United Kingdom’s 15 nuclear reactors are responsible for producing 21% of its total electricity. 14 among 15 of those are Advanced Gas Cooled (AGR) reactor which uses graphite as the moderator. The problem with graphite moderated reactor is that the mechanical and physical properties of the graphite components are changed by fast neutron irradiation and radiolytic oxidation. It is observed that the graphite components has a 3% reduction in dimension under these exposure. Thus, it is crucial for EDF to have a device which can inspect these kind of changes to increase the power plant’s lifespan. The current technique used involve the use of a Channel Bore Monitoring Unit (CBMU) and a Channel Viewing Camera (CVC), but both of this devices require a lot of time and cost a lot of money to the company. Thus, this project tested the reliability of a Light Detecting and Ranging (LiDAR) system as a replacement to the CBMU device in detecting the deformation in the graphite core. The RPLiDAR A2, which has been chosen for this project, was attach onto a ET50 Screw Drive Electric Actuator and placed inside a start of life graphite core. A 2-dimensional and 3-dimensional scan was carried out and the data was processed using AutoCAD 2017. The result obtained shows a very high possibility of this device replacing the CBMU as deformations can be easily monitored and large surface crack can also be seen on the 3-dimensional scan clearly as line of missing points. Not only that, the LiDAR system also requires a very short period of time to gather the required informations.

Contents

1 Introduction

1.1 History of United Kingdom’s Nuclear Power Generator

2 Literature Review

2.1 Graphite Reactor Core

2.2 Current Inspection Techniques

2.2.1. Channel Bore Monitoring Unit (CBMU)

2.2.2. Channel Viewing Camera (CVC)

2.2.3. Prototype Eddy Current Inspection Tool (PECIT)

3 Materials and Methods

3.1 Introduction to LiDAR

3.2 RPLiDAR A2 360° Laser Scanner

4 Results

4.1 2-Dimensional Point Cloud Data

4.1.1. MATLAB R2016b

4.1.2. AutoCAD 2017

4.1.3. 1mm Wire Test

4.2 3-Dimensional Point Cloud Data

5 Discussion

5.1 Experiments using RPLiDAR A2

5.2 Future Research

5.2.1. Improving the LiDAR System

5.2.2. New Technology

6 Conclusion

List of Figures

Figure 2.1: Material properties of an isotropic graphite.

Figure 2.2: A channel bore monitoring unit (CBMU) with wheels and feelers.

Figure 2.3: Data gathered using the CBMU at Channel 24:70, Hunterston B Reactor 4.

Figure 2.4: Sample of data obtained from the CBMU and a load trace data.

Figure 2.5: Television Remote Inspection Unit Multi-Purpose Head (TRIUMPH) assembly.

Figure 2.6: Data obtained using the Prototype Eddy Current Inspection Tool (PECIT).

Figure 3.1: Class of lasers and its descriptions.

Figure 3.2: The concept of a Light Detecting and Ranging (LiDAR) system.

Figure 3.3: Result for the weighted scoring method.

Figure 3.4: Parts received with the RPLiDAR A2.

Figure 3.5: Output result from the ‘ultra_simple’ application.

Figure 3.6: Representation of the variable ‘theta’ and ‘Dist’.

Figure 3.7: Output result from the modified version of ‘ultra_simple’ application.

Figure 3.8: Results saved in a text file using CSV format.

Figure 3.9: ET50 Screw Drive Electric Actuator.

Figure 4.1: Result obtained by using MATLAB R2016b.

Figure 4.2: Result obtained by using AutoCAD 2017.

Figure 4.3: Result obtained when a wire is placed onto a wall.

Figure 4.4: Top and side view of the point cloud data obtained using the RPLiDAR A2.

Figure 4.5: Presence of crack inside the point cloud data.

Figure 5.1: The Confocal Chromatic sensor.

1            Introduction

1.1          History of United Kingdom’s Nuclear Power Generator

United Kingdom’s first nuclear reactor was built in 1946 at the Atomic Energy Research Establishment at Harwell in Oxfordshire. The 3 kWth air-cooled graphite-moderated GLEEP (Graphite Low Energy Experimental Pile) started its operation in 1947. In 1948, the 6 MWth British Experimental Pile ‘0’ (BEPO) was commissioned, which was the first large reactor outside the USA. The reactor was used to demonstrate the viability of commercial power reactor.

