Visual and Profilometric Fuel Rod Inspection
A design concept was developed for the visual and profilometric segments of INL's MEITNER fuel examination platform.
|Final Design Concept|
|Team Name||Team MEITNER|
|Duration||Summer 2016 - Fall 2016|
Current inspection of nuclear fuel cells requires the use of cumbersome hot cells. INL is investigating an alternative process for inspecting these fuel cells that can be located closer to reactor sites and does not require a dedicated building for operation. Our task is to specify segments of this process that will perform non-destructive profilometric and visual inspection of irradiated test specimens. The processes must be fast and easily serviceable while maintaining a high level of fidelity and flexibility.
Inspection of nuclear fuel rods via current methods is quite cumbersome. Irradiated materials must be transported to an examination facility such as INL's HFEF. There, the fuel is inspected inside hot cells, large led lined rooms with argon atmospheres where technicians must use robot arms to interact with the material. It is difficult to implement new inspection solutions within this testing environment due to the long testing time. It is expensive to test and to improve testing methods. INL is devising a new environment for inspecting materials known as the Modular Examination Instrument for Transportable Nuclear Energy Research (MEITNER). This device is a modular stack of examination equipment that can simultaneously inspect materials via non-destructive, non-contact measurements.
Our mission is to design and specify the equipment of the profilometric and visual test cell. A mock-up cell featuring analog equipment is to be designed as a proof of concept of the INL cell design. An extensive design document is to be included detailing equipment specifications, testing procedure, limitations, and avenues for future development.
|Major Requirements||Acceptable Value||Ideal Value||Priority Level||Comments|
|Outer Diamter of module||N/A||104.5||high||The outer diameter is large enough to support the widest cask diameter.|
|Inner diameter of module||N/A||58.73 cm||High||The inner diameter is wide enough for the widest inner cask diameter.|
|Tortuous path||90 degree turn that prevents direct line of sight||Greater than 90 degree turns to impede scattering of radiation to exterior||Med||The tortuous path must prevent radiation from escaping to the exterior environment.|
|High degree of dimensional accuracy||<15 micrometer||<10 micrometer||high||Profilometry should be capable of detecting small deformations in part profile.|
|High degree of optical zoom||20x||100x||med||High degree of zoom is necessary to observe surface defects upon fuel.|
|radiation resistance||2 months at 1200 rad/hr||4 months at 1200 rad/hr||low||Technology must be capable of measuring part before failure.|
|Post irradiation inspection is completed in a timely manner||<24||<8||low||Ideal inspection should be completed in a work day.|
|Camera view frame||10.2cm >||28cm >||med||View angle able to view largest specimen and smallest specimen with visual fidelity.|
Within the HFEF at INL profilometric inspection is currently performed using a pair of LVDT sensors within their hot cell. These sensors are placed against the material and removed for every measurement at a resolution of 1,2,5, and 10 measurements per millimeter. This causes some examinations to last 8 hours for a single element. LVDT are contact tools for dimensional inspection though they do not present a significant risk of damaging test elements. Non-contact inspection methods are preferred, such as those using lasers. Laser systems may additionally save time by eliminating the time required to reset the sensor for every measurement. Promising laser imaging techniques include laser interferometry and laser triangulation, laser micrometry, and line scan imaging.
|LVDT(Linear Variable Differential Transformer
These sensors use electromagnetism to translate deflection of a plunger into a voltage/amperage. The output from the sensor must be conditioned. Based on the conditioning used, these sensors are capable of infinite resolution and high cycle life. They are rad resistant but require a very stable frame to maintain alignment and to prevent shaking loose. These are currently implemented in INL's hot cell. Though they are contact type sensors, there is little risk of surface damage.
|Line Scan Camera
Line scan cameras work like normal cameras but capture only a single row of pixels with each exposure (resolutions of 1x2048 are common). They possess the ability to capture images very quickly and are usually used on quick conveyor belts. The output of these cameras can be an image or a spectral intensity plot. A machine vision system could use this information to determine the width of a silhouette of its target. this system could be set up in the same way as a visual inspection system. Radiation tolerance is not standard on these cameras, frequently trading the line scan property for low-resolution area scanning and rad tolerance.
|Laser Triangulation Sensor
Laser triangulation sensors are capable of very high accuracy measurements on flexible ranges. The laser fires at high frequency from the sensor creating a spot of light on the surface of an object. A CMOS line scanning camera within the sensor housing detects the position of the laser spot on the surface and outputs a voltage. Resolution of the sensor is dependent upon its offset from its target, so it must be located within the MEITNER cell to possess adequate resolution. These sensors are not rad resistant and would need to be located far from fuel and/or behind shielding. Sub 10 micron accuracy would also require the sensor to be located less than 50 millimeters from the target, therefore this sensor type is unsuitable for the MEITNER environment.
These sensors operate similarly to laser triangulation sensors. A laser is shot inside of the sensor housing at a rotating mirror. The mirror deflects the laser across a lens creating a scanning “sheet” of laser light. A receiver translates the light that it retrieves per pass into dimensional information of parts that block light in between the lenses. These sensors only capture profile based information on a target, therefore the thickness of curved plates cannot be determined with this equipment. These sensors are built with specific scanning widths which define the largest object that could be measured by the sensor, up to around 2 inches. These sensors are also fairly expensive and contain more sensitive equipment than laser triangulation or LVDT. They are not rad resistant and would need to be frequently replaced.
