Abstract
The Italian Neutron Experimental Station, or INES for short, instrument beamline at the ISIS Neutron and Muon Spallation Source recently underwent an upgrade to its sample positioning area to improve the sample handling capability and capacity of the instrument. INES is a powder diffractometer instrument devoted to materials characterization and cultural heritage studies. Due to the types of samples received for cultural heritage studies the instrument scientists wanted to maximize the height of samples that could be scanned with the instrument. This paper covers the in-house design and delivery of a new, two axis, linear positioning stage for INES that due to the addition of a new linear height stage, needed to be as low profile as possible to give the instrument scientists the sample capacity they required.
Keywords
Introduction
The ISIS Neutron and Muon Spallation Source [4] is a research facility located at the Science and Technology Facility Council (STFC) Rutherford Appleton Laboratory in Oxfordshire, UK. Used for a wide array of academic research from cultural heritage, through to medicinal and materials studies the facility produces beams of neutrons and muons covering seven main science techniques including; Diffraction, Spectroscopy, Reflectometry, and Imaging.
Opened in 1985, and with the completion of a second instrument hall (Target Station 2) in 2008, there are now more than 30 operational instrument beamlines at the facility.
One of these instruments, the Italian Neutron Experimental Station [3], or INES for short, is a powder diffractometer run as a collaboration between ISIS and the Italian National Research Council (CNR) [1]. INES is devoted to materials characterisation, cultural heritage studies, and equipment tests; and has recently been used to study ancient Damascus swords, meteorites, and the helm of the Black Prince. Located in Target Station 1, this instrument has been running since 2006 and was due for a blockhouse sample area upgrade which concluded in 2019. The purpose of this upgrade was to increase the scanning range of the instrument and improve its capability and capacity for sample handling. This was to be achieved through replacement of the existing two axis linear translation stages for sample positioning, a new linear height stage to allow vertical scanning, and a new rotation stage. In addition to improving the size of samples that can be placed in the instrument, this upgrade also aimed to improve the experimental throughput time of samples as the new stages would allow sample repositioning without having to enter the instrument between scans which can be a slow process due to having to wait for the instruments neutron shutter to close before entry to the blockhouse is allowed, and then open after exiting to continue taking measurements. This paper covers the design and delivery of the replacement two axis sample stage. Due to the introduction of the new height and rotation stages there were significant height restrictions placed on the design of the new two axis stage so this had to be minimised as much as feasibly possible so that larger samples, specifically cultural artefacts, could be scanned.
Design and development
Replacement of existing stages
The old two axis sample stage in the INES blockhouse, shown in Fig. 1, provided 100 mm of travel in the X and Y directions and took up a physical space of 450 mm by 450 mm area and a height of 190 mm. These old stages were slow, bulky, and no longer able to do what the scientists needed them to do.
Often at ISIS, new motion projects are either externally tendered for design and manufacture or utilize off-the-shelf bought in components. For this project, due to the low profile requirement, a short search revealed no suitable purchasable hardware to meet requirements; whilst it would have been possible to go out to tender to meet these requirements, the decision was taken early on in the project that the new two-axis stage would be entirely designed and manufactured in-house. A potential issue with going to tender was that we had no fixed number-based requirement on the height of the stages, just that they should be as low as possible and we were not aware of what might be achievable, this could have resulted in a situation where a company designing these stages for us may not have been prepared to spend as much time and money minimizing the height as we were prepared to. Producing a system ourselves would also enable us to develop in-house experience, control the design, manufacturing, and assembly procedures and ensure the completed product completely met the scientific requirements without compromise. This in-house experience is very valuable moving forwards as it provides a better understanding for future projects, even tendered ones, in knowing what specifications can, or are easily, achievable and what aspects of stage design are most challenging.
Old INES sample stages.
The specification for the new system was generated through discussion with the instrument scientists to gain a grasp of their experimental needs and how they physically performed their experiments. As the INES instrument is often used for studying items of historical significance these can often not be mounted on to the sample plate too securely, this means there is a requirement for a low jerk system with high resolution and repeatability; accuracy is somewhat irrelevant in this experimental setup as scan points are identified using an external laser source and these points then need to be repeatedly moved between. The following are some of the requirements:
±50 μm repeatability.
<10 μm resolution.
Vacuum compatible (rough vacuum
50 kg load capacity.
±100 mm linear travel range in X and Y directions.
Fit within 780 mm internal diameter vacuum tank (avoiding vacuum feedthrough ports).
Combined stage height (including new rotational stage) of less than 190 mm.
