Integration of an Ultrasonic Sensor within a Robotic End Effector for Application within Railway Track Flaw Detection

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Defectoscopy plays a very important role in the process of detecting rail cracks. Maintaining the railway surface at the appropriate level requires systematic testing. In the article, the authors presented an interesting concept of testing railway rails using the classic ultrasonic research technique and type A imaging. They also presented the observed limitations in the research method.

 The purpose of rail testing is to detect defects occurring in the entire cross-section of the rail. Testing rails mounted on the track involves introducing ultrasonic waves from the head surface through the coupling liquid layer. Access from the foot surface is not possible and therefore developing operational cracks in the rail foot cannot be detected.

 

During the research process, the authors focused their research only but consciously on two elements, the so-called head (on its surface) and web (the element connecting the head with the base "foot" of the rail). However, in many cases, defects also occur in the rail foot (the lower part attached to the track). Such tests, i.e. in the rail foot, are difficult to carry out for rails embedded (fastened) on the track.

In the case of rail testing, automatic testing is an important issue as it speeds up the research process. Refining the research methodology is essential when performing tests on individual rail elements.

 

Comments and suggestions for authors:

1) The angle used in conventional tests is also the angle of 45 degrees, used e.g. in the Tandem technique for testing rails. When continuing the research, the authors may consider introducing this type of heads with this angle of introduction of the ultrasonic beam into the research methodology. Research using multi-transducer heads and PA imaging should also be considered.

It is important for the authors to indicate that they have knowledge on this subject but have consciously limited their research.

2) The authors made a significant contribution to developing the research process and drawing conclusions that will allow the continuation of the research methodology.

I ask the authors to indicate future research directions.

Comments on the Quality of English Language

-

Author Response

Comments and suggestions for authors:

1) The angle used in conventional tests is also the angle of 45 degrees, used e.g. in the Tandem technique for testing rails. When continuing the research, the authors may consider introducing this type of heads with this angle of introduction of the ultrasonic beam into the research methodology. Research using multi-transducer heads and PA imaging should also be considered. It is important for the authors to indicate that they have knowledge on this subject but have consciously limited their research.

 

# Reviewer 1 Response 1. This is a good observation. The research was very focused on the hypothesis that automation is viable. Now that we have shown this we hope to progress the research with a broader scope. As part of this the potential for an end effector with multi-transducer heads is certainly within the area of interest and we have a track record of designing novel end effectors so this is a direction we will pursue.

 

2) The authors made a significant contribution to developing the research process and drawing conclusions that will allow the continuation of the research methodology.

I ask the authors to indicate future research directions.

 

# Reviewer 1 Response 2. We have now added the following future research to the conclusion. ‘This research was focused on the hypothesis that automation is viable for detecting flaws on rail tracks. Future research should now focus on the design of end effectors that can accommodate multi-transducers for increased ultrasonic testing. In addition, an end effector design that can accommodate a Phased array (PA) transducer should be studied. Compared with single element systems the PA uses a total focusing method (TFM) that provides full matrix capture (FMC). This would make the mapping of tracks much easier and more accurate than with traditional ultrasonic equipment.’

Reviewer 2 Report

Comments and Suggestions for Authors

This paper combined a robotic end effector with ultrasonic sensors to demonstrate its flaw detection capability as an initiation to automate the flaw-inspection in railway track. I found the paper to be overall well-written and well-structured. Some clarification could be used to help readers better appreciate this work.

1. Page 5, Figure 3. More details are needed to describe the figure. Did the robot just place the sensor on top of the specimen? Or did it also perform data acquisition?

2. Page 6, Figure 4. More descriptions are needed for the figures. It's quite challenging for me to understand the meaning of this figure.

3. Page 6-7, Figure 5 and figure 6 seem redundant because the same information in Figure 5 and 6 can be found in Figure 7.