In 1953, UK’s government announced a civil nuclear power program which resulted in the Atomic Energy Authority Act 1954 to create the United Kingdom Atomic Energy Authority (UKAEA) in the following year which responsible for the development of nuclear reactor technology. Their first achievement was in October 1956 when they successfully built the first of eight small Magnox prototype units. The Magnox reactors served two purposed which was to generate power as well as producing plutonium for the military. The latter Magnox reactors were built at a much larger scale and optimised for production of electricity. There was a total of 26 Magnox reactors in the UK with the last reactor, Wylfa 1, being shut down in December 2015.

A paper titled ‘The Second Nuclear Power Programme’ published in 1964 marked the next phase of the UK nuclear power program. The UKAEA’s advanced gas-cooled reactor (AGR) was adopted as the UK standard for the design. AGR is more efficient due its operating temperature of around 650°C compare to Magnox reactor which only operates at around 360°C ^[3]. As of October 2016, the UK has 15 nuclear reactors in which 14 of them are AGR units and one of them is a water reactor (PWR) technology ^[9]. The problem with AGR units was most of them are running at significantly less than the original design capacity.

In the late 1990s, nuclear power plants generated almost 25% of UK’s total electricity but this number has gradually decreased as old plants have been shut down due to ageing-related problems. Now, UK’s 15 nuclear reactors are producing 21% of UK’s total electricity, but again, this number is expected to decreased as most of the reactors are expected to shut down in the 2020s and 2030s. However, EDF Energy, the owner of all 15 nuclear reactors, are currently spending millions of pounds to upgrade and maintain the reactors to increase its total lifespan.

One of the biggest problem with AGR reactors are within its graphite reactor cores. They are made up of Gilsocarbon graphite which tends to deform and crack under high stress. The solution for this was to use a device which can scan for deformations and deliver the information straight to the people responsible. The only problem with the devices available now is that it takes too long to scan the entire reactor core. Thus, a more efficient device is needed to overcome this problem.

2            Literature Review

2.1          Graphite Reactor Core

Most of UK’s nuclear reactor are AGR units which uses graphite as the moderator and carbon dioxide as the gas coolant at a pressure of 41 bar ^[8]. The graphite used was manufactured from a special coke obtained in the USA known as Gilsonite ^[8]. Nuclear graphite requires a very high purity to allow the moderation process of neutrons unaffected. This was controlled by, (a) carefully inspect the purity of the raw materials, (b) treatments with halogens to form volatile halide salts which is then removed, also (c) by utilising high graphitisation temperatures to help impurities diffuse out ^[9]. Note that the nuclear graphite is an anisotropic material due to the high amount of coarse porosity made by the large filler particles. Nuclear graphite has densities of around 1.7 g cm^-3. The difference compared to that of pure isotropic graphite, refer table 1 for full properties, is about 20%, due to the porosity.

Material Property of Isotropic Graphite	Values
Density	1.81 g cm-3
Mean coefficient of thermal expansion	4.35 x 10-6 (°C)-1
Poisson’s ratio	0.2
Dynamic Young’s modulus	10000 MPa
Ration of SYM to DYM	0.84

Figure 2.1: Material properties of an isotropic graphite.

The problem with graphite moderated reactor is that the mechanical and physical properties of the graphite components are changed by fast neutron irradiation and radiolytic oxidation. Neutron irradiation occurs when neutrons from the fission process collides with the nuclei of the moderator causing displacements of atoms which leads to vacancies and interstitial loops in the moderator. Not only that, it also causes stress build up in the moderator and other additional strains such as irradiation creep (deform plastically) and irradiation induced dimensional change. On the other hand, radiolytic oxidation, unlike thermal oxidation, occurs significantly in a low temperature carbon dioxide-cooled graphite moderator, 700°C and lower. This happens when the carbon dioxide is decomposed by ionising radiation to give reactive oxidising species. The equation for the reaction can be seen below:

C+CO2=2CO

(1)

In 1960s, methane was shown as a powerful inhibitor which was then added along with carbon monoxide (~0.025% volume and ~1% volume respectively) to the carbon dioxide coolant to control graphite oxidation on the internal porosity ^[9], but this also has its own consequences as methane without a doubt reduced the oxidising effects but also has the possible detrimental effects on the fuel ^[5].

Neutron moderation is a process whereby the neutrons are slowed down to a thermal energy level of <1 eV by elastic collision with the carbon atoms inside the graphite. This is likely to cause carbon atoms to be displaced from the original lattice, thus, producing vacancies and interstitials. Eventually causing dimensional changes to the graphite moderator. Review by Tucker and Wickham suggested that during the initial period of the graphite moderator, there will be a bulk shrinkage due to the closure of small pores and cracks. After few working years, there will be a ‘turnaround’, occurring between 100 to 125 x 10²⁰ n cm^-2 (EDND), when the shrinking cracks are unable to withstand new irradiation-induced crystalline growth. Thus, an expansion of graphite moderator occurred.

