Shrinking thinking

Written by: Tom Shelley | Published:

Moves are afoot to make very small, complex products – many for medical markets – cheap enough to be disposable. Tom Shelley reports

It is one thing to make very small components, but quite another to assemble them into micro products cheap enough to be disposable, and reliable enough to be ingested or implanted in human beings.
The challenges in developing such products lie both in the need to understand the difference between ordinary size devices and those that are very small, as well as developing machines and technologies capable to making and verifying them in large numbers at low cost.
With large markets in view, research and development is well underway, both in academia, industry, and in joint research projects embracing both. This is especially so in the medical device field, which – with more and more people globally wanting to live longer and better – is likely to continue to expand, regardless of whatever happens to the rest of the world economy.
One such organisation is EUPASS - Evolvable Ultra Precision Assembly Systems. Academic partners include the Tampere University of Technology in Finland, some of whose work Eureka described in its June 2007 edition, and the University of Nottingham. Speaking at the recent TCT/MM Live event in Coventry, Professor Svetan Ratchev, director of the precision manufacturing centre at Nottingham, explained that the purpose of the programme was to develop common standards, so that different pieces of machinery and software could be joined together to, as he phrased it, “plug and produce”. The reason for this is that, while micro-machines can be assembled by technicians with eye lupes and tweezers – using the sorts of techniques developed to make top-end mechanical Swiss watches – this would mean such products could only be sold at equally top-end prices. The only alternative is to make and assemble parts using robotics, programmed using standard software that is quick to deploy.
This affects both the design of the products – cheap watches are based on entirely different technology to expensive watches, which makes them easy to mass manufacture – and the design of the machines and their accompanying software.
The EUPASS programme is coming towards its end, but Ratchev said it had been technologically very successful, with an achieved target positioning accuracy of 1.5 microns, against a target positioning accuracy of 5 microns. In his own laboratory, he reports, he can achieve a positioning accuracy of 10nm.
The final Eupass demonstration machine, PV2, was coming together at the Fachhochscule Nordwestschweiz in Windisch, Germany, at time of writing, using a slot-based architecture called ‘Bay Concept’, which allows all modules to slot into a generic docking frame in a workstation. Power, control signals and compressed air are supplied through a standard Eupass Bay interface. Two control approaches are being tested, one based on IEC61499 and the other on agent technology. Test assembly operations will include the ultra precise adhesive joining of a small valve from Festo, and the assembly of a door handle and tubes onto a valve from Electrolux. Feintool has built all the workstations for PV2, while Beckhoff is supplying all of the software and control hardware.
But these are nothing to the manufacturing problems that will have to be overcome to produce the next generation of products already contemplated, such as micro machines that attack only cancerous or otherwise diseased cells, and stem cell manufacturing plants to construct replacement human tissue.
When machining very small parts, or very small features on large parts, thermal and mechanical effects start to become very important. Arthur Turner, managing director of Deckel Grinders in St Helens, referring to micro milling at TCT, said close attention has to be paid to anything that might cause changes in temperature, such as draughts.
Eureka heard much the same thing when at the opening of the new Renishaw Centre for Manufacturing Productivity at Greenwich University in Chatham on 22 October. Renishaw already offers short courses based on the ‘Applying the Productive Process Pyramid’. Now, according to Grahame Baker, senior lecturer in engineering management and manufacturing systems at the university, there are plans to use this as the basis for a new final year engineering course. He also hopes this will lead on to research into auto compensation for thermal effects in machining.
Introducing a forum at TCT/MM Live, David Aspinwall, reader in advanced manufacturing technology in the machining research group at Birmingham University, said that, for micro parts, it is essential to keep them on the same machine, with the same working fluid. It is also necessary to remember that, when machining small features, different parts of a cutter are at different temperatures.
For very small features, the general thinking is to move away from mechanical machining and turn to electro discharge machining and lasers, so there are no mechanical forces involved that can cause deformation and hence errors. Oxford Lasers believes this to be the case, but uses Delcam’s PowerMill CAM software to programme their machines, even though the package is really intended to be used with traditional milling machines.
Chris Griffiths, from the Manufacturing Engineering Centre at Cardiff University, talked about some of his work on micro injection moulding. “De-moulding is a big problem for small parts,” he reports, because surface area to volume ratios become greater as size reduces, with increased risk of parts being broken when they are ejected. He described the results of both finite element modelling using Moldflow and practical experiments, and reported that coating mould tools with diamond like carbon solved a lot of challenges.
But one big problem that is far from solved – and one that is crucial, if micro and nano scale products are to be used in implanted medical applications – is verifying that parts have truly been manufactured according to the design intent.
True 3D metrology becomes difficult or impossible when verifying sub micron and nm sized features on surfaces, according to Professor Richard Leach, from NPL. “There are no ISO standards for measuring surface texture”, he pointed out. He then explained how the scanning white light interferometers used by many enterprises for verifying such surfaces can often produce results that bear little relation to what is really there, because of various optical effects caused by the structures.
The solution, he believes, is to produce standard surface structures that can be used to calibrate optical instruments and verify the standards to which sub standards can be traced by a special stylus instrument that NPL has spent six years developing. This machine has a working volume of 8mm x 8mm x 1mm, but it can measure surface features with an uncertainty of only 10nm x 10nm x 1nm. The work piece is supported on a co-planar X-Y air bearing within a 3-axis mirror block. Position is established by 4 degree of freedom interferometry. The probe enters through a hole in the block and has a mirror on the rear of the stylus. Since any significant force would damage either the surface or the stylus, or both, it is not possible to use a static mass or spring to load the stylus. Instead, another air bearing laterally supports the probe, while having an attached toroidal magnet float between two vertically separated coils controls its vertical motion. This ensures that the same, tiny, downward net force is applied, regardless of vertical displacement.
It seems that, in our ever-shrinking world, the big opportunities are still expanding – while the commercial possibilities are equally there, waiting to be exploited.


* A modular, multi-vendor system standard has been developed for plug-and-produce micro manufacturing machines

* Manufacturing small components and small features on large components requires paying attention to matters that do not affect the production of large components and features, and may require modifying their design

* Validation of exact dimensions of very small features on surfaces can be problematic, but a new machine has been developed that can reliably and accurately verify nm scale features on standards, to which it will be possible to trace test pieces to calibrate optical metrology systems

This material is protected by MA Business copyright see Terms and Conditions.
One-off usage is permitted but bulk copying is not. For multiple copies contact the sales team.


Please view our Terms and Conditions before leaving a comment.

Change the CAPTCHA codeSpeak the CAPTCHA code