Source-linked AI summary
A one-piece 3D printed flexure translation stage for open-source microscopy
James P. Sharkey, Darryl C. W. Foo, Alexandre Kabla, Jeremy J. Baumberg, Richard W. Bowman
TL;DR
Precise positioning is essential in scientific apparatus, but manufacturing difficulty can make integrated mechanisms hard to realise. This paper uses plastic compliance to build a monolithic 3D-printed flexure translation stage and a microscope around it. The stage combines sub-micron position accuracy with greater range than equivalently sized metal flexures and low drift over days.
Problem
Custom integrated positioning mechanisms can offer lower drift, higher stiffness, and better fit than stacked commercial stages, but can be difficult to manufacture.
Source: §I · p. 1Method
The authors design a monolithic 3D-printed flexure stage for plastic and use it as the basis of a microscope to assess mechanical performance.
Source: §I · pp. 1–2Results
The stage offers greater motion range than equivalently sized metal flexures and sub-micron position accuracy, while focus typically drifts less than 10 µm over five days.
Source: A. Drift · p. 4, §VI · p. 6Takeaways & Limitations
The stage can form the basis of a mechanically stable microscope for timelapse experiments without autofocus or active drift correction.
Source: §VI · p. 6Usable travel is limited by the maximum flexure bend angle of about 6°, giving around 20% of the lever length.
Source: §III · p. 3
Abstract
from arXiv · showhide
Open source hardware has the potential to revolutionise the way we build scientific instruments; with the advent of readily-available 3D printers, mechanical designs can now be shared, improved and replicated faster and more easily than ever before. However, printed parts are typically plastic and often perform poorly compared to traditionally machined mechanisms. We have overcome many of the limitations of 3D printed mechanisms by exploiting the compliance of the plastic to produce a monolithic 3D printed flexure translation stage, capable of sub-micron-scale motion over a range of $8\times8\times4\,$mm. This requires minimal post-print clean-up, and can be automated with readily-available stepper motors. The resulting plastic composite structure is very stiff and exhibits remarkably low drift, moving less than $20\,\mu$m over the course of a week, without temperature stabilisation. This enables us to construct a miniature microscope with excellent mechanical stability, perfect for timelapse measurements in situ in an incubator or fume hood. The ease of manufacture lends itself to use in containment facilities where disposability is advantageous, and to experiments requiring many microscopes in parallel. High performance mechanisms based on printed flexures need not be limited to microscopy, and we anticipate their use in other devices both within the laboratory and beyond.
I. INTRODUCTION
Precisely positioning samples is essential, yet custom integrated mechanisms can be difficult to manufacture. The paper presents a 3D-printed flexure stage and microscope as an accessible route to stable, precise positioning and open-source instrument development.
I. INTRODUCTION
- Custom integrated mechanisms can offer lower drift, higher stiffness, and better fit than stacked commercial stages, but manufacturing difficulty often leads designers to choose the latter.The paper identifies precise positioning as a widespread experimental need.Source: §I · p. 1
- Open-source scientific hardware primarily enables researchers to investigate, verify, and improve methods, rather than simply reducing costs.Shared designs also facilitate customisation and improvements that can be shared with others.Source: §I · p. 1
- The monolithic printed flexure stage is designed for plastic rather than machined metal, requires minimal post-print assembly, and offers greater motion range than equivalently sized metal flexures.Its parametric OpenSCAD design can be altered and printed on most currently available printers.Source: §I · p. 1
- The microscope uses the printed stage to assess mechanical stability and positioning precision in a realistic setup, addressing microscopy’s need to hold samples stably in the focal plane and translate them across the field of view.The authors identify timelapse studies and constrained settings such as incubators or fume hoods as potential applications.Source: §I · pp. 1–2
II. 3D PRINTED HINGES
The paper adapts flexure hinges to the compliance and layer-by-layer fabrication of plastic 3D printing. It constrains hinge bending and selects geometry intended to keep deformation reversible while allowing useful motion.
II. 3D PRINTED HINGES
- Flexure hinges, or “living hinges,” bend reversibly when deliberately thinned regions deform; plastic’s greater compliance than metal can allow a longer motion range.The paper frames this as an advantage that requires careful design to account for plastic’s lower stiffness.Source: §II · p. 2
- No support material is needed for the microscope, but thin flexures and bridging spans are challenging to print and may require printer-specific adjustments.A test object is included to help optimise the printer before printing the full microscope.Source: §II · p. 2
- The design limits each flexure hinge to bending no more than α = 6° to avoid deformation beyond the elastic limit and resulting permanent damage.The hinge axes are printed horizontally so the flexures can be thinner and stronger than with vertical axes.Source: §II · p. 2
- Corner-filleted hinges better match layer-by-layer printing than circular cut-outs, and the selected links are 0.72 mm thick, 1.5 mm long, and 4 mm wide.An arc approximation gives an estimated maximum strain of about 0.024, close to the yield strain of commonly used PLA and ABS.Source: §II · p. 2
- The authors report no snapped flexures after six months of use, while less-constrained designs have snapped when bent beyond safe limits.The maximum strain is concentrated at the flexure edge, and the link dimensions can be adjusted in the parametric CAD design for different materials.Source: §II · p. 3
III. MECHANICAL DESIGN
The microscope uses flexure-linked mechanisms to position the sample in x and y and the objective in z. A compact table-like xy stage and a one-piece printable body support the design, with screw actuation suited to manual or motorized control.
