded metrology solutions become more commonplace and sensornets are incorporated across the manufacturing environment, fully autonomous control will become increasingly possible. Accompanying the development of these autonomous systems is the capability to create true digital twin models, and the ability to accurately simulate the manufacturing process is fundamental to the concept of agile manufacture; optimally choosing the right machine and process within a set of given performance constraints.

    Increased capability in terms of product quality and production throughput is not the only benefit that can be achieved through better metrology practice. Reduced scrappage, more efficient processes and less waste material are all key components of sustainable manufacture.

    All manufacturing companies stand to benefit from increased metrology capability. From small companies manufacturing low volume/high-cost parts, right through to mass manufacture of low-cost consumer goods, the benefits are obvious. Any company in the process of considering upgrading their manufacturing capability would be overlooking a critical factor if they did not also consider the potential productivity benefits which can be made through the use of better measurement tools and techniques. Taking a proactive approach to incorporating metrology into all aspects of production, through design, manufacture and verification, is likely to yield a significant long term competitive advantage. The UK Metrology Research Roadmap, recently published by the Future Metrology Hub, highlights many ways in which industry and academia can work together to develop the next generation of measurement technology.

    The Future Metrology Hub – Author

    The EPSRC Future Metrology Hub, led by Professor Dame Jane Jiang, is a consortium of universities with a world leading track record in metrology research. The project brings together experts from a wide variety of fields such as instrumentation, data science, metrology applications and manufacturing technology.