The height was one of the significant challenges throughout this design project and the scientists were interested in reducing this value as much as possible to increase their sample height options. The example given was that of a sword hilt which they would need to mount vertically, the lower down these stages physically were, the more of the sword they would be able to scan.
Stage design
Off-the-shelf low-profile rails, just 12 mm tall, were selected for minimizing the height of the design. Generally, low profile rails will be rated for much smaller moment loads than direct load, to accommodate for this, the stage was designed with a wide base to minimize any potential load conditions that would contribute to significant overhang. The design also provided relief cutouts for motion hardware, such as the limit switches and ballscrews, to bring the overall height down further.
Other notable design considerations are the integration of the wiring harness into the stage and the vacuum compatibility. The wiring harness integration is achieved through p-clip mounting points throughout the cavities of the device; for any moving parts of the harness it was initially desired to use a small energy chain but none were found to be compatible, instead a thin gauge closed-coil spring was used to secure and protect these parts of the harness. The spring is fastened to the stages at p-clips and the individual wires for the electrical harness routed through. More details of this can be found in Section 3.3.
For the vacuum compatibility, through holes are used wherever possible and, where not possible, venting routes between any trapped volumes have been added.
With the motion hardware selection for this system, this was an iterative cycle of finding hardware small enough and then validating it could meet the performance requirements. For example, with the selection of the motor it was necessary to find something small enough that would fit in the height available but still have the torque and inertia match required. Given the small loads associated with this system and the use of linear rails, the torque required remained very small (7 mN·m with a 4× safety factor applied), and the inertia of the system was calculated at approximately 0.082 kg·cm2 based on the masses predicted by the CAD model. Using some of the common motor suppliers of ISIS motion hardware, a motor was selected by assessing the torque, size, and inertia characteristics of available options. Given the limited space available in this system for a motor, the selection was first filtered by frame size looking to allow for the biggest motor that could fit in the space available. This is not the usual approach for motor selection but given the space available in this system was so small and motor options around this size are not expensive, it did not make sense to downsize the motor any further than necessary as this would make assembly and maintenance more difficult. Following this the torque and inertia were assessed to find a suitable option. In this case it was also possible to find a motor option with integrated back-axle encoder which saved time from having to select a separate encoder and then finding or designing mounting hardware to attach this encoder to the motor. Generally, motors with integrated encoders also provide a smaller footprint than trying to mount a separate encoder to a motor. Through this process a 28 mm frame size motor was selected which provided 81 mN·m of torque (direct drive) with a rotor inertia of 0.009 kg·cm2. Whilst the motor torque was significantly more than required for the system, the inertia mismatch worked out to about 9 to 1. Inertia mismatch can cause issues in motion systems as it can create an excessive resistance to velocity changes; for stepper driven systems, a serious inertia mismatch can cause skipped steps, excessive vibration and noise, and completely prevent a system from working or accelerating up to desired velocity. For very high-performance or high speed systems it is recommended to try and reduce this inertia match to as close to
Following all this initial design work and hardware selection, the design was continuously optimized from its initial concept shown in Fig. 2, through discussion with machinists and operational staff to alter access and dimensions to better suit our in-house manufacturing capabilities as well as try to make assembly simpler which was a major benefit of the in-house design/manufacturing approach selected for this project.
The final stage design measured in at a 78 mm height for X and Y translation and was designed to work with an off-the-shelf Z rotation stage sitting on top for an additional 60 mm of height. The total footprint for the translation stages was 440 mm by 440 mm for the main base with motors protruding an additional 100 mm. For the CAD model of this system, Figs 3 and 4 show some of the finer details and Fig. 5 provides a complete view of this final design.
Initial concept design for two axis system with Z rotation stage on top.
Open side view of a single stage.
Venting routes for trapped volumes.
Final design view.
Manufacturing
In-house manufacturing presents some significant advantages over the traditional route used at ISIS of sending parts out for quote and manufacture. A direct and consistent line of communication, including face-to-face meetings, between designer and machinist allows any issues or changes to be discussed, corrected, and approved with minimal time delay or potential for confusion. In the case of this project one of the significant changes made was with the removal of the venting routes between trapped volumes in favor of vented screws. It also enabled the parts to be test fit as they were made which in turn allowed for more frequent updates to the rest of the project stakeholders, such as when the parts the manufactured parts came back from anodising shown in Fig. 6. Overall manufacturer of the system went without issues and thanks to careful planning during the 3D design stage of the project, such as with the creation of assembly instructions illustrated in Fig. 7, everything assembled as intended.
Parts ready from anodising for assembly.
Example of assembly instructions provided as part of mechanical drawings.
CAD section view showing striker (yellow) and end-of-travel switch.
