Week 11:Technical Paper
Table of Contents
Click here to download Word file.
AbstractThe Corning Tropel Senior Design team was tasked with the development of the next generation, dual head Lightgage Metrology System. The system will be the only product available that is able to simultaneously measure both sides of a part and report a thickness. The design fulfills the stringent requirements for accurate measurement. Motion was successfully incorporated mechanically and with the TMS software. The prototype system was successfully manufactured, assembled and tested.
Technical Background"Corning's LightGage system is an advanced, full surface interferometer for measuring flatness, parallelism, and depth." Using a tunable laser, an infrared camera and mathematical computer algorithms the LightGage can accurately characterize the surface features of a part. The system then processes the data and graphically displays details of the part's surface and all relevant measurements to the user. While this accurately describes how the LightGage operates, it is merely a top-level, black-box description. In order to fully understand the scope and mission of this project it is necessary to review the technology behind the LightGage interferometer.
As stated above, the LightGage uses a camera to collect multiple interferometric images of the part under measurement. For each image, the laser frequency (alpha wavelength) is changed by an equal, predefined increment (frequency stepping). Because the part's surface is not perfectly flat, the intensity of each pixel will vary by a frequency that corresponds to the mismatch in the distance between the reference and test arms of the interferometer (see Figure 1).
For this type of interferometer (frequency-scanning), the modulation frequency is measured to obtain the absolute distance between reference and test surfaces. The result is that the height of each pixel is calculated independently of its neighbors using fast Fourier transform algorithms.
Thus, the LightGage can measure parts with rough and diffuse surfaces, as well as parts with discontinuous surface features with very large differences in height pixel-to-pixel (>20mm).
The final result of an individual part measurement is a complete set of data characterizing the part's surface(s) that can then be used to extract a variety of relevant measurements such as parallelism, flatness, and low-order shape.
The LightGage technology is targeted at industries where part surfaces require tolerances on flatness, parallelism and/or part thickness on the order of microns; and is intended to be used in quality and process control. Applications include measuring hard-drive platters and components, fuel injectors, and high-precision pumps to name a few.
The technology offers customers the opportunity to gain a better understanding of their manufacturing processes, allowing them to reduce re-work and variability, ultimately saving them valuable time and reducing costs, while simultaneously increasing product quality and reliability.
Historical ContextThe LightGage system has been under development at Tropel for nearly 6 years, and during that time, the LightGage tool has been exposed to potential customers in many different industries. One request that Tropel engineers kept hearing from industry was that they develop a tool that can measure both sides of a part at the same time; or a tool that can deliver a precise measurement of a part's thickness. With the LightGage, both requests can be fulfilled simultaneously: use two LightGage sensors; one for each side of the part.
To that end, Tropel has developed various dual-head LightGage prototypes of varying complexity; but ultimately did not have the time or resources to dedicate a design team to developing a flexible platform for a dual-head LightGage system; one that can be easily assembled, modified, and demonstrated to potential customers. And most importantly, it had to work well enough to be presented to customers that may want to fund further development of the system to meet their individual needs.
Thus, Tropel needed a working proof-of-concept of a dual head LightGage that had some degree of automation (for throughput and usability purposes), was simple to assemble and operate, and, most importantly, cost-effective.
At the same time, the company was looking to develop new relationships in the high-tech community, and some of their engineers had gone through RIT's Multidisciplinary Senior Design. The decision was obvious--and so the task of developing the dual-head LightGage system has been given to the design team of P09701.
Design ObjectiveThe objective of P09701 (Corning Tropel LightGage Metrology System) is to develop a system around 2 LightGage interferometer"heads" (below)that allows for both the bottom and top surfaces of a part to be measured simultaneously. To do this, two LightGage sensors will be "aimed" at each other vertically.
The design team has been given responsibility for developing a fixture capable of supporting the interferometer heads as well as the part being measured. The system is to consist of a desktop computer, tunable laser diode (TLD), two LightGage sensors, mechanical structure, as well as a motorized head positioning system.
The final deliverable will be a working proof-of-concept for use in Tropel applications engineering and research; with the ultimate goal of sparking customer interest in the dual-head system, which will fuel further development of the system into a commercially viable package.