Author Response

1. Page 5, Figure 3. More details are needed to describe the figure. Did the robot just place the sensor on top of the specimen? Or did it also perform data acquisition?

 

# Reviewer 2 Response 1. The robot is programmed to place the sensor on top of the specimen and then move in a manner that keeps it within the correct proximity for a full signal. The potential data acquisition and recording in real time exists but in the case of this research the signals were observed and recorded in person. To improve the clarity, we have added 3 new figures (fig 2, 3 and 5).

 

2. Page 6, Figure 4. More descriptions are needed for the figures. It's quite challenging for me to understand the meaning of this figure.

 

# Reviewer 2 Response 2. LC to do this.

LC reply : Figure 4 depicts examples of what to expect in regions where no flaws (cracks) where present and where cracks were present. Figure 7a shows that no signal present (no green spiked peak) hence meaning that in that specific region there are no cracks. Figure 7b and 7c both show a signal (green distinct peak) when using the probe manually and when using the robotic configuration therefore proving that cracks can both be detected manually and automatically. We have now added this description to the text in 4.2.

 

3. Page 6-7, Figure 5 and figure 6 seem redundant because the same information in Figure 5 and 6 can be found in Figure 7.

 

# Reviewer 2 Response 3.  

We agree both the original figure 5 and figure 6 are summed up in figure 8. So, we have removed figure 5 and 6.

Reviewer 3 Report

Comments and Suggestions for Authors

I would like to provide the following comments on the manuscript.

This study focuses on intergrating the novel intergration of an ultrasonic detection technology Epoch™ 650 within the end effector of a KUKAKR 16 robotic system for enhancing the detection of railway track flaws. The research results are innovative and have the potential to enhance the understanding of scholars in the same field.

Additionally, the reviewer has suggested several necessary amendments, such as:

1) The sensor arrangement described in this paper, with one sensor positioned atop the rail and another at the rail waist, is innovative. The efficacy of the waist-mounted sensor in detecting the location and depth of penetrating injuries, as demonstrated in the article, is noteworthy. However, this study appears to overlook the prevalent types of rail defects encountered in real-world scenarios, particularly those arising from rolling contact fatigue. Such fatigue-induced damages predominantly manifest below the rail's top surface and around the rail head. This raises a question about the capability of waist-level sensors to detect these fatigue cracks. It may be worth exploring the potential of positioning sensors at the track gauge angle to address this concern. Furthermore, the paper would benefit from empirical data or case studies demonstrating the effectiveness of this robotic and ultrasonic probe-based flaw detection system in actual rail service conditions. Incorporating such data would significantly strengthen the practical applicability and validity of the proposed system.

2) In Section 3, the paper refers to 60° and 70° wedged probes, yet the structural distinctions between these two types of probes are not clearly delineated. For enhanced clarity and a better understanding of their respective functionalities and applications, it would be beneficial if the paper could include a comparative illustration. Visual representations, such as diagrams or schematics, would greatly aid in comprehending the physical differences and potential impact these variations have on the probes' performance. Such additions would not only enrich the section but also provide the reader with a more comprehensive understanding of the technology being discussed.

3) The assertion in Section 4.2 regarding the accuracy of the 70° wedged probe in identifying all hole sizes, as referenced in Figure 6, warrants reconsideration. A closer examination of the data reveals that the diameter of the 6mm hole, as measured by the Robotic 70° probe, is recorded as 6.557mm. This measurement significantly deviates from the actual size and demonstrates a lower accuracy compared to the 5.847mm measurement obtained manually with the same probe. This discrepancy raises questions about the claimed precision of the robotic probe, and it would be beneficial for the authors to address this inconsistency. A more detailed analysis or discussion regarding the factors contributing to this deviation would enhance the reliability and comprehensiveness of the findings.

4) The location and depth of the defect in Figure 1 are not clearly shown in the picture in Section 3. Please mark it in Figure 1 so that the true location of the defect is clearly visible.