Under simultaneous exposure to neutron irradiation and radiolytic oxidation, Brocklehurst ^[11] observed that (a) the ‘turnaround’ is delayed and occurred at a higher dose of neutron irradiation and (b) the subsequent growth rate is lower. For 32% radiolytic weight loss, the dimensional change curve shows a decreased from -1.8% to -3.5% and the ‘turnaround’ occurred at 190 x 10²⁰ n cm^-2 (EDND). By adopting Brocklehurst and Kelly model on dimensional change, an AGR peak-rated brick typically will undergo a radiolytic weight loss or around 38% at a dose of 200 x10²⁰ n cm^-2 (EDND).

In 2006, Tsang and Marsden observed a 3% decreased in dimension for both the height and the bore radius when they did a material model test involving simultaneous exposure of neutron irradiation and radiolytic oxidation ^[12]. Ernest D Eason together with Graham Hall and Barry J Marsden analysed the dimensional change of Gilsocarbon graphite when irradiated under inert environment. Based on 1269 points obtained, they conclude that the typical dimensional change is around 1% to 3% with uncertainty of a few tenths of a percent ^[13].

To ensure a safe working nuclear reactor, the graphite structure needs to remain undistorted and strong to maintain fuel cooling, permit loading and unloading of fuel, and allow the necessary movements of control rods in both normal and fault condition.

Since the graphite cores are not replaceable, its condition is often seen as one of the limiting factor of the plant ^[9]. Thus, a number of techiques and devices has been used to monitor these changes that occur inside the graphite cores. Sub-section 2.2 will discuss on some of the devices used to perform this task.

2.2          Current Inspection Techniques

2.2.1. Channel Bore Monitoring Unit (CBMU)

Figure 2.2: A channel bore monitoring unit (CBMU) with wheels and feelers.

This device is being used to examine reactor cores during its periodic shutdowns, roughly every 3 years during the refuelling process. The data obtained from this device is processed by a computer to produce information about the core diameter, ovality, channel distortion and brick tilts. It is also used together with load trace data to detect cracks that are growing inside reactor cores.

The CBMU device weigh at around 230 kg with 8 wheels, 4 at the top and 4 at the bottom. There are also 4 feelers, situated 90° apart from each other, which are used to measure the diameter of the core as the device is raised from the bottom of the core by a delivery hoist. The feelers act as a Linear Variable Displacement Transducer which takes measurement every one-millimetre interval of height. The feelers have a resolution of 0.02 mm. Diagram below shows the data gathered regarding the diameter of the core using this device for channel 24:70 at Hunterston B Reactor 4 ^[14]:

Figure 2.3: Data gathered using the CBMU at Channel 24:70, Hunterston B Reactor 4.

The CBMU device also contains two tilt transducers, which allow the device to determine the tilt of the core up to 0.1°. For the device to be able to scan the whole core, there is also a rotary joint located at the upper section of the device which allow it to rotate about the vertical axis. This allow scans to be carried out at a 30° increments to eliminate systemic errors caused by equipment manufacturing tolerances ^[14].

During these periodic shutdowns, large volumes of data can be collected from a few selected reactor cores, roughly eight, which were selected by using a software called ‘Channel Selection Assistant’. They are selected based on factors that are relevant to the safety of the power plant. By using these accurate data from these few selected cores, an overall assessment of the health of the whole plant can be predicted. This however comes with a disadvantage as gathering those information takes a lot of time due to several procedures that must be go through, the channels selected must be emptied from any fuel for the device to be lowered down, the device must be raised very slowly through the entire channel in order to get a high accuracy data. This may take several hours. Once done, the fuel must be reinserted into the channels.

All of this must be done during its shutdown period, which for a nuclear power plant can cost up to one million pound for one full day of not working.

This device is also being used during the refuelling process, which takes place every week, to gather a large volume of less detailed information relating to height and load of a fuel assembly. The data obtained are used to provide information relating to the presence of cracks within the core ^[15]. Diagrams below shows two different data, one obtained using the CBMU device, and the other is a load trace data:

Figure 2.4: Sample of data obtained from the CBMU and a load trace data.

This two data will be fed into a system called “Intelligent Crack Detection System”, whereby the data relating to the region of interest is extracted. These data will then be examined for peaks as this will show the location of brick layers and also the presence of cracks inside the core.