Concept DevelopmentA well designed concept selection process was implemented utilizing the provided tools for the course. The process began with a detailed needs assessment and continued through better defining the goal via metrics and priorities. Solutions were generated and measured to the needs of the costumer. The concept was developed and began to take shape when the costumer informed the team of a new, specification which required the abandonment of the first design. Alternatives were investigated and a new solution chosen through the same process and matured. The process used for concept selection is best summarized through the flow chart presented in Figure 5.
The detailed design phase followed an analogues process. After the System Level design Review, where the chosen solution was universally accepted, the design was refined and components selected. The team was again given a obstacle that would require the system to, once again, dramatically change. Due to the economy, the usable budget was considerably less than expected. In order to accommodate this change, the system was redesigned, and features removed. The result was a prototype system. The process flow chart and concept evolution are shown in Figures 6 and 7, respectively.
During the first few weeks of MSD2 it was obvious that the team was going to have to make due with a much smaller than anticipated budget and limited manufacturing resources. With that in mind, the team took the customer's feedback and, in an effort to save time and meet budget, redesigned the system so that it consisted of as few unique parts as possible, and ensured that the parts were as simple as possible in order to expedite fabrication. The result: out of 18 manufactured parts, only 8 are unique.
Not only was this significantly less expensive to make, but manufacturing took two weeks less than anticipated, and assembly was trivial.
How We Addressed the Customer's Needs
System ArchitectureThe system architecture was developed in order to break the system down into the fundamental components. Sub-systems were identified and developed with regards to inter-relation with one another. The system architecture is best shown visibly in Figure 8, where the boxes represent sub-systems and lines show which systems interact.
Thermal DriftWhen approaching the thermal drift problem in the design of metrology instruments, two solutions typically present themselves: use mass or exotic materials. The mass approach means using a large mass of thermally stable material to stabilize the system. Thus, all vertical components would be mounted to a granite slab in order to prevent thermal drift, as granite takes a very long time to react to any temperature change. However, due to size, budget, and transportation constraints, this option was not feasible.
The second approach, and the one used in this project, is to find a material with very low coefficient of thermal expansion such that it is affected minimally by ambient temperature changes. While materials such as quartz and carbon fiber meet these criteria, they were also not suitable to this application, as they would be too difficult to manufacture.
The material that the team chose, Invar, is a unique nickel steel alloy known for its unusually low coefficient of thermal expansion (alpha=1.2 x 10-6 K-1). The material has been used by Tropel for many years and has been proven to be very effective when used as a mounting structure in optics applications. This approach also yields the simplest solution to the problem of thermal drift: three 100mm tall invar posts are used to space the two heads.
A quick calculation can be performed given the material properties in order to estimate the amount the invar posts will drift over the course of a measurement cycle, given that the only heat introduced to the system is a change in ambient temperature. Perhaps more valuable, however, is knowing by how much the ambient temperature in the room needs to increase during a measurement cycle in order for the system to drift out of spec. This calculation is as follows:
When solved, it is found that the ambient temperature change would need to be on the order of 100 K for convective heat transfer to cause critical thermal expansion. It is important to note that this is due to the short period of time of a given measurement. Thermal expansion, due to convection, will occur over long periods of time, but this slow expansion will not affect individual measurements.
Furthermore, since the top head lifts off of the invar posts between measurements, the system requires that, when lowered, the top head comes to rest in exactly the same position every time; otherwise, measurement repeatability is severely compromised. To solve this problem, a steel vee pad is inset into the top of each invar post such that each is at precisely the same height. Coupled with spherical-tipped micrometers, this type of mount ensures that the micrometer tips come to rest in exactly the same position every time, and that this is the only possible position. The figure below illustrates that, when resting in the pad, the micrometer tip is only in contact with two points tangent to the sphere.
Furthermore, the vee pads are oriented such that the valley of each pad points to the center of the LightGage fizeau, as seen in figure. This orientation further ensures that there is no possibility that the micrometers settle in a different position measurement to measurement. The kinematic mount effectively constrains the system in every direction such that there is only one solution during the HeadtoMeasure routine.
In order to make the measurement process as straightforward as possible, the team decided to use a part support fixture, or "harp" on the bottom fizeau. A harp is a fixture with a wire of known diameter strung across in a weaving pattern, such that it looks like a harp. This fixture is then placed on the fizeau of the bottom head and is used to support the part under scrutiny. The user simply places the part on the harp and presses Measure. Because the part is a known distance away from the bottom fizeau (130-140 microns), setting up a recipe for the bottom side of the part is very straightforward. Thus, the bottom head in the dual head system is configured exactly like a single-head LightGage system, allowing the system to be disassembled and used as a single head system.