III. MECHANICAL DESIGN
- Flexure-linked levers move the sample table in two dimensions, while a four-bar mechanism moves the objective vertically; all three axes use M3 screw-and-nut actuation.Printed thumb-wheels or gears meshing with compact stepper motors can control the screws.Source: §III · p. 3
- The four-bar mechanism moves the objective vertically while preserving its orientation and lateral position, with near-linear travel over the usable range.The hinge bend is limited to about 6°, giving a range of around 20% of lever length and a maximum path deviation of 1% of lever length.Source: §III · p. 3
- The table-like xy stage is stiffer and more compact than stacked stages, and its body-mounted actuators transmit less vibration to the stage.Its legs can tilt for motion in both x and y, while each actuator moves only along one axis.Source: §III · pp. 3–4
- The microscope body prints as one piece, with bridges linking eight legs to avoid requiring components compliant along two axes.The design also uses a 2.6:1 mechanical reduction for z-axis motion to improve precision.Source: §III · p. 4
IV. MECHANICAL PERFORMANCE
Mechanical performance was assessed by tracking the relative position of the sample stage and objective. Particle tracking measured drift and the automated stage’s positioning accuracy and repeatability.
IV. MECHANICAL PERFORMANCE
- Particle tracking of 6 µm latex spheres on a glass slide measured stage-objective separation to quantify drift, positioning accuracy, and repeatability.Most tracking ran in real time on a Raspberry Pi; high-speed frames were recorded at 90 Hz for offline analysis.Source: §IV · p. 4
- For high-speed tracking, camera frames were recorded to RAM at 90 Hz, then written to disk and analysed offline.Source: §IV · p. 4
- Drift along z was measured by folding the imaging path with a prism and mounting the slide vertically on printed supports.The camera’s x axis then represented the z axis.Source: §IV · p. 4
A. Drift
The stage’s axial position typically drifted less than 10 µm over five days, while lateral axes can show greater creep after substantial movement. Allan deviation indicates performance within an order of magnitude of highly optimized optical-tweezers systems, with a noise–stability trade-off minimized around one second.
A. Drift
- Less than 10 µm of axial drift occurred over five days; lateral axes showed similar drift but can creep more after large movements.The y trace appeared to relax exponentially at the start of a measurement after substantial prior movement.Source: A. Drift · pp. 4–5
- The z axis has lower creep because it uses fewer flexure joints and requires less force to move.Axial drift is low enough to maintain focus with the lens used here, though a higher numerical aperture lens may require improvement.Source: A. Drift · p. 5
- Within an order of magnitude of highly optimized optical-tweezers systems, the stage’s Allan deviation represents adequate performance for many experiments.The paper describes this as few-nanometer performance.Source: A. Drift · p. 5
- Around one second, high-speed measurements show the lowest variance between time chunks, balancing measurement noise against instrument stability.At short timescales, Allan deviation decreases as τ^-1/2 from measurement noise; at longer timescales it rises due to drift, with a slope between τ^1/2 and τ.Source: A. Drift · p. 5
B. Accuracy & Repeatability
The automated stage returned to within 1 µm for small moves, with repeatability worsening to about 15 µm after 1 mm moves. Its measured single-microstep travel was 67 ± 6 nm, below the 78 nm estimate, and small steps show motor-driver vibration.
B. Accuracy & Repeatability
- Less than 1 µm repeatability for small moves increased to about 15 µm for 1 mm moves.Repeatability is the root mean squared error after moving a given distance and returning to the start; an anti-backlash algorithm improves performance by approaching the target from the same direction.Source: B. Accuracy & Repeatability · p. 5
- Figure 8 plots RMS positioning error after moves in random directions and returns to the origin, with inset return-position scatterplots for two move distances.The smaller and larger inset scale bars are 1 µm and 4 µm, respectively.Source: Figure 8 · p. 6
- Single microsteps are small, but pulse-width-modulation vibration is visible at small length scales; larger steps are better defined.Figure 9 shows the stage response to alternating X and Y steps.Source: Figure 9 · p. 6
- 67 ± 6 nm per microstep was measured, about 86% of the 78 nm estimate for a 2:1 geared motor.The authors attribute the discrepancy to repeatable flex in the actuating lever while the material remains within its elastic limit.Source: B. Accuracy & Repeatability · p. 6
V. OPTICS
The microscope uses a reversed Raspberry Pi camera-module lens to provide a 300 × 230 µm field of view and around 2 µm optical resolution. Transmission illumination supports bright-field imaging, while a printed condenser holder adds dark-field and basic phase contrast.
V. OPTICS
- The reversed Raspberry Pi camera-module lens provides a 300 × 230 µm field of view and around 2 µm optical resolution.Each pixel corresponds to 120 nm in the sample; the authors compare the resolution and field of view to a good modern 10× or 20× objective with numerical aperture 0.15.Source: §V · p. 6
- Transmission illumination provides bright-field imaging, and a printed condenser holder adds dark-field and basic phase contrast.The inverted microscope images samples generally through slides or coverslips; fluorescence imaging, a better low-cost lens, and higher resolution are identified as possible improvements.Source: §V · p. 6
VI. CONCLUSIONS
The authors present a 3D-printed translation stage with sub-micron position accuracy, greater motion range than metal flexure stages, and low drift in ambient conditions. It supports a stable microscope for timelapse experiments, and its design files are freely available for adaptation.
VI. CONCLUSIONS
- The stage combines a motion range greater than metal flexure stages with sub-micron position accuracy and remarkably low drift in ambient conditions.The low drift holds over days or weeks.Source: §VI · p. 6
- The stage supports a mechanically stable microscope for timelapse experiments without autofocus or active drift correction.The microscope also serves applications where space and weight are at a premium.Source: §VI · p. 6
- Removing the dark-field stop switches the microscope to bright field; removing the condenser lens reduces brightness but does not stop operation.Source: Figure 10 · p. 7
- Freely available design files and assembly instructions can be printed on most currently available machines, and the authors plan to explore other flexure mechanisms.They hope the design will contribute to the open-source hardware library.Source: §VI · p. 7