- Email: [email protected]
Thermographic online monitoring system for Automated Fiber Placement processes
Accepted Manuscript Thermographic Online Monitoring System for Automated Fiber Placement Processes Berend Denkena, Carsten Schmidt, Klaas Völtzer, Tristan Hocke PII:
S1359-8368(16)30528-5
DOI:
10.1016/j.compositesb.2016.04.076
Reference:
JCOMB 4275
To appear in:
Composites Part B
Received Date: 17 March 2016 Accepted Date: 24 April 2016
Please cite this article as: Denkena B, Schmidt C, Völtzer K, Hocke T, Thermographic Online Monitoring System for Automated Fiber Placement Processes, Composites Part B (2016), doi: 10.1016/ j.compositesb.2016.04.076. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Thermographic Online Monitoring System for Automated Fiber Placement Processes BEREND DENKENA, CARSTEN SCHMIDT, KLAAS VÖLTZER*, TRISTAN HOCKE Institute of Production Engineering and Machine Tools, Leibniz Universität Hannover, Ottenbecker Damm 12, 21684 Stade, Germany
RI PT
*Corresponding author. E-Mail: [email protected]
M AN U
SC
Abstract: Automated Fiber Placement processes are commonly used to manufacture lightweight structures e. g. for highly demanding aerospace applications. In general, quality inspection is usually carried out manually and it is considerably time consuming in terms of high lot sizes and huge part dimensions. An online AFP process monitoring based on thermal camera combined with process depending image processing is presented. The visible temperature difference between the laid-up tow and its surface underneath is analyzed and information of lay-up defects such as overlaps, gaps, twisted tows and bridging derived. This monitoring system can reduce the efforts in quality inspection and will help to increase process reliability significantly. Keywords: A. Prepreg, D. Thermal Analysis, D. Process Monitoring, E. Lay-Up
1 Introduction
EP
Positioning defects (gaps, overlaps, missing tows, twisted tows) Bonding defects (bridging, air pockets) Foreign bodies Tow defects
AC C
TE D
The Automated Fiber Placement (AFP) process evolves into one of the leading manufacturing processes in lightweight industries, especially in aerospace applications. This process is able to reach high productivity by guaranteeing continuous high quality and economic manufacturing. Within this process, several pre-impregnated carbon fiber slit-tapes are laid-up by a Fiber Placement head on a tooling surface side-by-side simultaneously. During the lay-up process, certain defects may occur as schematically shown in Fig. 1. They can be divided in four categories:
[FIGURE 1 HERE]
Today, the operator does the quality control during AFP processes manually by visual inspection. To do so the lay-up process has to be interrupted. The visual inspection and the repair process in cases of detected defects are done for every single ply during the AFP process and therefor are very time consuming. In literature, the effort is numeralized from 32 % [1] to 63 % [2] of the overall AFP production time. Hence, the time of non-productivity caused by visual inspection and defects provoke high costs. If defects are not detected during the AFP process they need to be detected and repaired after curing the part which is even more expensive. Therefore, the reliability and the time of the inspection process are intended to be optimized. 1
ACCEPTED MANUSCRIPT
RI PT
A camera system mounted on the Fiber Placement head can be a support for the operator [3]. By this, he is able to do a continuous inspection of currently laid-up slit-tapes. However, this method demands a high degree of concentration, thus it is not practical for an online process monitoring. Furthermore, the black prepreg slit-tapes prevent a high visual contrast in between the single tows and the plies on the tooling surface. Consequential an intensive illumination and a complex image analysis are necessary, thus a visual monitoring is very ambitious to realize [4]. However, Miene [5] presents an approach wherein an algorithm calculates the fiber orientation for single rovings during the manufacturing process of carbon fiber reinforced plastics (CFRP) preforms and reasons occurring defects. The reliability depends highly on the ambient conditions and the required computing time for the image analysis is significantly high.
SC
The quality control for fiber-reinforced components via ultrasonic probes is prevalent [6]. This method is conducted after the manufacturing process. Until now, there is no case known where ultrasonic probes are used for online monitoring of the placed tows.