2.2.2.    Channel Viewing Camera (CVC)

Visual inspection is usually done after CBMU measurements as results from CBMU may indicate areas that require extra attention. This is where this device comes in handy, TRIUMPH (Television Remote Inspection Unit Multi-Purpose Head) camera, is a general purpose close-circuit television (CCTV) inspection tool for AGRs. It is a modular system meaning that it can be assembled with different modules depending on the requirement of the task ^[22].

Figure 2.5: Television Remote Inspection Unit Multi-Purpose Head (TRIUMPH) assembly.

A TRIUMPH assembly will consist of four main modules:

1. The Rotate Module, which situated at the top-most position of the TRIUMPH assembly. Its function is to enable the other module attached below it to rotate 360° together to avoid picture tumbling.
2. The Channel Viewing Mirror Module, which consists of a camera, that can look sideways at the channel wall and straight down the channel, and two lighting assembly.
3. The Zoom Module, which has an attached 6:1 zoom lens that allow closer inspection of certain area inside the channel.
4. The Prism Module, which is usually attached to the in front of the Zoom Module to increase the field of view of the camera to 105° from forward to side.

2.2.3. Prototype Eddy Current Inspection Tool (PECIT)

As previously mentioned, graphite core has the potential to crack as a result of dimesional change through its operating lifetime. The first two methods discussed above mainly focused on the dimesional change and surface cracks. Non of both can detect subsurface cracking that initiated from the external keyways and does not propagate to the core surface.

Since graphite is an electrical conductor, eddy current testing has been proposed as it is a suitable technology to solve this problem. The main aim was to detect cracking which does not show on the surface of the core. In 2007, Phoenix Inspection manufactured the phase 1 of eddy current testing device. It uses four 70mm-diameter coils that can perform rotational, axial, and helical scan patterns. In the late 2008, James Fisher Nuclear Limited designed the proof-of-principle eddy current tool (PoPECT). It relies on the same hoist system used by the CBMU and uses the same coil as the phase 1. The downside of PoPECT was it takes about 45 minutes for the device to warm up to the channel temperature, but that does not stop the device from being use at Heysham 1, reactor 1 in June 2010.

Following the successful deployment of PoPECT, a prototype eddy current inspection tool (PECIT) was manufactured. The difference between PoPECT and PECIT was that PECIT has four wheels at the top and bottom section with the centre section able to rotate. Pecit also has three probes – impedance bridge probe, transmit-receive probe, and differential probe. Figure below shows the result obtained by using PECIT:

Figure 2.6: Data obtained using the Prototype Eddy Current Inspection Tool (PECIT).

From the data above, blue region represents the subsurface cracks and also brick layers. This confirms the ability of this technology to detect subsurface cracks.

3            Materials and Methods

To improve the existing method for measuring the deformation of the graphite core and detecting surface cracks, a light detecting and ranging system (LiDAR) was used for this research.

3.1          Introduction to LiDAR

In 1960, Theodore H. Maiman at Hughes Research Laboratories build the first ever laser based on theoretical work by Charles Hard Townes and Arthur Leonard Schawlow. The primary feature of a laser beam that made it very usable in this field compare to the other light sources is that it produces a highly-collimated light, meaning that the light will spread minimally as it propagates. This tight spot of light enable it to be used for multiple applications, i.e. laser cutting and LiDAR.

In 1970s, four classes of laser with a few subclasses have been introduced. This was to prevent any unwanted injuries from occurring as high-power lasers are hazardous due to its ability to cause permanent eye injuries and even burn the skin. Below is a table showing the description of each class and its subclasses:

Class	Comments
1	Intrinsically safe in foreseeable conditions
1M	Intrinsically safe, from the visible to near IR, in foreseeable conditions, unless when using beam visualisation optics.
2	No danger for people with a normal vision system (blink reflex).
2M	Intrinsically safe in the visible under foreseeable confitions, unless when using beam visualisation optics.
3R	Possible hazards for the direct vision of the beam.
3B	Hazards for the direct, but not for the diffused vision.
4	Hazards for the direct and diffused vision, in addition to fire hazard.

Figure 3.1: Class of lasers and its descriptions.

A LiDAR system was first invented in 1960s shortly after the discovery of laser beam. It works by calculating the time it takes for a small packet of laser beam to travel from the transmitter to a surface and bounces back to the receiver. Below is a figure demonstrating how a LiDAR system works and the required formula to calculate the surface distance.

Figure 3.2: The concept of a Light Detecting and Ranging (LiDAR) system.