The most promising method of profilometric inspection is laser interferometry. This method was investigated by the project, Fuel Rod Defect Detection. This system uses the interference of a reflected lasers beam when compared to a reference mirror in order to determine distance. This method boasts very fine accuracy and radiation tolerance but is limited by the short scanning depth, though its standoff distance is large enough to keep it a reasonable distance from the irradiated material.
Within HFEF standard digital cameras take pictures of test elements through 4 feet of oil filled glass. The discoloration of the glass causes images to lack true color. It is also difficult to image the entire specimen due to the hot cells fixed viewpoint and cumbersome manipulation of specimens. Some inspection is done by using binoculars located at each window. Due to MEITNERs closed off environment, the current techniques for visual inspection are not possible, but the ability to significantly reduce the radiation impact on test equipment presents an opportunity. Standard digital cameras, who live for a very short time within the hot cell, will be able to withstand the duration of a parts testing cycle allowing true color photos of the element in 360 degrees. This is due to the lower rad/hr dose that is present within MEITNER as well as the potential to position the camera where there is a significant amount of shielding between it and the fuel specimen. Rad shielded cameras may be used to extend the time between replacement and may allow them to be placed much closer to the test element. Snake cameras present an opportunity for testers to look around the far side of a specimen within the cell through a tortuous through hole, though the small length of articulating armature reduces versatility and may necessitate multiple insertion points.
Each lead shielded ring of MEITNER will be capable of housing multiple inspection tools. Stepped wedge-shaped plugs will be through cut from the walls of MEITNER. Testing methods may be devised using one of these wedges as their access point to the nuclear material within. In order to minimize radiation escaping from the vessel, any through cut hole in the shielding must follow a tortuous path, a path that does not have a line of site connection between each end. This is critical to the safety of the facility and allows equipment to operate in a lower rad environment which reduces degradation. Due to the meandering of the through hole light must be bounced to meet testing equipment or equipment must possess enough radiation tolerance to last within the inner cell.
In our investigation, the use of mirrors to reflect light through the walls of MEITNER to reach a camera on the exterior would prove more problematic than needed. Mirrors would be difficult to replace and the cameras field of view would be restricted to a window the size of the through hole without the use of fish-eye lenses that would distort its image. For this reason, radiation tolerant cameras were preferred. Few non-contact inspection methods were found that possessed the radiation tolerance and flexibility to function to specifications in the MEITNER system, though through our investigation we concluded that the best course of action would be to internalize this part of the module as well. The method that stood out was laser interferometry, specifically NovaCams Microcam. To accurately position the laser within the cell a linear motion stage is needed with at least a stroke length equal to the largest plate with to be inspected within MEITNER, 11.5 inches. Though radiation tolerant equipment has a higher investment it will need to be replaced less frequently, which will outway the costs of decontamination and service.
In order to simulate the restrictions of the modules in the MEITNER platform, a frame was constructed with similar dimensions. This frame is the mounting point for all the analog equipment that will be in the cell. It is constructed from angle iron. The total height is 18 inches. The total length is 12 inches. Total width is 8 inches. To eliminate some of the unnecessary design complications the wedge was designed without curvature of MEINTERs walls considered. The top half of the wedge is left open to provide viewers context for the tortuous path that cables must follow through the wall. The sides are covered with sheet metal. Inlaid plates bolted to the rim of the sheet metal frame provides mounting points for the camera and laser system.
The Univeristy MEITNER mockup will be constructed from 24-inch Sch 40 plastic pipe. The inspection module will not enter MEITNER perpendicularly, instead, it will interface at an offset angle to provide the necessary room for equipment. The camera will be mounted to the flat front surface of the module. The profilometric inspection will be performed by a laser triangulation device that is mounted to a linear motion stage which is also mounted to the inspection module. In order to provide additional stability, the linear motion stage will be mounted to a long plate running through the wedge which is mounted at the front and back faces of the wedge.
INLs MEITNER setup will differ from our own mostly due to different equipment. The camera used by INL will mount in the same position as the analog camera. Instead of a laser triangulation device, we recommend that INL uses a MicroCam 4D sensor for profilometric measurements. This piece of equipment will be a small probe of less than 1 inch in diameter that release information to an external data processing unit via rad hardened fiber optic cable. The same linear motion stage can be used here, though the mounting plate should be made from a material that will minimize deflection.