M AN U
In the context of research activities laser triangulation sensors are used to monitor single tows [7]. The monitor system can localize the position of the tows by an edge detection.
TE D
Shadmehri et al. developed a Laser-Vision inspection system to detect the ply location, the fiber orientation and gaps [8]. Therefore, he combined a laser projector system with a vision system. Initially a laser is scanning the targets before projecting the characteristic properties on the surface. Afterwards the operator can inspect certain sections of the surface on a monitor. As [8] concludes there is no complete solution for an inspection of the fiber placement process available and in addition it is challenging to integrate the process monitoring in the process itself for an online inspection. The US patent 7513964 B2 [9] describes a thermal AFP monitoring system. During the Fiber Placement, heat or cold is applied to the tows and an IR-camera pictures the process. The composition of these pictures shows the whole ply and can be analyzed to detect certain defects in lay-up. There is no realization of this patent known so fare.
AC C
2 Description
EP
In the following, a new approach for a thermographic online process-monitoring system for analyzing the thermal effects during the automated lay-up process is presented.
2.1 Thermographic Process Monitoring Looking at the AFP process the thermoset prepreg tows need to be kept cool inside the Fiber Placement head to prevent fouling and to ensure good material quality during processing. A second mayor quality aspect is the tack between the tow and it’s subsurface. To ensure a good tack, the surface of the tooling or previous laid tows is heated right before the lay-up and the tow is compacted under specific, controlled pressure. The schematic setup is shown in Fig. 2.
[FIGURE 2 HERE]
Right behind the compaction roller, the surface temperature of the laid-up tow exhibits a lower temperature than the surrounding surfaces. An Infra-Red (IR) camera can detect this tem2
ACCEPTED MANUSCRIPT perature difference as shown in Fig. 3. Due to the position and the angle of the camera, the picture shows a 40 mm sharp area wherein certain Pixel rows are used to determine the edges of the laid-up tows. The IR camera that was used for the first experiments has a resolution of 320 x 256 Pixel, which corresponds to about 7
.
The goal is to develop a process monitoring system that is able to detect defects for lay-up
RI PT
speeds up to 1 . Therefore, the camera frame rate of 60 and the computing speed of
the developed algorithm limit the real-time capability of the monitoring system.
[FIGURE 3 HERE]
2.2 Tow Localization
TE D
M AN U
SC
By evaluating specific regions behind the roller and by applying an edge detection algorithm, the tow geometry and its position can be extracted and monitored. This information reveals many placing defects like overlaps, gaps, twisted tows and spliced tows. Furthermore, the bond influences heat transfer from the substrate to the tow surface. Therefore, the temperature of the placed tow can be used as a quality indicator of the bond, as well. The temperature of the heated surface behind the compaction roller depends on the heat absorption, the convection and the heat transfer into the tooling. A homogeneous temperature transition is expected. If there are inhomogeneous sections in the laid-up laminate, their temperature differs from the surrounding surface temperature. These hot or cold spots indicate anomalies such as bridging defects or foreign objects. Knowing the normal temperature picture of the actual process, hot and cold spots are detectable by dynamic thresholds depending on the actual process. Both the edge detection and the surface inspection algorithms are separately applicable.
AC C
EP
Tow localization is a major part of the process monitoring. To detect the tow position and geometry the algorithm analyzes the temperature profile of the laid-up tows in the Region of Interest (ROI), right behind the compaction roller, as pictured in Fig. 4. The area close to the compaction roller is most suitable to localize the tows since the temperature contrast is highest right behind the compaction roller. The ROI consists of a certain number of Pixel rows to minimize computing errors through the noise of the raw data.
[FIGURE 4 HERE]
The developed algorithm calculates the edges of the tows so that the position can be determined and compared to the planned position. Information about the tow position is stored in the robot coordinate system. As can be seen in Fig. 5, the temperature profile also includes further information about the transition between single tows.
[FIGURE 5 HERE]
3
ACCEPTED MANUSCRIPT Hence, the algorithm can estimate the tow width, the gap width between the simultaneous placed tows and check for overlap of placed tows.