Distance=Speed of light × Time of flight2

(2)

LiDAR was first used by the National Centre for Atmospheric Research to measure clouds. Few years after, in 1971, Apollo 15 mission uses LiDAR system to map the surface of the moon which in turn makes the public aware of its accuracy and usefulness.

Nowadays, LiDAR have been used in all sort of industry, from terrestrial mapping to autonomous vehicle, and even in robots. This is due to the fact that most of the LiDAR systems, if not all, uses a Class 1 Laser, meaning that it is safe to be use by the public and the fact that it is fairly cheap to buy. The first step that was done in order to carry out this project was to choose a suitable LiDAR system. In choosing which LiDAR system to be used, three LiDAR systems have been chosen to compare its specifications and price to suite the project’s need.

1. LIDAR-Lite 3 Laser Rangefinder with HS-422 Servo Motor.
   A one-dimensional LiDAR system which has an accuracy of +/-2.5cm with a range of zero to 40m. For it to work for this project, it requires a servo-motor to be attached beneath it which would allow it to get a 360° reading. This is the cheapest LiDAR which only cost £154 ^[20].
2. RPLIDAR A2 360° Laser Scanner.
   A 360° LiDAR system which has a built-in servo motor. It has an angular resolution of 0.9° with a distance range of 0.15m to 6m. This device takes 4,000 samples every second and it cost £430 ^[22].
3. Velodyne VLP-16 LiDAR
   A high-end LiDAR system which has 360° horizontal field of view and 30° vertical field of view. It has an accuracy of +/-3cm with angular resolution of 0.1°. This device takes 300,000 samples every second and it cost £6400 ^[21].

A weighted scoring method has been setup to aid the choosing process. Few main criteria were selected to match the purpose of this project. Range was scored based on the lidar’s ability to scan distance of 135mm from the transmitter, thus the scoring is either 100 or 0. Table below shows the result of the weighted scoring method.

		Requirement Score
Criteria	Weight	1	2	3
Horizontal FoV	10	20	100	100
Range	20	100	100	100
Samples per second	10	20	40	60
Distance Resolution	30	50	35	20
Angular Resolution	20	0	50	70
Dimension	5	80	60	20
Cost	5	100	80	0
Weighted Scores	100%	48	61.5	57

Figure 3.3: Result for the weighted scoring method.

As seen in table above, RPLiDAR A2 has the highest overall score with Velodyne VLP-16 positioned second. Although VLP-16’s specifications are better than RPLiDAR A2, the price at which it is selling currently was too expensive for this project’s budget. Thus, RPLiDAR A2 has been chosen for this project.

3.2          RPLiDAR A2 360° Laser Scanner

Setting up RPLiDAR A2 was an easy process, as it already comes with its USB adapter and a mini USB to USB 3 cable.

Figure 3.4: Parts received with the RPLiDAR A2.

The three components were simply connected together and the USB 3 connection can be connected to the computer’s USB slot. The driver was also given in their official site together with three applications developed by SLAMTEC for ease of use. The application that of interest was called “ultra_simple”, where it will run the lidar system and output the results obtained to a window. Figure below shows the result obtained by using the “ultra_simple” application:

Figure 3.5: Output result from the ‘ultra_simple’ application.

The variables ‘theta’ represents the rotation angle, from 0° to 360°, which increases as the lidar rotates clockwise, while the variable ‘Dist’ represent the distance of the surface from the centre of the RPLiDAR A2.

Figure 3.6: Representation of the variable ‘theta’ and ‘Dist’.

The code for this application was modified by using Visual Studio 2015 to output x and y coordinates instead of ‘theta’ and ‘Dist’, refer Appendix. By using Pythagorean theorem, the value of x and y coordinate of a single point can be obtained from its ‘theta’ and ‘Dist’ values.

x = Dist * cos(theta)

(3)

y = Dist * sin(theta) * -1

(4)

The result of integrating these two equations into the original code can be seen in figure below, each line represents a single point cloud data.

Figure 3.7: Output result from the modified version of ‘ultra_simple’ application.

The data shown in the window will be automatically exported into a text file in a comma separated variable (CSV) format. This allows the data to be examined by multiple software without the needs of changing the format every time.

Numbers in the black, and red box represents the x, and y-coordinates respectively.

Figure 3.8: Results saved in a text file using CSV format.

This allows the RPLiDAR to obtain a 2D scan of the inside of the graphite core. To transform the numbers obtained into a figure, two software were chosen, MATLAB R2016b and AutoCAD 2017. Both software has the function which enable user to input data from a text file which makes them suitable for this process.