Vibration IsolationAs stated previously, the LightGage system is very sensitive to both thermal and vibration instabilities. While Tropel will use a granite table for both thermal and vibrational stability in a production model, this project used a pneumatic isolation table (donated by Tropel) in order to isolate the system from any ambient low frequency vibration. Measurement data from before and after the isolation table was floated suggests this significantly reduced noise and a marked improvement in measurement reliability was seen.
With the help of Dr. Raisanen, a vibration model of the LightGage head support structure was developed in order to further understand how the system will react to various vibration loads.
Coarse Motion IntegrationA major design goal for this project involved minimizing the necessary equipment and protocol interfaces while maximizing the flexibility of both the rate and range of positions required to be reached in an automated mechanical fashion. Additionally, it was determined that a minimum of 120-150 mm of travel was required to allow easy access to the measured part, and that this motion be achieved in minimal time for a high rate of throughput. Perhaps most importantly, it was necessary to be able to operate three such actuators in unison with virtually no discernable difference in speed or location.
After discussions with distributors of several types of linear actuators, it was determined that Mirai Intertech produced actuators that would be capable of all these requirements. The SCN5-150 Dyadic actuators have a total linear range of 150 mm with a variable speed of between 0.1 mm/sec to 200 mm/sec and repeatability of ~0.01 mm. This large variance in speed would allow for fast movement but also for precise and gentle lifting of the top head assembly. Additionally, each actuator is individually capable of producing thrust of 10.2 kgf, and thus three such devices would be more than capable of lifting the weight of the top head and mount with little trouble. A large weight tolerance also allows for a future iteration of the dual-head system to integrate additional functionality on the top head mount with little consequence to coarse motion accuracy.
In addition to their physical operating characteristics, each actuator contains a built-in motor controller, eliminating the need for additional hardware. This would meet the customer's requirement for minimal equipment, but unfortunately the controllers operate on the RS-485 protocol. While this is a well known and widely implemented protocol, it was not employed by the current Lightgage system. Fortunately, Tropel had used RS-485 on previous products and was able to supply a USB-to-RS-485 converter. This converter was able to conveniently and seamlessly allow all three actuators to require only a single USB connection. In addition to saving space, this single USB connection could be integrated into the existing Lightgage system control box.
A final specification which the Dyadic actuators were able to fulfill was the need for an emergency shutoff. The built-in controller provided just such functionality: a constant 24 VDC is required to be supplied to an ILK line. If this voltage is removed, the actuators are programmed to perform a maximum deceleration and cancel all movement. This action can also be executed on all three actuators simultaneously. A Keyence optical sensor is used to control this kill switch. This connection draws a maximum of 4 mA, and thus is negligible when compared to the power draw of the actuators themselves.
For power, the Dyadic actuators use a constant 24VDC and a ground. Each has a maximum current draw of 1.5 amps, for a total maximum power draw of 108 watts. Additionally, the RS-485 communications bus required a constant 5 VDC supply with minimal current draw. To satisfy this power consumption, Tropel was able to provide the design team with a 24VDC power supply with a maximum current of 10 amps. This power supply simply plugs into the wall, and is more than capable of providing the 4.5 amps for the actuators in addition to any power required by the communication systems and the Keyence optical sensor.
To provide the necessary 5VDC to the RS-485 communications bus, a simple resistive voltage divider was used. Since this connection would be drawing almost no power, resistor values were chosen to keep the overall current flow through this connection to a minimum.
CodingTo control the Dyadic actuators through a software interface, pre-written code was supplied by Mirai Intertech. This code was simply a collection of methods which allowed functionality of the motors to be controlled in a programming environment. Languages provided for were C, C++, and Visual Basic. Tropel's TMS software package is written in C++, so the supplied code proved to be an ideal solution. The pre-written methods were assimilated into the TMS software so that the Dyadic actuators could be controlled in a manner similar to any other motor used by Tropel. Furthermore, the speed of execution of this code allowed for all three actuators to move repeatedly in a synchronous fashion. Failure to do this could possibly have disastrous consequences.