RI PT
Therefore, the temperature in between the tows is analyzed in detail (Fig. 5). If the temperature is similar to the subsurface temperature, a gap is identified next to the tows. If in contrast the temperature between two tows is lower compared to the tow temperature, the criterion for an overlap is fulfilled. The lower temperature in this case results from the double amount of material placed in this region. For the calculation of the tow width, a small gap width is presupposed to detect the edges of the tow reliably. If the gap width increases a certain value and differs from planned tow path, a rejection criterion is fulfilled, which is identified by the monitoring system. The same appears to any overlap.
SC
2.3 Surface Inspection
M AN U
In addition to the tow localization, it is also possible to evaluate the surface temperatures. An anomaly of temperature can represent bonding and planar defects, for example bridging or foreign bodies.
[FIGURE 6 HERE]
EP
3 Results
TE D
For a detailed analysis, certain ROIs are defined as pictured in Fig. 6. The ROIs adapt the shape of the laid-up tows and mask them as well as the left and right area next to the tows. The temperature transition in a single ROI should be homogeneous. If there is a region of higher or lower temperature it can be detected by checking with a calculated threshold value. Combined with the area size of these temperature anomalies the surface inspection can determine defects.
AC C
To test the online process monitoring an infrared camera is integrated in an experimental Fiber Placement setup (Fig. 7).With this setup different process conditions are simulated like accurate and faulty lay-up to test and improve the developed algorithm.
[FIGURE 7 HERE]
Fig. 8 shows an exemplarily lay-up result with different defects. As a first event, a tow twist can be identified in the plot of temperature profiles over the time. The twisted tow induces an increased gap between two tows. Additionally, the left tow edge is leaving the tolerance band according to the path planning. This is also captured in the contour plot, which is supplied by the algorithm, as well. This deviation will result in a defect indication. In the top of Fig. 8 a second event occurs, which is here an unexpected tow ending. Due to the missing tow edge, this defect is also easily detectable. Compared to the planned tow path, this missing tow will lead to a second defect indication. 4
ACCEPTED MANUSCRIPT [FIGURE 8 HERE]
RI PT
In addition, the surface inspection monitors local temperature anomalies of the tows and the surrounding surface. By comparing the temperature inside the ROIs to a calculated threshold, foreign bodies and adhesion defects get visible. The developed algorithm detects this defects by analyzing the temperature anomalies and interpreting of hot and cold spots.
SC
Fig. 9 illustrates this surface inspection procedure for a bridging tow over a gap. Due to the fact that there is no connection between the tows and it´s subsurface while the tows are bridging, the heat conduction substantially decreases. Caused by the decreased heat transfer the tows remain cold. Therefore, the colder areas of the tows are detected as cold spots (Fig. 10).
M AN U
[FIGURE 9 HERE]
[FIGURE 10 HERE]
TE D
The detection of foreign bodies during the Fiber Placement process is highly significant. Figure 11 shows an example for the lay-up over a carrier foil on a tooling surface. Due to the minor foil thickness compared to the tows no bonding defects (bridging) between the tows and the tooling surface occur. Moreover, the thermal properties of the foil differ substantially from the CFRP and the tooling material. Hence, the surface inspection in this area detects the extreme temperature transition from the tooling surface to the carrier foil as Hot Spots. Fig. 12 illustrates the ROIs wherein Tow1 does not have any temperature anomalies that are higher than the calculated threshold.
AC C
EP
Table 1 gives an overview of relevant occurring defects during the fiber placement process corresponding to Fig. 1 and the possibilities to identify them via the tow localization and the surface inspection. The combination of both technics has the ability to detect all defects. 4 Conclusion and Outlook The monitoring of Automated Fiber Placement processes via thermal analysis is a promising approach for steady quality assurance in composite structure manufacturing. The shown thermographic monitoring system can determine and store the tow position and is able to detect process relevant defects.
[FIGURE 11 HERE]
[FIGURE 12 HERE]
5
ACCEPTED MANUSCRIPT [Table 1 HERE]
Table 1. Overview of detectable defects via thermographic process monitoring.