A few lines of code were written to enable MATLAB to read the text file and plot each point to a new window. This was done by using the command line ‘csvread’ which reads the text file that was made in the previous section of this report. The two values in each line of data were separated into two arrays, one for the x-axis and one for the y-axis. This creates two 1×429 arrays each representing x and y-axis respectively. From this, a ‘scatter’ command was used to plot each data point as a single filled dot inside the new window.

For AutoCAD 2017, the same .txt file was modified by adding an additional ’POINT’ command at every line of the .txt file. This file will then need to be saved with a .scr extension to create an AutoCAD Script file which allow the use of a ‘SCRIPT’ command inside the AutoCAD software which will open the .scr file and plot it automatically.

For a 3D scan of the graphite core, the RPLiDAR was attached to a ET50 Screw Drive Electric Actuator which will move the RPLiDAR through the entire length of the graphite core starting from the bottom all the way to the top. The code for the application previously used was altered by adding another line which corresponds the z-axis. The new application allows the RPLiDAR to take a full 360° scan before adding a value of one to the z coordinate and the process repeats. The process will be done automatically while the RPLiDAR is being move upward.

Figure 3.9: ET50 Screw Drive Electric Actuator.

Processing the 3D scan data uses the same method as the 2D scan data. The only difference is that only AutoCAD 2017 will be use. This is due to MATLAB having problem its internal memory, thus reading thousands of data in a single file is an impossible task. To create 3D points inside AutoCAD, the same process can be done as the 2D points. All of the results obtained will be measured for the diameter, as the nuclear graphite core that was used is a start of life graphite core meaning that it is a perfect cylinder with diameter of 135mm.

The final test was to used a 1mm wire placed inside the graphite core to test the reliability of the LiDAR system in detecting cracks. This test requires the same application and process as to the 3D scan.

4            Results

4.1          2-Dimensional Point Cloud Data

4.1.1. MATLAB R2016b

Figure 4.1: Result obtained by using MATLAB R2016b.

4.1.2. AutoCAD 2017

Figure 4.2: Result obtained by using AutoCAD 2017.

4.1.3. 1mm Wire Test

Figure 4.3: Result obtained when a wire is placed onto a wall.

4.2          3-Dimensional Point Cloud Data

Figure 4.4: Top (a) and side (b) view of the point cloud data obtained using the RPLiDAR A2.

Figure 4.5: Presence of crack inside the point cloud data.

5            Discussion

5.1          Experiments using RPLiDAR A2

From the 2-dimensional scan result, it is obvious that although the overall point plot can be seen as a circle, as it should be, but there are points that are situated further or nearer to the centre point. This was caused by the LiDAR system itself as the sensitivity was not high enough. Thus resulting in a less accurate plot of the graphite core.

The presence of crack in figure 4.1 and 4.2 are not very obvious, but a closer look shows that there are some missing points at the top part of the circle, proving that there is a crack in the graphite core. Due to this observation, an assumption was made that a laser beam that goes through a crack does not return to the recieving-end of the LiDAR system resulting in the LiDAR system unable to measure any distance, thus, no point can be plotted. This assumption can be use as a rough estimate to detect any presence of crack and locate its position for further analysis using other forementioned technique such as the CVC device.

As for the 1mm wire experiment, the LiDAR system could not detect any presence of the wire due to its angular and horizontal resolution were not high enough to detect small changes in displacement. This is a different scenario to that of a surface crack, as wires can still reflect the laser beam back to the recieving-end of the LiDAR system. To eliminate these problems, a higher sensitivity LiDAR system can be use as a replacement to the RPLiDAR A2.

Since the 2-dimensional scan result shows a high possibility of using this system to measure the deformation of the graphite core and its ability to detect large cracks, the next step was to attached the RPLiDAR A2 to a ET50 Screw Drive Electric Actuator for a 3-dimensional scan. The movement of the actuator brings the RPLiDAR A2 from the bottom to the top of the core, and the output result can be seen in figure 4.4 and 4.5. The whole process of setting up, gathering informations and post processing for a single core takes no longer than 10 minutes. Comparing this to that of the CBMU device shows a substanstial decrease in work time. Thus, implementing the LiDAR system as one of the inspection technique means that all of the graphite cores can be scanned instead of just a few. This will produce a high volume and high detail data of the health of the plant in the same amount of time that it takes for the CBMU device to scan just a few number of core.

From the data gathered and processed, the inside of the core can be carefully inspected by using AutoCAD 2017. The top and side view of the core can be easily obtained by rotating the plot making inspection and comparison for the deformation easy. This means that all of the core can be scanned at the start of its life and saved for future reference for accurate deformation analysis without the need of using ’Channel Selection Assistant’.