Thank you for considering these comments. I hope they are helpful for the improvement of the manuscript.

Author Response

1) The sensor arrangement described in this paper, with one sensor positioned atop the rail and another at the rail waist, is innovative. The efficacy of the waist-mounted sensor in detecting the location and depth of penetrating injuries, as demonstrated in the article, is noteworthy. However, this study appears to overlook the prevalent types of rail defects encountered in real-world scenarios, particularly those arising from rolling contact fatigue. Such fatigue-induced damages predominantly manifest below the rail's top surface and around the rail head. This raises a question about the capability of waist-level sensors to detect these fatigue cracks. It may be worth exploring the potential of positioning sensors at the track gauge angle to address this concern. Furthermore, the paper would benefit from empirical data or case studies demonstrating the effectiveness of this robotic and ultrasonic probe-based flaw detection system in actual rail service conditions. Incorporating such data would significantly strengthen the practical applicability and validity of the proposed system.

 

#Reviewer 3 Response 1. These are good observations. We are aware that there are different rail defects encountered in real-world scenarios and certainly contact fatigue is one of major concern. The research was very focused on the hypothesis that automation is viable. Now that we have shown this we hope to progress the research. We plan to invest in a phased array sensor, this will widen the scope of inspection capabilities and help identify additional flaws, and as pointed out by another reviewer we hope to integrate a multi sensor end effector so that detection is increased. As mentioned the main focus of this research was to prove that we could successfully detect flaws using a rudimentary automation firstly before expanding to more realistic rail service conditions.

 

2) In Section 3, the paper refers to 60° and 70° wedged probes, yet the structural distinctions between these two types of probes are not clearly delineated. For enhanced clarity and a better understanding of their respective functionalities and applications, it would be beneficial if the paper could include a comparative illustration. Visual representations, such as diagrams or schematics, would greatly aid in comprehending the physical differences and potential impact these variations have on the probes' performance. Such additions would not only enrich the section but also provide the reader with a more comprehensive understanding of the technology being discussed.

 

#Reviewer 3 Response 2. LC to do this.

 

LC reply: essentially there are no structural distinctions between the probes. The only difference between the 2 is that the piezoelectric crystal within the probe is angled θ at 60 degrees for one and 70 degrees for the other. The affect that this has is that the incident angle of the wave that is produced from the probe is ultimately dependant on the angle that the piezo electric crystal is set at. Therefore, the wave is released at a 60 and 70 degree angles. During testing, the sound beam will normally travel down to the test piece's bottom at the created angle and then reflect back up at the same angle. The complete height of a flaw is covered by the sound beam as the probe is moved back and forth. The entire flaw volume may be inspected and discontinuities can be found due to this scanning action. The accuracy of the results of the fault detection is ultimately dependent based on the size of angle of probe chosen. In common practice a probe of 60 or 70 degree is used for thicknesses of up to 40 mm. Anything above 40mm, a probe of 45 degree or less is used. To emphasise this we have added a figure (figure 3) to the text and the following additional descriptive text to section 3. ‘The Epoch ™ 650 used in this study utilises a piezoelectric transducer with an integrated receiver to propagate ultrasonic energy into the specimen. The selected Angle Beam Wedges Transducers are single element transducers for flaw detection and sizing. The unit transmits a primary shear wave into a test piece and the design allows them to be easily scanned back and forth over the inspected part. Unlike straight beam testing where the sound beam will travel at the generated angle down to the bottom of the test piece and then reflect upward at the same angle. In the selected angle beam testing by moving the probe back and forth the sound beam will sweep across the full height of a detected flaw. This scanning motion enables inspection of the entire flaw either through direct reflection from a second acoustic path (Figure 3 a and b) or through second acoustic path (figure 3 c and d). These transducers are typically used for the inspection of welds, and cracks on pipes, tubes, forgings, castings and machined components. They are Atlas European Standard Transducers designed to meet inspection criteria referenced throughout Europe and feature standard connectors and common frequencies. The two sensors use a 70 and 60 Degree Refracted Shear Wave (frequency range between 4 MHz and 5 MHz) and have a near field distance in steel of 30 mm. Before inspection both sensors were tested using NDT Calibration and Reference Test Blocks. The block used were a U8880016 2214M 5-STEP 1018 steel block and a U8880046 TB1065-1 ISO 7963 miniature steel block (ISO 7963 Ultrasonic testing).’