Tow geometry Gap/Overlap
End of tow
Twisted tow
Surface detection
RI PT
Edge detection
Bridging
Foreign bodies Change of surface area material
SC
Type of defect
M AN U
For industrial application it is important to increase the reliability and to ensure the real-time capability. Therefore, the system will be integrated in the newly developed Fiber Placement head (Fig. 13). Within a new project called “Therm-O-Plan”, funded by the Central Innovation Programme for SMEs, this monitoring system will be further developed in cooperation with industrial partners to become a reliable system.
EP
TE D
Furthermore, improvements are planned to amplify the developed module. For example, the tack between the tows and the surface can be connected to the time depending temperature adjustment of the laid-up tows. In addition, the edge detection algorithm should also detect the tows in a previous laid-up row to control the repeat accuracy of the placing machine. Moreover, a model should predict the temperatures to evaluate the actual temperature.
AC C
[FIGURE 13 HERE]
5 Acknowledgements
The authors would like to thank the federal state of Lower Saxony and the European Regional Development Fund (ERDF) for funding this research work. They also thank the Central Innovation Programme for SMEs (ZIM) for funding the ongoing research. For further information visit the website www.hpcfk.de.
References [1]
Rudberg T., Neilson J., Henscheid M., Cemenska J. Improving AFP cell performance. In: SAE International Journal of Aerospace Manufacturing and Automated Fastening Conference, September, 2014.
6
ACCEPTED MANUSCRIPT Halbritter A., Harper R. Big Parts Demand Big Changes to the Fiber Placement Status Quo. SME Composites Manufacturing, Mesa, AZ, 2012.
[3]
Soucy K. In-process monitoring for quality assurance of automated composite fabrication. In: review of progress in quantitative nondestructive evaluation, plenum press, New York 1996.
[4]
Schmitt R., Pfeifer T., Orth A. Feasible production of fiber-reinforced composites through inline inspection with machine vision. In: Proc IMEKO World Cong. Rio de Janeiro, Brazil; 2006.
[5]
Miene A. Genau in die Textur geschaut, Kunststoffe, Vol. 5, 2009, p. 62-65.
[6]
Drinkwater B. W., Wilcox P. D. Ultrasonic arrays for non-destructive evaluation: A review, NDT&E International, Vol. 39, 2006, p. 525-451.
[7]
Nguyen C. D., Krombholz C., Röstermundt D. Einfluss einer Online Bahnkorrektur auf die Materialeigenschaften von Prepreg Tows im Fiber Placement Prozess, Deutscher Luft- und Raumfahrtkongress 2012, DocumentID: 281408.
[8]
Shadmehri F., Ioachim O., Pahud O., Brunel J.-E., Landry A., Hoa V., Hojjati M. LaserVision Inspection System For Automated Fiber Placement (AFP) Process, 20th International Conference on Composite Materials Copenhagen, 2015.
[9]
Ritter J. A., Sjogren J. A. Real-time infrared thermography inspection and control for automated composite marterial layup, Patent US7513964, 2009
M AN U
SC
RI PT
[2]
Figure captions
TE D
Fig. 1. Occurring defects during Automated Fiber Placement process. Fig. 2. Schematic draft of the Fiber Placement head.
Fig. 3. Thermal image of the tow and surrounding surface during the lay-up process.
EP
Fig. 4. Definition of the region of interest for the tow localization by an IR camera; left: view of the laid-up tows and right: thermal image of the corresponding process. Fig. 5. Tow localization within the temperature profile of the ROI. Fig. 6. Definition of the regions of interest for the surface inspection by an IR camera.
AC C
Fig. 7. Fiber Placement system including an integrated infrared camera. Fig. 8. Result of the edge detection during a single path with two events. Fig. 9. Surface inspection of bridging over a gap; left: view of the laid-up tows and right: 3D-thermal image of the defect. Fig. 10. Detected cold and hot spots in the ROIs. Fig. 11. Surface inspection of a foreign body; left: view of the laid-up tows and right: 3D-thermal image of the defect. Fig. 12. Detected hot spots in the ROIs. Fig. 13. IFW Fiber Placement head including IR camera.
