Name: John Hart School: Massachusetts Institute of Technology (MIT), Cambridge, Massachusetts Impact: Virtual Testing Saves Time in Product Development
When most of us think of microscopes, we probably think back to junior high school biology class and the challenge of peering at a plate glass slide through the eyepiece of an old-fashioned cantilevered microscope. While those microscopes might have been adequate for viewing smaller-than-the-eye-can-see organisms from the backyard, today's scientists studying human genes and proteins require microscopes capable of nanometer-scale precision. This calls for a whole new way of designing and building microscopes.
To meet the stringent precision requirements while providing powerful magnification, the Precision Engineering Research Group at the Massachusetts Institute, including John Hart, a 23-year-old mechanical engineering Ph.D. student, collaborated with the High Precision Microscope (HPM) project at the University of Illinois Laboratory for Fluorescence Dynamics to design and test the structure for a new type of optical microscope for nanometer-scale biological experiments. Structural designs of current microscopes, which have retained similar cantilevered shapes for decades, make advanced setups such as those needed by the HPM cumbersome, difficult to reconfigure, and unacceptably sensitive to thermal and mechanical disturbances. Significant improvements in the flexibility, stability and resolution of the microscope are needed.
The team at MIT and Illinois designed a new type of microscope structure, featuring five tube sections connected by high-precision kinematic couplings, which made the microscope significantly more thermally stable than the old design. Single-molecule measurements are particularly sensitive to even the slightest mechanical fluctuations. To the stability of a molecular experiment, a one-degree temperature change can be like a human experiencing the temperature change radically from summer to winter.
One of the microscope's unique features is that the sample stage and microscope lens are mounted inside the tube's rings. The symmetrical ring design makes the structure significantly less sensitive to thermal errors and the kinematic couplings allow the microscope to be disassembled and reassembled without recalibration (see Figures 2 and 3). The microscope's structure lets researchers focus on the same tiny spot for a much longer duration with the structure drifting very little, which means that scientists may not have to put the structure in an expensive climate-controlled chamber to prevent temperature changes in the lab from causing the structure to drift.
Figure 2: Pro/ENGINEER model of the new microscope
Figure 3: Photo of the prototype microscope structure
As with any new design, performance testing was a crucial part of the building process. The team had to be sure the proposed microscope structure, created in a virtual environment would perform as expected when it was built. Building multiple microscope structures to test the feasibility of modifications to the microscope's original design would be exhaustingly time consuming and impractical.
"Once we agreed on the design of the microscope, we turned to PTC's Pro/ENGINEER and Pro/MECHANICA to model and simulate the thermal performance of its structure," said Hart. "I first learned about this software when I came to MIT in September 2000. Although other design and testing packages are available at MIT, we chose PTC's products because they offered the most complete integration of design and testing tools."
The team used Pro/ENGINEER, PTC's flagship design software, to build a virtual model of the microscope structure. A similar physical prototype was built from drawings created in Pro/ENGINEER and subjected to a variety of thermal stability experiments. Through computer modeling in Pro/MECHANICA, PTC's testing and analysis software, the virtual model faced the same tests. At the end of the experiment, Hart found that the computer generated model performed within 20 percent of the actual measured estimate, an excellent result.
"Based on those experiments," said Hart, "We knew that Pro/MECHANICA's virtual testing mapped back exactly to real world conditions, which, in the long run, saved us a lot of time. It meant that as we were optimizing the performance of the microscope structure, we could run virtual computer generated tolerance tests, confident that Pro/MECHANICA's simulations were a valid representation of how the model would perform in the real world. We didn't have to build additional physical prototypes, saving us thousands of dollars and a few weeks of labor."
Another major challenge of the project was the mutual dependency of the mechanical and optical design. Pro/ENGINEER was extremely valuable for integrating the mechanical and optical models. The optics and optical raytrace were modeled with in a separate optical design package and then exported to an IGES file. Using Pro/ENGINEER's IGES import capabilities, the optical parts and raytrace were seamlessly assembled with the mechanical parts created in Pro/ENGINEER. The creation of a unified mechanical and optical model, which could be visualized together, greatly facilitated a consistent design. This reduced the possibility for errors while eliminating the need for building prototypes.
Now finishing the final prototype of the microscope, scientists are looking forward to examining a world too small to have envisioned without the aid of the virtual testing capabilities of PTC's software.