 

3) The assertion in Section 4.2 regarding the accuracy of the 70° wedged probe in identifying all hole sizes, as referenced in Figure 6, warrants reconsideration. A closer examination of the data reveals that the diameter of the 6 mm hole, as measured by the Robotic 70° probe, is recorded as 6.557mm. This measurement significantly deviates from the actual size and demonstrates a lower accuracy compared to the 5.847mm measurement obtained manually with the same probe. This discrepancy raises questions about the claimed precision of the robotic probe, and it would be beneficial for the authors to address this inconsistency. A more detailed analysis or discussion regarding the factors contributing to this deviation would enhance the reliability and comprehensiveness of the findings.

 

#Reviewer 3 Response 3.

The main reason as to why the data seems inconsistent when measuring the diameter of the 6 mm hole using the 70 degree probe via the robotic configuration when compared to the data gathered when measuring this manually could be due to the rate of speed at which the probe passed over the flaw. When performing this manually the human will invariably change the inspection speed when a flaw is detected therefore adding accuracy to the flaw sizing. In the automation path programming we kept it consistent therefore not allowing room for variation of speed whenever a flaw is detected. In the future we could consider a real-time subroutine that allows for a reduction in inspection speed for accurate sizing of the flaw. This level of intelligent automation is possible but was not the focus of this research.

 

4) The location and depth of the defect in Figure 1 are not clearly shown in the picture in Section 3. Please mark it in Figure 1 so that the true location of the defect is clearly visible.

 

#Reviewer 3 Response 4. We have added another figure Fig 2) to make this clearer for the reader.

Reviewer 4 Report

Comments and Suggestions for Authors

The paper proposes that they investigates the novel integration of an ultrasonic sensor within a robotic platform specifically for the application of surface crack and internal defects within rail tracks. The performance of the robotic sensor system has been assessed on a rail track specimen containing sacrificial surface cracks and internal defects and then compared against a manual detection system. The investigation concludes that the robotic sensor system successfully identified internal defects in the web region of the rail track when utilising a 60° and 70° wedged probe, with a frequency range between 4 MHz and 5 MHz. However, the surface crack investigation proved that the transducer was insensitive to the detection of cracks, possibly due to the inadequate angle of the wedged probe. The overall outcome of the study highlights the potential that robotic sensor systems have in the detection of internal defects and characterises the limitation of surface crack identification to assist in enhancing rail safety. The paper is good, and more details about the comments should be provided:

1、The wedge probe does not have enough testing angles and frequency, and additional experiments should be conducted.

2、The distance between the surface crack detection area and the internal defect detection area in the test sample is not large, which will affect the measurement results. Additional experiments should be conducted to eliminate interference.

3、The surface crack detection results described in the paper were not directly provided, and experimental results should be supplemented.

4、The performance parameters of the sensors used in the paper are not directly provided, please supplement them.

5、There are not enough experimental samples for ultrasonic crack detection, and additional experiments such as different shapes and depths of track crack holes are needed.

Author Response

1. The wedge probe does not have enough testing angles and frequency, and additional experiments should be conducted.

 

#Reviewer 4 Response 1. In this research we showed that the 60° and 70° wedged probes, with a frequency range between 4 MHz and 5 MHz were able to detect track flaws. In order to establish this, we performed many hours of inspection and this included the use of a 45° sensor.  These pre-trials established the detection efficacy of the 60° and 70° wedged probes and 4 MHz to 5 MHz. Additional experiments would provide more information but we believed that we had evidenced the main goal which was the proof that this normally human inspection task could be done using a robot. With this achieved we now hope to peruse further experimental work to build on our findings,

 

2. 2. The distance between the surface crack detection area and the internal defect detection area in the test sample is not large, which will affect the measurement results. Additional experiments should be conducted to eliminate interference.