As seen in figure 4.5, there is a big crack that runs from the top of the core straight down to the bottom of the core, this is represent in figure 4.5 by a straight line of missing points, labelled crack, confirming the usability of the LiDAR system to detect surface crack in the graphite core. Although to be noted that due to the angular and horizontal resolution of the RPLiDAR A2, fine cracks cannot be measured. As previously mentioned, this problem can be easily solved by using a much sensitive LiDAR system with smaller angular and horizontal resolution.

5.2            Future Research

In this subsection, there will be two sub-section which will discuss further on research that can be done to improve the method of analysing the deformation of nuclear graphite core. The first sub-section will be more on how to improve the current method used, in this case, the LiDAR system, and the other sub-section will discuss on the newer method that will be more reliable in the next few years.

5.2.1. Improving the LiDAR System

Since the RPLiDAR A2 and the ET50 Screw Drive Electric Actuator are two different devices, the data obtained may have some degree of inaccuracy since the displacement of the actuator may not be the same to that registered by the ’ultra_simple’ application. A device which implement both, the LiDAR system and the actuation system, into one single controller would be the best way to obtained accurate data in terms of its vertical displacement.

As for the LiDAR system itself, a higher accuracy LiDAR system with a greater horizontal and angular resolution would be ideal for this application. This would not only gives a high volume and high detail data but the use of a CBMU device can be eliminated meaning that analysing graphite cores would not consume a lot of time and money.

Note that, the RPLiDAR A2 that was used in this project is not suitable to be put in the actual radioactive plant as the body that encases it was not made to be put in a radioactive and high temperature environment. Thus, the new LiDAR system’s body must be made from a material that has tolerance with the radiation present, for example Polyether Ether Ketone (PEEK), also materials that can withstand high temperature without melting.

5.2.2. New Technology

Figure 5.1: The Confocal Chromatic sensor.

Confocal Chromatic Sensors is a type of sensor that uses a very simple light phenomenon, which is refraction. This is arguably the most accurate and most resolute distance sensor technique. It works by passing through white light (polychromatic) over a series of lens to which spreads the focal length over several points, creating a full spectrum of colours (monochromatic) as shown in figure above. Different wavelengths of different coloured lights are focused at different distances away from the sensor head. This allows for an easy measurement of displacements as surfaces will reflect only a specific wavelength off of it surface. A light sensor is then used to measure the reflected wavelength and calculate the displacement relative to the sensor head ^[17]. Figure below shows the actual result obtained by using this sensor to measure the surface of a silicon chip ^[18].

The advantage of using this device is that, it has no moving parts and no electronic components inside the sensor head, meaning that noise and other type of errors are likely to be avoided as no heat or vibrations is produced inside the sensor head. This will in turn produce an accurate result without having to take multiple readings. The white light source (white LED) is situated inside the sensor unit which will be passed to the sensor head via an optical fibre cable. This allows different sensor head with different measuring capabilities to be connected to the same sensor unit for different application purposes. The measuring distances ranges from 30µm to 30mm with resolution of up to 0.024µm ^[16]. There are also a 90° version that allows for measurement of inner diameter, this increases the likelihood of this device to be used as a measuring device for the deformation of a reactor core.

Just like any other new technology, this device takes time to fully developed to be able to be used in other industries. As of now, this technology hasn’t been able to detect distances further than 30mm, which makes them unsuitable to be used inside the reactor core that has inner diameter of 270mm. Other than that, this device is absolutely an ideal device to be used inside the reactor core as no heat or radiation can ruined with the sensor head, provided it was made by using materials that can withstand radiations.

6            Conclusion

From this experiment, it shows that a LiDAR system is a reliable technique when it comes to measuring the deformation inside AGR’s graphite cores. The technique requires only a short period of time to be set up and processing the obtained data takes only a couple of steps. The overall time it takes is roughly around 10 minutes, depending on the power of the computer used to process the data. But, after processing the data, the 3-dimensional point cloud could be easily examined for any deformation or sizable cracks. The process of comparing the new data from the old data is also easy as overlapping point cloud can be made easily using AutoCAD software.

As of now, the data obtained could not be made into a 3-dimensional solid, but with the knowledge of others and CAD softwares, a 3-dimensional solid of the inside of the graphite core could be made by using the point cloud data obtained using this system.