7
ACCEPTED MANUSCRIPT Table captions
AC C
EP
TE D
M AN U
SC
RI PT
Table 1. Overview of detectable defects via thermographic process monitoring.
8
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S1359-8368(16)30528-5
DOI:
10.1016/j.compositesb.2016.04.076
Reference:
JCOMB 4275
To appear in:
Composites Part B
Received Date: 17 March 2016 Accepted Date: 24 April 2016
Please cite this article as: Denkena B, Schmidt C, Völtzer K, Hocke T, Thermographic Online Monitoring System for Automated Fiber Placement Processes, Composites Part B (2016), doi: 10.1016/ j.compositesb.2016.04.076. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Thermographic Online Monitoring System for Automated Fiber Placement Processes BEREND DENKENA, CARSTEN SCHMIDT, KLAAS VÖLTZER*, TRISTAN HOCKE Institute of Production Engineering and Machine Tools, Leibniz Universität Hannover, Ottenbecker Damm 12, 21684 Stade, Germany
RI PT
*Corresponding author. E-Mail: [email protected]
M AN U
SC
Abstract: Automated Fiber Placement processes are commonly used to manufacture lightweight structures e. g. for highly demanding aerospace applications. In general, quality inspection is usually carried out manually and it is considerably time consuming in terms of high lot sizes and huge part dimensions. An online AFP process monitoring based on thermal camera combined with process depending image processing is presented. The visible temperature difference between the laid-up tow and its surface underneath is analyzed and information of lay-up defects such as overlaps, gaps, twisted tows and bridging derived. This monitoring system can reduce the efforts in quality inspection and will help to increase process reliability significantly. Keywords: A. Prepreg, D. Thermal Analysis, D. Process Monitoring, E. Lay-Up
1 Introduction
EP
Positioning defects (gaps, overlaps, missing tows, twisted tows) Bonding defects (bridging, air pockets) Foreign bodies Tow defects
AC C
TE D
The Automated Fiber Placement (AFP) process evolves into one of the leading manufacturing processes in lightweight industries, especially in aerospace applications. This process is able to reach high productivity by guaranteeing continuous high quality and economic manufacturing. Within this process, several pre-impregnated carbon fiber slit-tapes are laid-up by a Fiber Placement head on a tooling surface side-by-side simultaneously. During the lay-up process, certain defects may occur as schematically shown in Fig. 1. They can be divided in four categories:
[FIGURE 1 HERE]
Today, the operator does the quality control during AFP processes manually by visual inspection. To do so the lay-up process has to be interrupted. The visual inspection and the repair process in cases of detected defects are done for every single ply during the AFP process and therefor are very time consuming. In literature, the effort is numeralized from 32 % [1] to 63 % [2] of the overall AFP production time. Hence, the time of non-productivity caused by visual inspection and defects provoke high costs. If defects are not detected during the AFP process they need to be detected and repaired after curing the part which is even more expensive. Therefore, the reliability and the time of the inspection process are intended to be optimized. 1
ACCEPTED MANUSCRIPT
RI PT
A camera system mounted on the Fiber Placement head can be a support for the operator [3]. By this, he is able to do a continuous inspection of currently laid-up slit-tapes. However, this method demands a high degree of concentration, thus it is not practical for an online process monitoring. Furthermore, the black prepreg slit-tapes prevent a high visual contrast in between the single tows and the plies on the tooling surface. Consequential an intensive illumination and a complex image analysis are necessary, thus a visual monitoring is very ambitious to realize [4]. However, Miene [5] presents an approach wherein an algorithm calculates the fiber orientation for single rovings during the manufacturing process of carbon fiber reinforced plastics (CFRP) preforms and reasons occurring defects. The reliability depends highly on the ambient conditions and the required computing time for the image analysis is significantly high.
SC
The quality control for fiber-reinforced components via ultrasonic probes is prevalent [6]. This method is conducted after the manufacturing process. Until now, there is no case known where ultrasonic probes are used for online monitoring of the placed tows.
M AN U
In the context of research activities laser triangulation sensors are used to monitor single tows [7]. The monitor system can localize the position of the tows by an edge detection.
