Modelling and control of neutron and synchrotron beamline positioning systems S.O. Nneji a,b,n , S.Y. Zhang b , S. Kabra b , R.J. Moat a , J.A. James a a The Open University, Materials Engineering, Walton Hall, Milton Keynes, Buckinghamshire MK7 6AA, UK b Science and Technology Facility Council , Rutherford Appleton Laboratory, Harwell Science and Innovation Campus, Didcot, OX110QX Oxfordshire, UK article info Article history: Received 16 November 2015 Received in revised form 18 December 2015 Accepted 21 December 2015 Available online 14 January 2016 Keywords: Neutron diffractometer Sample alignment Kinematic calibration Simulation Automation Robotics abstract Measurement of residual stress using neutron or synchrotron diffraction relies on the accurate alignment of the sample in relation to the gauge volume of the instrument. Automatic sample alignment can be achieved using kinematic models of the positioning system provided the relevant kinematic parameters are known, or can be determined, to a suitable accuracy. In this paper, the use of techniques from robotic calibration theory to generate kinematic models of both off-the-shelf and custom-built positioning systems is demonstrated. The approach is illustrated using a positioning system in use on the ENGIN-X instrument at the UK's ISIS pulsed neutron source comprising a traditional XYZΩ table augmented with a triple axis manipulator. Accuracies better than 100 microns were achieved for this compound system. Discussed here in terms of sample positioning systems these methods are entirely applicable to other moving instrument components such as beam shaping jaws and detectors. & 2016 Elsevier B.V. All rights reserved. 1. Introduction Neutron and synchrotron diffraction are non-destructive methods of determining residual stress from measurements of strain within crystalline or polycrystalline materials. Residual stresses are those stresses present in an object in the absence of any external load or force. Such stresses can be very detrimental to the performance of a material, or the life of a component. Alter- natively, beneficial residual stresses may be introduced deliber- ately [1]. Neutron and synchrotron beam line experiments, as used for residual stress measurements, rely on accurate alignment of the sample in relation to the beam and hence the instrument hardware. Such instruments often incorporate XYZΩ tables for positioning the sample in the beam, with additional hardware such as rotation tables or goniometers being added to meet par- ticular requirements. Interest has also recently increased in exploring the use of industrial robotic arms as sample positioning systems, [2,3]. The motivation for the interest in robotic arms is the potential improvement in flexibility and the possibility of automation. For example, when controlled by suitable software, such systems may enable all required strain components to be obtained without manual intervention, even within large and geometrically complex samples. Whatever positioning system is in use the motor positions (or, in robotics terminology, ‘joint variables’) required to bring the measurement point to the measurement position, with sufficient accuracy, will need to be determined. Traditionally this has been achieved by performing a series of ‘wall scans’ in which the sample surface is passed through the beam while the scattered intensity is recorded. The position of the sample in relation to the beam is then determined from the intensity profile. By repeating this process, using scans through different points on the sample sur- face, the sample alignment may be determined. This process however can take considerable time, particularly with a sample of complex geometry, thereby considerably reducing the time avail- able for making scientifically useful measurements. For this reason other sample alignment methods have been developed. Ratel et al. [4] proposed doing sample alignment with a modular sample holder, coordinate measurement machine (CMM) and a XYZΩ positioning table. Their sample was mounted on the positioner and digitised with the CMM. Measurement points were specified using the model acquired from the CMM and the coordinate frame of the CMM was set as the inverse of the axes of the positioning table, enabling motor positions to be generated by simply invert- ing the coordinates of the measurement point. The solution pro- posed by Ratel requires samples to be mounted on sample holders which are then used to align the coordinate systems on the instrument and the CMM, the method also assumes that a Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/nima Nuclear Instruments and Methods in Physics Research A http://dx.doi.org/10.1016/j.nima.2015.12.067 0168-9002/& 2016 Elsevier B.V. All rights reserved. n Corresponding author at: The Open University, Materials Engineering, Walton Hall, Milton Keynes, Buckinghamshire MK7 6AA, UK. E-mail address: Stephen.nneji@open.ac.uk (S.O. Nneji). Nuclear Instruments and Methods in Physics Research A 813 (2016) 123–131