Testing of the system was led by one of the operational support staff motion technicians. The hardware was all tested in a workshop prior to final wiring which confirmed that all components worked, and the correct specification had been achieved. This part of the process also identified an issue with the end-of-travel limits of the system in that the switches were triggering too early and not allowing the full range of travel intended by the system. The strikers that hit these end-of-travel switches were modified to try and alleviate this but not enough to gain the full travel range desired. The modification of the strikers to try to gain additional travel involved reducing the height slightly and reducing the angle of contact, in retrospect the strikers should have been made shorter too but by this stage the project was pushed for time and needed to be installed. Figure 8 shows a CAD section view of one of the strikers and limit switches. Table 1 shows the measured performance of the stages; unfortunately at the time of this project being run, the design division at ISIS did not have the hardware available for performing accuracy and repeatability tests however after extensive use during experiments the scientists have confirmed that the stage meets their performance expectations when measured with neutrons.
Commissioning and integration
The final wiring of the system was led by the ISIS Electrical Operations team before final installation and commissioning by the motion technician on the instrument beamline. There had been some concerns throughout the project about how the low-profile design and integrated wiring harness would impact the wiring process, and these did prove to be challenging. Whilst it was possible to wire the stage, it took much more time than anticipated and some more modifications had to be made to accommodate for the wire spring guard on the upper stage as it would sometimes get trapped. A member of the mechanical operations team added a bracket to help guide the wiring spring which fixed this issue. The two figures in this section show some sections of the wiring harness and the cable management used throughout. Figure 9 highlights how tight some sections were to wire. Figure 10 shows more of an overview of the electrical harness; on the upper gold stage it is possible to see the closed-coil spring used to protect and manage the wiring between moving elements of the stages, on the lower red stage it is possible to see how the harness is split in to three tracks (highlighted by red boxes), this separation is to reduce the overall bundle diameter and make wiring easier for each section by reducing the number of wires in each bundle.
Performance of built stages
Performance of built stages
End-of-travel limit switch wiring.
View showing the electrical harness of the two translation stages.
The best available view of the stages can be seen in Fig. 11 prior to their final wiring however as of writing, the new stages have been installed and operating in INES for just over 6 months. Lack of final testing in a workshop after full assembly and wiring did result in the failure to capture one electrical issue, a non-functional encoder. This meant the scientists had to use one of the axes of the stage in open-loop control mode until time available in a maintenance shutdown could be found to repair this issue. Even in this state, the new linear axes have been running without issue and continue to do so since the repair to the single encoder. The instrument scientist feedback to this project has been incredibly positive and they are satisfied with the performance that this new linear stage offers them.
Stage assembly prior to final wiring.
Reflecting on this project, through more complex part design it was possible to significantly reduce the overall height of the linear stages for the INES beamline but this lead to more complex assembly, manufacturing challenges, as well as significant extra design time. Given the initial discussions with the instrument scientists and understanding how important scan height was to them, these extra complications were less significant for this system, but this would certainly not be the case for every application. For this project, the instrument scientists were able to go from a 100 mm travel two-axis stage that occupied 190 mm of height, to around 170 mm of travel in a 68 mm height (around a 64% reduction).
This completely in-house system enabled the development of skills and knowledge that are not obtained by buying an off-the-shelf solution, it also provided an opportunity for many teams across the facility to work together who otherwise would not do so and an unforeseen benefit of this project has been seeing these team working relationships improve. These relationships are something that will, hopefully, carry forward and prove beneficial in other future projects with everyone having a greater understanding of what each team is able to bring to a project to ensure its success and when all these separate teams need to be involved in the project process.
Some potential developments for in-house stage design in the future would look to standardize the selection of hardware in stages and improve the ease of wiring stages either through the use of small in-line connectors in the electrical harness or development of custom PCBs to completely replace the wiring harness within the stages utilizing push-fit, PCB mounted connectors between the stages to allow them to be ‘plugged’ together.
As briefly discussed in Section 3.2, it was not possible at the time to test the accuracy and repeatability of these stages however, over the last year the ISIS Design Division has been building a new Mechatronics Lab with appropriate equipment to allow this sort of testing in the future which should hopefully help push the performance of in-house designs even further.
Footnotes
Acknowledgements
The author would like to thank the instrument scientists A. Scherillo and A. Fedrigo for their support and enthusiasm with regards to this project and providing the original requirements, S. Cox, N. Webb and D. McPhail for their support of the design phase, J. Lewis and S. Griffiths for the manufacture and mechanical support of the system, J. Nutter and J. Chandler for the electrical installation of the system, and finally I. Johnson for testing and commissioning the final system on the instrument. This project would not have been a success without the involvement of all these parties.