TE D
Shadmehri et al. developed a Laser-Vision inspection system to detect the ply location, the fiber orientation and gaps [8]. Therefore, he combined a laser projector system with a vision system. Initially a laser is scanning the targets before projecting the characteristic properties on the surface. Afterwards the operator can inspect certain sections of the surface on a monitor. As [8] concludes there is no complete solution for an inspection of the fiber placement process available and in addition it is challenging to integrate the process monitoring in the process itself for an online inspection. The US patent 7513964 B2 [9] describes a thermal AFP monitoring system. During the Fiber Placement, heat or cold is applied to the tows and an IR-camera pictures the process. The composition of these pictures shows the whole ply and can be analyzed to detect certain defects in lay-up. There is no realization of this patent known so fare.
AC C
2 Description
EP
In the following, a new approach for a thermographic online process-monitoring system for analyzing the thermal effects during the automated lay-up process is presented.
2.1 Thermographic Process Monitoring Looking at the AFP process the thermoset prepreg tows need to be kept cool inside the Fiber Placement head to prevent fouling and to ensure good material quality during processing. A second mayor quality aspect is the tack between the tow and it’s subsurface. To ensure a good tack, the surface of the tooling or previous laid tows is heated right before the lay-up and the tow is compacted under specific, controlled pressure. The schematic setup is shown in Fig. 2.
[FIGURE 2 HERE]
Right behind the compaction roller, the surface temperature of the laid-up tow exhibits a lower temperature than the surrounding surfaces. An Infra-Red (IR) camera can detect this tem2
ACCEPTED MANUSCRIPT perature difference as shown in Fig. 3. Due to the position and the angle of the camera, the picture shows a 40 mm sharp area wherein certain Pixel rows are used to determine the edges of the laid-up tows. The IR camera that was used for the first experiments has a resolution of 320 x 256 Pixel, which corresponds to about 7
.
The goal is to develop a process monitoring system that is able to detect defects for lay-up
RI PT
speeds up to 1 . Therefore, the camera frame rate of 60 and the computing speed of
the developed algorithm limit the real-time capability of the monitoring system.
[FIGURE 3 HERE]
2.2 Tow Localization
TE D
M AN U
SC
By evaluating specific regions behind the roller and by applying an edge detection algorithm, the tow geometry and its position can be extracted and monitored. This information reveals many placing defects like overlaps, gaps, twisted tows and spliced tows. Furthermore, the bond influences heat transfer from the substrate to the tow surface. Therefore, the temperature of the placed tow can be used as a quality indicator of the bond, as well. The temperature of the heated surface behind the compaction roller depends on the heat absorption, the convection and the heat transfer into the tooling. A homogeneous temperature transition is expected. If there are inhomogeneous sections in the laid-up laminate, their temperature differs from the surrounding surface temperature. These hot or cold spots indicate anomalies such as bridging defects or foreign objects. Knowing the normal temperature picture of the actual process, hot and cold spots are detectable by dynamic thresholds depending on the actual process. Both the edge detection and the surface inspection algorithms are separately applicable.
AC C
EP
Tow localization is a major part of the process monitoring. To detect the tow position and geometry the algorithm analyzes the temperature profile of the laid-up tows in the Region of Interest (ROI), right behind the compaction roller, as pictured in Fig. 4. The area close to the compaction roller is most suitable to localize the tows since the temperature contrast is highest right behind the compaction roller. The ROI consists of a certain number of Pixel rows to minimize computing errors through the noise of the raw data.
[FIGURE 4 HERE]
The developed algorithm calculates the edges of the tows so that the position can be determined and compared to the planned position. Information about the tow position is stored in the robot coordinate system. As can be seen in Fig. 5, the temperature profile also includes further information about the transition between single tows.
[FIGURE 5 HERE]
3
ACCEPTED MANUSCRIPT Hence, the algorithm can estimate the tow width, the gap width between the simultaneous placed tows and check for overlap of placed tows.