The basic functionality implemented by the TMS software includes the ability to set the actuators to move at one of three preset speeds, move to a relative position, move to an absolute position, turn the servos on/off, home the actuators, and finally to receive feedback from the controller. The actuator controllers are able to report their absolute position, on/off status, whether motion is in progress or completed, and a variety of error codes which could be implemented by the TMS software to respond accordingly. These error codes include the deactivation of the ILK line, indicating that the emergency shutoff has occurred. All of this functionality is used to write scripts in the TMS software, which are in turn used to control the timing and motion of the actuators.
ScriptsTropel's TMS software package is very flexible, so integrating the Dyadic actuators into the measurement routine was very straightforward. In the "recipe" for each measurement, a script operator is inserted, which calls a desired function within a VB script defined by the user. The software then runs the appropriate function when the operator is called, either before or after the measurement (the user decides). For this project only 2 scripts were written: one that raises the head before the measurement, and one that lowers the head once the part has been placed in the tool. The basic order of operations during a measurement is as follows:
1. User presses Measure Button 2. Head to Load Position (send motor collection to absolute position 0). This raises the top head to the load position so user can place part in tool. 3. User is prompted by a message box to place part in tool. User must press OK to proceed. 4. Head to Measure Position (send motor collection to absolute position 95). This lowers the top head onto the invar posts, allowing the actuators to retract so they are no longer in contact with the lift plate. 5. User is prompted once again to verify that fringes are present in the live video. User must press OK to proceed. 6. System Measures Part. 7. Head to Load Position. Top head is once again raised after the measurement so that the user can remove the part.
The beauty of TMS is that the user can very easily tell the software to turn off all motions. So if the user would like to measure the same part multiple times, they can simply disable the script operators and the motors will not move the next time a measurement is taken.
Fine Motion: MicrometersWhile the customer had originally desired an entirely automated system, budget and time constraints meant that the fine adjust portion of the head motion would have to be manual. Because the instrument is so sensitive, in order to finely adjust the head position such that fringes could be found, it was necessary to use micrometers as the most sensitive manual adjustment method within the team's means. Thus, in order to ensure that the two fizeaus were parallel with one another, the user has to adjust the micrometers manually to find fringes. Originally it was thought that this would be challenging, as each fringe represents roughly 0.4 microns of tilt, meaning that in order to see fringes, the system would have to be within 5 microns of parallel. However, it was found that by setting all three micrometers to the same height (ruling), one was able to find fringes by simply moving the left and right micrometer up and down slightly. This is due to the system components being very precisely fabricated.
Part Detection: Keyence SensorWhile using actuators significantly simplifies the operation of the system, it poses another problem: what if the system is not set up correctly and the actuators crash the top head into the part? This situation would prove very costly, as both fizeaus would be damaged and need to be replaced. Thus, in order to solve this problem, the team decided on a method that has been proven by both Tropel and industry in general: a Keyence part detection sensor. Conveniently, Tropel used this exact method on another tool and the team was able to acquire the thrubeam mounts and fibers from Tropel. Dr. Raisanen had a spare Keyence sensor that he allowed the team to use, allowing us to complete this subsystem without using up any of our budget.
The sensor is wired through a relay directly to the motor controller kill switch. Both the sensor and the actuators run on 24V, so it was possible to configure the sensor in a way that when the beam is clear, 24V is flowing through the sensor to the motors; basically the system operates as if the sensor is not there and the actuators are allowed to move as instructed. However, when the beam is broken, the sensor switch is tripped and the voltage to the actuators drops to zero, causing the motors to perform maximum deceleration and canceling all motion before the head can strike the part. The thrubeam is mounted 4mm below the top fizeau in order to give the actuators sufficient distance to stop.
Results and Discussion
Final ProductAs written in the problem statement, the final deliverable for P09701 is a working prototype of a dual-head LightGage system for use in Tropel's engineering and applications research. The product being delivered to the customer is capable of: 1. Characterizing both sides of a part's surface over the period of one measurement. 2. System features actuators that lift the top head between measurements so that user can place or remove part from system. 3. Kinematic mounts ensure system returns to same position after lifting for every measurement. 4. Motions are fully integrated into TMS--no external software is needed to make system operate. It can be used just like any other Tropel Metrology Instrument. 5. Actuators have integrated motion controllers, connect to computer via USB, and run on COTS 24V power block. There is no external motion controller and no need for an expensive power supply. 6. Finding fringes is simple using micrometers, and only need to be adjusted for first measurement. There is no detectable drift in the micrometers. 7. Invar posts reduce the possibility of thermal drift. There is no detectable thermal drift during the measurement.