 

#Reviewer 4 Response 2. The observation about the proximity of the surface crack detection area and the internal defect detection area, potentially affecting the measurement results, is valid. To mitigate interference and enhance the accuracy of measurements, additional experiments and adjustments can be considered. Most importantly we will require a more complex sensor (Phased array). This is certainly our future plan future and we have added the following future research to the conclusion section. ‘This research was focused on the hypothesis that automation is viable for detecting flaws on rail tracks. Future research should now focus on the design of end effectors that can accommodate multi-transducers for increased ultrasonic testing. In addition, an end effector design that can accommodate a Phased array (PA) transducer should be studied. Compared with single element systems the PA uses a total focusing method (TFM) that provides full matrix capture (FMC). This would make the mapping of tracks much easier and more accurate than with traditional ultrasonic equipment.’

 

3、The surface crack detection results described in the paper were not directly provided, and experimental results should be supplemented.

 

#Reviewer 4 Response 3. LC do you have this data? “Initial results generated for the surface crack study demonstrated that the transducer was insensitive to the detection of surface cracks.”  No results can be provided because the transducer was insensitive. This is specifically substantiated & explained in the discussion.

 

4、The performance parameters of the sensors used in the paper are not directly provided, please supplement them.

 

#Reviewer 4 Response 4. The following has been added to section 3. . ‘The Epoch ™ 650 used in this study utilises a piezoelectric transducer with an integrated receiver to propagate ultrasonic energy into the specimen. The selected Angle Beam Wedges Transducers are single element transducers for flaw detection and sizing. The unit transmits a primary shear wave into a test piece and the design allows them to be easily scanned back and forth over the inspected part. Unlike straight beam testing where the sound beam will travel at the generated angle down to the bottom of the test piece and then reflect upward at the same angle. In the selected angle beam testing by moving the probe back and forth the sound beam will sweep across the full height of a detected flaw. This scanning motion enables inspection of the entire flaw either through direct reflection from a second acoustic path (Figure 3 a and b) or through second acoustic path (figure 3 c and d). These transducers are typically used for the inspection of welds, and cracks on pipes, tubes, forgings, castings and machined components. They are Atlas European Standard Transducers designed to meet inspection criteria referenced throughout Europe and feature standard connectors and common frequencies. The two sensors use a 70 and 60 Degree Refracted Shear Wave (frequency range between 4 MHz and 5 MHz) and have a near field distance in steel of 30 mm. Before inspection both sensors were tested using NDT Calibration and Reference Test Blocks. The block used were a U8880016 2214M 5-STEP 1018 steel block and a U8880046 TB1065-1 ISO 7963 miniature steel block (ISO 7963 Ultrasonic testing).’

 

5. There are not enough experimental samples for ultrasonic crack detection, and additional experiments such as different shapes and depths of track crack holes are needed.

 

#Reviewer 4 Response 5. This is a good observation. We are aware that there are different rail defects encountered in real-world scenarios. The research was very focused on the hypothesis that automation is viable. Now that we have shown this we hope to progress the research. As mentioned to another reviewer we plan to invest in a phased array sensor, this will widen the scope of inspection capabilities and help identify additional flaws, and as pointed out by another reviewer we hope to integrate a multi sensor end effector so that detection is increased. As mentioned the main focus of this research was to prove that we could successfully detect flaws using a rudimentary automation firstly before expanding to more realistic rail service conditions.

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

I have no more suggestions.

Reviewer 4 Report

Comments and Suggestions for Authors

The manuscript could be accepted for publication as the present form.