Apart from that, the downside to this system is that it would not able to detect any fine surface cracks and subsurface cracks. Other than that, this system would be suitable to replace the CBMU device, provided that the LiDAR system has been made to withstand the radioactive and high temperature environment, also having a reasonable accuracy and sensitivity.

Bibliography

1. (2017) Nuclear Power in the United Kingdom, Available at: http://www.world-nuclear.org/information-library/country-profiles/countries-t-z/united-kingdom.aspx (Accessed: 28 October 2016).
2. (2016) Nuclear Development in the United Kingdom, Available at: http://www.world-nuclear.org/information-library/country-profiles/countries-t-z/appendices/nuclear-development-in-the-united-kingdom.aspx (Accessed: 30 October 2016).
3. D.K.L. Tsang and B.J. Marsden (2006) ‘The development of a stress analysis code for nuclear graphite components in gas-cooled reactors’, Journal of Nuclear Materials, 350(3), pp. 208-220.
4. Stanislav Golubov, Roger Stoller, Alexander Barashev, and Bachu Singh (n.d.) Modeling Irradiation Creep and Growth with VPSC.
5. J.V. Best, W.J. Stephen, and A.J. Wickham (1985) ‘Radiolytic graphite oxidation’, Progress in Nuclear Energy, 16(2), pp. 127-178.
6. M. Wadsworth, S.T. Kyaw, and W. Sun (2014) ‘Finite element modelling of the effect of temperature and neutron dose on the fracture behaviour of nuclear reactor graphite bricks’, Nuclear Engineering and Design, 280(), pp. 1-7.
7. A.D. Fletcher, A.J. Peyton, D. Armitage, M. Brown, and R. Chapman (2014) ‘Subsurface cracks in AGR graphite bricks’, NDT & E International, 66(), pp. 72-81.
8. B. T. Kelly (1984) The Radiolytic Corrosion of Advaced Gas-Cooled Reactor Graphite, UKAEA Springfields Nuclear Power Development Laboratories, Salwick, Preston, Lancashire, U.K..
9. P. G. Tipping (2010) Understanding and Mitigating Ageing in Nuclear Power Plants.
10. B.J. Marsden, G.N. Hall (2012) ‘4.11 – Graphite in Gas-Cooled Reactors’, Comprehensive Nuclear Materials, 4(), pp. 325-390.
11. (1999) The First Information Exchange Meeting on Survey on Basic Studies in the Field og High Temperature Engineering.
12. D.K.L. Tsang and B.J. Marsden (2006) ‘ The development of a stress analysis code for nucleargraphite components in gas-cooled reactors D.K.L. Tsang * , B.J. Marsden’, Journal of Nuclear Materials, 350(), pp. 208-220.
13. A. T. Ellis and K. M. Staples (n.d.) Management of Ageing Processes in Graphite Reactor Cores.
14. Reactor Core Inspections – Channel Bore Measurement Unit (CBMU), Available at: https://nucleus.iaea.org/sites/graphiteknowledgebase/wiki/Guide_to_Graphite/Reactor%20Core%20Inspections%20_%20Channel%20Bore%20Measurement%20Unit%20(CBMU).aspx (Accessed: 5 January 2017).
15. G. M. West , S. D. J. McArthur , J. R. McDonald , A. Ballantyne , J. Reed , and S. Beynon (n.d.) Graphite core brick crack detection through automated load trace analysis.
16. Confocal chromatic sensors, Available at: http://www.micro-epsilon.com/displacement-position-sensors/confocal-sensor/confocal-chromatic-sensors/ (Accessed: 23 January 2017).
17. Principles of Measurement Used by Laser Sensors, Available at: http://www.acuitylaser.com/measurement-principles/ (Accessed: 23 January 2017).
18. Confocal Chromatic Measurement System, Available at: http://www.micro-epsilon.co.uk/download/products/cat–confocalDT–en.pdf (Accessed: 23 January 2017).
19. LIDAR-Lite 3 Laser Rangefinder, Available at: http://www.robotshop.com/uk/lidar-lite-3-laser-rangefinder.html#Specifications (Accessed: 4 February 2017).
20. PUCK™ VLP-16, Available at: http://velodynelidar.com/vlp-16.html (Accessed: 6 February 2017).
21. TRIUMPH Camera Inspections, Available at: https://nucleus.iaea.org/sites/graphiteknowledgebase/wiki/Guide_to_Graphite/TRIUMPH%20Camera%20Inspections.aspx (Accessed: 17 February 2017).
22. (2017) RPLIDAR A2 The Thinnest LIDAR, Available at: https://www.slamtec.com/en/Lidar (Accessed: 19 October 2016).