How it WorksThough it has been mentioned previously, the system is very easy to operate. The measurement process has been made as simple as possible for the end-user. The measurement process is as follows:
1. User opens DefaultLG_P09701.rcp recipe using TMS and presses Measure button. 2. System will lift top head to load position. If head is already lifted, it will stay there. User is then prompted to place part on system. User must then press OK button to continue. 3. Top head is then lowered to measure position. Top head is now resting on Invar posts and lift pads have disengaged from the lift plate. User is then prompted to verify fringes are present in live video. If no fringes are present, user must find them, and then press OK button to continue. 4. Part surfaces are measured sequentially--bottom surface is measured first, and then top surface. 5. When data collection is complete, top head lifts back up to load position so part can be removed. 6. Data is processed and displayed to user. Measurement complete. The measurement process takes about two minutes from start to finish.
How We Know it Works
FringesBuilding a dual head LightGage system as proposed includes two variables (risks) that could not be modeled before assembly: micrometer drift and ambient air mixing effects. While it was known that the invar posts would provide the needed thermal stability in the support structure, it was not known how much, if any drift there would be in the micrometers, especially after lowering the top head to the measure position. Lubricating fluid in micrometer threads can be especially frustrating, as it tends to settle when loaded--that is, the fluid migrates up the threads, causing the micrometer to continue to settle for a period of time. During this time, any measurement would be meaningless. Furthermore, because the LightGage is sensitive to thermal effects, there was no way to quantify the amount of unmixed air in the room (air currents of different temperatures swirling in the room), which could also negatively affect the measurement data.
The micrometer drift was to be explored by placing a weight on the tips for a period of time and then measuring the difference in the micrometers after the weight has been added. The desired specification of overall mechanical drift for the system was stated as 6nm. This would require the team be able to accurately measure the micrometer length and change down to the nm level. This is nearly impossible and infeasible for the team given that the equipment available was not accurate enough, nor could the measurement be outsourced due to the budget constraints. Therefore the amount of drifting done by the micrometers, or the system in general, could only be tested once the system was completed.
To test this design's viability, the only solution is to take a measurement using the top head and qualitatively analyze the measurement data. The top head is used for this verification because the upward looking head design has already been commercialized by the Customer and is known to work under various conditions. The downward looking head is the challenge, as it is mounted more than 100mm above the bottom head, and keeping the two fizeaus perfectly parallel is crucial. As can be seen below, it is indeed possible to take a reasonable measurement using the top head.
After a verification of the top head's ability to see a part below and have fringes available the drift was addressed. A very simple way to qualitatively see drift is to set the micrometers such that there are 3 to 5 fringes showing in the live video. If the fringes move to one direction or another after a period of time there has been a significant mechanical drift. For perspective, if the fringes move one fringe to the left or right (0.4 microns), the top head has drifted nearly 5nm. It can be noted that the fringes in the top head live video did not move and the micrometer drift was acceptable.
Though it cannot be shown in this paper, the effects of air mixing on the system can be easily seen in the live video as well. If the room is properly ventilated, the fringes in the live video should not move. However, if there is a current of warm air circling the room, the fringes will "dance" looking almost like waves on a pond. In this case, either the system should be moved into a more stable place, or an enclosure should be put around it. This depends mainly on where the machine is to be placed.
RepeatabilityThe system designed will eventually be sold to manufacturers that will be able to use it continuously with few adjustments. To achieve this level of use the lifting of the top LightGage head must be repeatable to avoid continuous changes to the micrometers.
To attain a maximum throughput for the measurements in a day the lifting action of the motors must be synchronized and smooth. This will ensure the same placement of the top assembly each time and therefore have a greater probability of consistency in measurements without readjustments every time. To check this a part can be loaded and measured, but not removed. The motors would then lift and settle as if a new part was being measured, but the same part would be measured for a second time. This measurement would be recorded, along with more measurements depending on the tolerance required by the customer. The measurements would be analyzed to assess their consistency of the measurements and problems in the system could be addressed and modified. Due to time constraints and the final acquisition of the last motor at such a late date, the repeatability tests were unable to be completed. These tests are suggested for further prototypes as this is merely a proof of concept.