The analog camera used within our version of the visual inspection section of our module is IntelliSecu 4 camera, model number ISC-H4RO-T. This camera boasts specifications nearly identical to the radiation tolerant camera that we recommend for INLs module. The camera is capable of 10x optical zoom and additional digital zoom up to 100x. The camera automatically switches to infrared in darkness. It also features pan and tilt functions to direct the field of view. It can be programmed to follow specified tracks or configured for manual control. It was designed for viewing objects in the range of tens of feet and therefore has some difficulty focusing on objects within the first foot or two of the camera.
|Laser Triangulation Sensor
In order to collect information about the dimensions of a specimen, our system uses Wenglors OPT 2005 photoelectric laser sensor. This sensor operates by laser triangulation principles. It operates within a 50 to 350 mm sensing range, outputs a 0 to 10-volt dc analog signal and requires 18 to 30 volts dc to operate. It is capable of taking 800 measurements per second and has a reported resolution of >50 micrometers. The sensor works with interfering light and on reflective surfaces.
|Arduino Control System
Power is delivered to the linear motion stage control motor and laser via an 110 to 24 volt AC to DC switching power supply. The linear motion stage is driven via a voltage source from this power supply and a signal from the Arduino board directed to an A4988 stepper motor driver board. The Arduino is configured with connections to this boards step size pins, direction pin, and step pin. Every logic high pulse to the step pin directs the stepper motor to proceed one step, by this function the pulses per second can be controlled within Arduino code. The analog signal produced by the laser triangulation sensor is directed to the LabJack U3-HV model analog to digital converter. This ADC has 16 12 bit analog inputs which will bottleneck the accuracy of the laser sensor if the analog output is not taught to a short deflection range.
|Linear Motion Stage
To provide the versatility to move the laser sensor,e.g. for sweeping measurements of specimens, a Thomson MicroStage Linear Unit was used. This stage was selected to have a large enough stroke length to accommodate the widest specimens that are to be tested within MEITNER, which is 377 millimeters or about 14.8 inches. This stage is screw driven and ball guided. Every revolution of the drive screw provides 0.025 inches of travel and the stages max speed is 0.83 inches per second. Possesses anti-backlash nut for high repeatability factor. This same motion stage can be used in INLs system though rad resistant stepper motors and screw grease are necessary. There is also a concern that this motion stage is discontinued by the manufacturer and an alternative must be scouted.
In order to support the motion stage, an aluminum bar was screwed to its base. This bar extends the length of the stage and through the wedge module where it is fixed with bolts. The laser sensor is fixed to the motion stage by a tapped aluminum plate which attaches to the stages moving platform.
|Rotation Platform and Controller
The rotation platform and controller are manufactured by Optis-Focus. The rotation platforms stage has an outer diameter of 100mm, a transmission ratio of 180:1, and a max speed of 25 degree per second. The controller runs in step units. Capable of rotating the platform to a specified position at a given speed, both in steps. The controller is capable of delivering full, half, quarter, and eighth steps. This system is used to emulate the capability of the elevator that will move fuel through MEITNER, sans the ability to change elevation. By controlling the rotational speed of this device scans can be made radial around the object, and the part can be precisely rotated to view the front and backside.
The frame was constructed using angle irons, cut at forty-five degrees on the corners and welded together. The sides are sheet metal fixed via jb-weld. The linear motion stages cantilevered beam was bolted an angle iron which was bolted to the inside of the frame. The beam rests on a fulcrum made of an eighth-inch steel bar bolted to the front of the frame. The camera is bolted to a pair of eighth-inch steel bars bolted to the top of the frame. The frame was painted black with automotive paint to give it a consistent look.
The MEITNER cylinder was mocked up via a 23.25-inch pipe with a cut out for the module to fit as it would in the INL system. The internal stresses in the pipe caused it to contract after being cut. To remedy this a circular piece of MDF board was cut to the standard diameter of the pipe. This wood piece hold the pipe to the proper diameter.
Two different testing procedures were investigated to demonstrate how the profilometric setup could be used to capture information about the dimensions of a subject. The first was a radial inspection for viewing cylindrical specimen. The second is a sweeping method for more diverse geometry such as curved plates.
For the process of radial inspection, the laser was manually moved to the position where it read the lowest distance value. This corresponds to the closest part of the surface of a cylindrical specimen. The specimen is then rotated at a fixed rate while the laser records measurements. From this data, a plot can be created of time versus distance from the laser sensor to the subject. These can be transformed by a known distance to the center of rotation and a rotational speed into the angle and radius domains. From the data collected bumps were found on several parts in consistent locations over several passes. Though they cannot be tied to physical deformations of the subjects as the bumps were quite large, and the subjects very circular, these bumps may correspond to other defects. They could be tied to changes in surface condition or reflectivity or to a hitch in the mechanism of the rotational platform. The reflectivity of the specimen may have reflected the laser light into the sensor in a way that interfered with its accuracy as well. Further investigation of the equipment may be focused on determining what the cause of these defects is.
Sweeping inspections are to be used for objects with irregular geometry. For this method, the laser is moved on the linear stage at a fixed rate across the surface of the specimen. This provides a good scan for determining the relative deformation of a subject. The change in curvature of curved plates can be determined from these scans as well. By rotating the specimen 180 degrees with the rotation stage an additional pass can be made of the reverse side. In order to pair this scan to the front side, it must be flipped horizontally and vertically in such a way that it accurately represents the distance between itself and the front side of the plate. Pairing the points on the front side with the corresponding point on the reverse side presents an opportunity for additional development on this project the techniques evaluated during this project present opportunities for the loss of alignment.