RI PT
Therefore, the temperature in between the tows is analyzed in detail (Fig. 5). If the temperature is similar to the subsurface temperature, a gap is identified next to the tows. If in contrast the temperature between two tows is lower compared to the tow temperature, the criterion for an overlap is fulfilled. The lower temperature in this case results from the double amount of material placed in this region. For the calculation of the tow width, a small gap width is presupposed to detect the edges of the tow reliably. If the gap width increases a certain value and differs from planned tow path, a rejection criterion is fulfilled, which is identified by the monitoring system. The same appears to any overlap.
SC
2.3 Surface Inspection
M AN U
In addition to the tow localization, it is also possible to evaluate the surface temperatures. An anomaly of temperature can represent bonding and planar defects, for example bridging or foreign bodies.
[FIGURE 6 HERE]
EP
3 Results
TE D
For a detailed analysis, certain ROIs are defined as pictured in Fig. 6. The ROIs adapt the shape of the laid-up tows and mask them as well as the left and right area next to the tows. The temperature transition in a single ROI should be homogeneous. If there is a region of higher or lower temperature it can be detected by checking with a calculated threshold value. Combined with the area size of these temperature anomalies the surface inspection can determine defects.
AC C
To test the online process monitoring an infrared camera is integrated in an experimental Fiber Placement setup (Fig. 7).With this setup different process conditions are simulated like accurate and faulty lay-up to test and improve the developed algorithm.
[FIGURE 7 HERE]
Fig. 8 shows an exemplarily lay-up result with different defects. As a first event, a tow twist can be identified in the plot of temperature profiles over the time. The twisted tow induces an increased gap between two tows. Additionally, the left tow edge is leaving the tolerance band according to the path planning. This is also captured in the contour plot, which is supplied by the algorithm, as well. This deviation will result in a defect indication. In the top of Fig. 8 a second event occurs, which is here an unexpected tow ending. Due to the missing tow edge, this defect is also easily detectable. Compared to the planned tow path, this missing tow will lead to a second defect indication. 4
ACCEPTED MANUSCRIPT [FIGURE 8 HERE]
RI PT
In addition, the surface inspection monitors local temperature anomalies of the tows and the surrounding surface. By comparing the temperature inside the ROIs to a calculated threshold, foreign bodies and adhesion defects get visible. The developed algorithm detects this defects by analyzing the temperature anomalies and interpreting of hot and cold spots.
SC
Fig. 9 illustrates this surface inspection procedure for a bridging tow over a gap. Due to the fact that there is no connection between the tows and it´s subsurface while the tows are bridging, the heat conduction substantially decreases. Caused by the decreased heat transfer the tows remain cold. Therefore, the colder areas of the tows are detected as cold spots (Fig. 10).
M AN U
[FIGURE 9 HERE]
[FIGURE 10 HERE]
TE D
The detection of foreign bodies during the Fiber Placement process is highly significant. Figure 11 shows an example for the lay-up over a carrier foil on a tooling surface. Due to the minor foil thickness compared to the tows no bonding defects (bridging) between the tows and the tooling surface occur. Moreover, the thermal properties of the foil differ substantially from the CFRP and the tooling material. Hence, the surface inspection in this area detects the extreme temperature transition from the tooling surface to the carrier foil as Hot Spots. Fig. 12 illustrates the ROIs wherein Tow1 does not have any temperature anomalies that are higher than the calculated threshold.
AC C
EP
Table 1 gives an overview of relevant occurring defects during the fiber placement process corresponding to Fig. 1 and the possibilities to identify them via the tow localization and the surface inspection. The combination of both technics has the ability to detect all defects. 4 Conclusion and Outlook The monitoring of Automated Fiber Placement processes via thermal analysis is a promising approach for steady quality assurance in composite structure manufacturing. The shown thermographic monitoring system can determine and store the tow position and is able to detect process relevant defects.
[FIGURE 11 HERE]
[FIGURE 12 HERE]
5
ACCEPTED MANUSCRIPT [Table 1 HERE]
Table 1. Overview of detectable defects via thermographic process monitoring.