Meeting Customer NeedsThe system that the group will be delivering to the customer is a working proof-of-concept of the dual-head LightGage system. It is able to measure both sides of a part during the same measurement cycle; it can be assembled on existing LightGage granite, making it possible to install as a bolt-on upgrade to single-head systems; and it uses low-cost, compact, and easy to set up actuators that have been fully integrated into Tropel's measurement software package. The team has demonstrated that the system is virtually fool-proof, very easy to operate, and successfully measures both sides of a part. The product that the team is delivering is mechanically complete, and will be 100% functional once the customer works out bugs inherent in their system that only they have the ability to fix.
It is not known, however, how meaningful the measurement data is. There are various reasons for this:
1. The team is not familiar with the optical set up of the LightGage head. Adjustments were made to the alignment in the heads such that measurements could be taken, but it is not known whether or not this will affect the accuracy of the data collected.
2. Though familiar with the TMS software, the team did not have the depth of understanding or the capability to fix bugs that were found in the software. Thus, a true thickness measurement could not be taken because the software would not allow the user to input the correct part feature height. This problem was not resolved before delivery. The customer made it known that they were unsure what effect their more recent software revisions would have on our ability to get the system working properly.
3. The team was unable to reach an acceptable level of measurement repeatability due to the conditions inside the metrology lab. The temperature in the room was rarely stable--often ranging in temperatures between 65 degrees and 75 degrees, and various levels of humidity. Furthermore, the air currents constantly flowing through the room caused the fringes in the live video to "dance" further eroding any attempt at repeatable measurements. The customer, however, has appropriately environmentally controlled space where the system will reside once delivered. This should significantly increase measurement repeatability.
Overall, however, it is believed that the team is delivering a successful product. While the customer wanted a dual-head LightGage, it was just as important to them that they see how a team operates with very few resources. Indeed, the team had less than $3,000 to develop the next generation LightGage system (not including any Tropel-sourced LightGage hardware). The customer was interested in finding actuators that could be connected to a computer via USB without the need for an external motion controller. The team found actuators that fit their description and were able to power them using an off-the-shelf 24V power pack. Furthermore, we were able to deliver a very robust and cost-effective system that re-uses many of Tropel's stock parts and uses a minimal amount of custom fabricated components.
4. The system is not without fault. The ability to measure both sides simultaneously is not yet within reach. The software limits the action as does the brightness of each LightGage. The light from one head can over saturate the other and ruin a measurement. This requires the LightGage heads to measure one at a time in order to have useful output.
5. The system is operational and proves the concept of a dual-head LightGage; however, there are improvements that can be made. This system was created on a modest budget and with more funds could have been made to work better. The worry of ambient temperature fluctuations affecting the LightGage, for example could have been avoided by the creation of an enclosure that had the option of temperature control. This would eliminate the issue of temperature no matter the environment.
Vibration was addressed by the air-table, although previously addressed by a granite slab generally used for other LightGage operations. The advantage of the granite was the pre-drilled holes in a strategic pattern for the placement of fixtures for the actuators and base plate. If the pre-drilled hole configuration could be placed onto the face of the air-table in lieu of the pre-drilled grid, the system would be more stable and the dismantle/re-build would be more accurate and repeatable if moving was an issue.
The system does not need constant adjusting between measurements, although some adjustment via the micrometers is necessary. The adjustment is not always intuitive and takes some skill to know where to move the micrometers and which micrometer to adjust first. For a customer, learning this task can be time consuming and the time lost to a learning curve can be expensive. This can be avoided by the replacement of the manual micrometers and replacing them with motors. This automated addition would simplify the process for the customer and help to avoid lost time while learning a new process by allowing for software to automatically adjust the fine motion.
Small adjustments to improve the system include a better computer to run the software. This would ensure quick measurements and faster changing times. The hardware of the entire system should be of one type. Currently some parts are English while the majority is in metric. The material for the lift structure could also be made of a more robust material to avoid wear from bending under the weight of the top assembly when lifted.
AcknowledgmentsSenior design team P09701 would like to thank Corning Tropel for all of their support, technically and financially. We would also like to express our gratitude for our advisor Dr. Raisanen and all of his help throughout the project.
References Thomas J. Dunn, "Frequency-Scanning Interferometry: Advances Precision Component Manufacturing and Assembly", Photonics Spectra, June 2005, pp. 96-100
 Thomas J. Dunn, Nestor O. Farmiga, Andrew W. Kulawiec, Joseph C. Marron, "Frequency-Scanning Interferometry", Corning Tropel White Paper