Tow geometry Gap/Overlap
End of tow
Twisted tow
Surface detection
RI PT
Edge detection
Bridging
Foreign bodies Change of surface area material
SC
Type of defect
M AN U
For industrial application it is important to increase the reliability and to ensure the real-time capability. Therefore, the system will be integrated in the newly developed Fiber Placement head (Fig. 13). Within a new project called “Therm-O-Plan”, funded by the Central Innovation Programme for SMEs, this monitoring system will be further developed in cooperation with industrial partners to become a reliable system.
EP
TE D
Furthermore, improvements are planned to amplify the developed module. For example, the tack between the tows and the surface can be connected to the time depending temperature adjustment of the laid-up tows. In addition, the edge detection algorithm should also detect the tows in a previous laid-up row to control the repeat accuracy of the placing machine. Moreover, a model should predict the temperatures to evaluate the actual temperature.
AC C
[FIGURE 13 HERE]
5 Acknowledgements
The authors would like to thank the federal state of Lower Saxony and the European Regional Development Fund (ERDF) for funding this research work. They also thank the Central Innovation Programme for SMEs (ZIM) for funding the ongoing research. For further information visit the website www.hpcfk.de.
References [1]
Rudberg T., Neilson J., Henscheid M., Cemenska J. Improving AFP cell performance. In: SAE International Journal of Aerospace Manufacturing and Automated Fastening Conference, September, 2014.
6
ACCEPTED MANUSCRIPT Halbritter A., Harper R. Big Parts Demand Big Changes to the Fiber Placement Status Quo. SME Composites Manufacturing, Mesa, AZ, 2012.
[3]
Soucy K. In-process monitoring for quality assurance of automated composite fabrication. In: review of progress in quantitative nondestructive evaluation, plenum press, New York 1996.
[4]
Schmitt R., Pfeifer T., Orth A. Feasible production of fiber-reinforced composites through inline inspection with machine vision. In: Proc IMEKO World Cong. Rio de Janeiro, Brazil; 2006.
[5]
Miene A. Genau in die Textur geschaut, Kunststoffe, Vol. 5, 2009, p. 62-65.
[6]
Drinkwater B. W., Wilcox P. D. Ultrasonic arrays for non-destructive evaluation: A review, NDT&E International, Vol. 39, 2006, p. 525-451.
[7]
Nguyen C. D., Krombholz C., Röstermundt D. Einfluss einer Online Bahnkorrektur auf die Materialeigenschaften von Prepreg Tows im Fiber Placement Prozess, Deutscher Luft- und Raumfahrtkongress 2012, DocumentID: 281408.
[8]
Shadmehri F., Ioachim O., Pahud O., Brunel J.-E., Landry A., Hoa V., Hojjati M. LaserVision Inspection System For Automated Fiber Placement (AFP) Process, 20th International Conference on Composite Materials Copenhagen, 2015.
[9]
Ritter J. A., Sjogren J. A. Real-time infrared thermography inspection and control for automated composite marterial layup, Patent US7513964, 2009
M AN U
SC
RI PT
[2]
Figure captions
TE D
Fig. 1. Occurring defects during Automated Fiber Placement process. Fig. 2. Schematic draft of the Fiber Placement head.
Fig. 3. Thermal image of the tow and surrounding surface during the lay-up process.
EP
Fig. 4. Definition of the region of interest for the tow localization by an IR camera; left: view of the laid-up tows and right: thermal image of the corresponding process. Fig. 5. Tow localization within the temperature profile of the ROI. Fig. 6. Definition of the regions of interest for the surface inspection by an IR camera.
AC C
Fig. 7. Fiber Placement system including an integrated infrared camera. Fig. 8. Result of the edge detection during a single path with two events. Fig. 9. Surface inspection of bridging over a gap; left: view of the laid-up tows and right: 3D-thermal image of the defect. Fig. 10. Detected cold and hot spots in the ROIs. Fig. 11. Surface inspection of a foreign body; left: view of the laid-up tows and right: 3D-thermal image of the defect. Fig. 12. Detected hot spots in the ROIs. Fig. 13. IFW Fiber Placement head including IR camera.
7
ACCEPTED MANUSCRIPT Table captions
AC C
EP
TE D
M AN U
SC
RI PT
Table 1. Overview of detectable defects via thermographic process monitoring.
8
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT