Ian Jones and John Rudlin
Paper presented at Joining Plastics 2006, London, National Physical Laboratory (NPL), 25-26 April 2006.
Biographical note of speaker
Ian Jones joined TWI after a degree in Materials Science at the University of Cambridge, UK. He has worked on a wide range of laser materials processing applications, including high power laser welding of metals and non-metals. Heis now in the polymer joining group at TWI and has led developments in many methods of using lasers for joining plastics and textiles.
Laser welding allows precise control over the position and energy applied for a weld. To make the most effective use of the process, a method is required of controlling the weld quality at the time of welding, preferably usingreal-time rapid monitoring and feedback to control the laser power. Transmission laser welding makes use of an absorbent medium, either as a coating or in the lower polymer resin, at the joint which heats in the laser beam andmelts/welds the surrounding plastics. Infrared (IR) thermography has been used effectively when carbon black is used as an absorbent medium for the weld. Use of low visible colour IR absorbers can use alternative procedures, whichoffer opportunity to identify whether a weld has been made or not. Monitoring methods based on thermography, visible and IR imaging and spectrometry are assessed and their performance compared when used in laser welding using carbonblack and low visible colour absorber alternatives.
Using laser radiation as the energy source enables precisely controlled welding both in terms of the position and the amount of applied energy. To optimise this, methods are required to verify the correct location of a laser absorbing dye that is applied as a coating, the welding process and to view the weld after completion to ensure that consistent quality of the joints is produced. The ideal monitoring method would provide real time measurements of the weld as it forms, allowing continuous control of the process in a feed back loop, but much can also be gained from methods that provide information about the parts and process before and after welding. To some extent the use of low visible colour infrared (IR) absorbers (ie. in the Clearweld ® process  ) instead of carbon black makes monitoring more complex, because the weld is difficult to see, but changes occur during welding that are specific to IR absorbing coatings that can be used as an aid to determine whether a weld has been completed successfully.
To determine the applicability of infrared and visible light methods for monitoring welds made using low visible colour infrared absorbers.
Description of the relevant monitoring methods
Although several commercial imaging techniques, such as infrared, ultraviolet, ultrasonic, and visible light imaging [2,3,4] , are available for viewing welding phenomena in metals and plastics, the transmission laser welding process poses a unique challenge, whereby the weld interface is typically behind two layers of polymer sheet (see Fig.1); the pressure transmitter sheet, typically made from poly(methyl methacrylate) (PMMA), and the top layer of the part to be welded Any top face monitored image of the weld region must therefore pass through these two layers. In addition, methods using IR absorbing coatings are designed to have minimal visible impact, making some monitoring methods difficult to use. Low colour IR absorbers can also be used in the resin of the lower part in a similar way to the normal use of carbon black absorber. Monitoring methods needed for this case are different again compared to use of IR coatings.
Fig.1. Schematic drawing of the transmission laser welding process using IR absorbing coating
When considering the imaging methods available, it is first important to consider the radiation transmission properties of polymers in general and how this might affect the performance of the monitoring method.
PMMA is transparent at visible and near infrared wavelengths, but absorbs in the infrared region beyond 2200nm as shown in Fig.2. In general, the transmission properties of other transparent polymers in the visible and near-infrared are similar to PMMA. This limits the operating wavelength for any imaging system that needs to see through to the weld interface.
Fig.2. Typical ultra-violet, visible and near-infrared transmission spectrum of the PMMA
Visible light imaging
Visible light imaging uses a video camera sensitive to visible light. It should be noted that some solid state CCD cameras are sensitive in the infrared as well as visible. If this is the case, then a filter can be used to block any radiation from the laser that would interfere with the images. In addition, an image analyser can be used to provide automated feedback of the items being viewed. Visible light imaging has been described when welding clear to blackparts  , where variations in intensity of the dark mark at the joint indicate changes in the quality of the joint. In theory this can be extended to welds between two clear materials if the changes in reflection from the surfaces to be joined can be identified.
Visible light imaging is appropriate when transparent and translucent materials are used as the top layer. It is a low cost option that uses off-the-shelf equipment. However, this method would be unsuitable when visibly opaque materials are used as the top layer.
Thermographic techniques remotely measure the radiation emitted from a hot surface by virtue of its temperature. The temperature of the surface can be calculated from the spectrum and intensity of the radiation emitted. The strength of the signal available is the product of the spectral radiance and the material's emissivity at a given wavelength. Emissivity values range from 1.0 for lampblack down to 0.02 for polished silver. Thermography is used effectively for monitoring laser welding where carbon black is used as the absorber. Use of the method presents difficulties if the emissivity of the heated region is low.
The relationship between radiance and the spectral radiation wavelength (cm), λ, at a given temperature (Kelvin), T, is given by Planck's equation , plotted in Fig.3. The Planck equation for spectral radiance describes the thermal emission from an ideal blackbody cavity of any material  , with an infinitesimally small aperture and in an absolute isothermal condition.
The wavelengths at which infrared thermometers are sensitive defines the range of temperatures that can be measured. Figure 3 indicates that to make measurements of thermal radiance in the melting range of plastics (150-400°C), it is necessary to monitor at wavelengths longer than approximately 1800nm where the radiance is at measurable levels (>0.1 lumens). If the materials being measured have low emissivity, then a longer measurement wavelength would be preferred to make use of the more intense region of the spectrum.
Fig.3. Relationship between spectral radiance from a black body emitter to the temperature of the emitter and the wavelength of the radiation
For transmission laser welding, Table 1 indicates that a short wavelength thermographic system is required to measure the weld interface temperature. Combining the information from Figs.2 and 3 it is concluded that it is preferable to monitor at infrared wavelengths shorter than 2200nm, where the PMMA transmission is greatest, and at wavelengths longer than 1800nm to read
Table 1 Infrared wavelength and detectable temperature range  for commercially available thermographic equipment.
|Infrared wavelength (nm)||Temperature range (°C)||Applicability to Polymers|
|8000 to 14000
||-50 to +500
||Low transmission; can only read temperature at top surface
|1000 to 2000
||+150 to +3000
||High transmission; can read temperature within component
Infrared imaging is carried out in the same way as visible light imaging, with cameras sensitive to the near infrared range of wavelengths. In principle, if wavelengths of the thermal emission spectrum from the weld are used, then a thermal picture of the weld can be seen. However, this procedure is subject to the same limitations as described above for infrared thermography. A method of using an IR camera to image a complete weld line immediately after the welding process has finished has been described when welding clear to black polycarbonate parts  . Regions of differing temperature could be identified and could indicate variations in weld quality along the joint.
Alternatively, a camera can be selected that is sensitive to the absorption wavelength of the infrared absorber used in the welding process. The camera would then be used to monitor the position of the coating before, during and after welding. The infrared absorption of the coating gradually decreases during welding, so potentially this could offer a good monitoring method. The difficulty in applying this method at the time of welding is that it is difficult to filter out interference from the laser wavelength, as this is selected to match the absorber. Because of this, the technique has only been applied to monitor the weld zone before and after welding. Image analysis may also be used in conjunction with infrared images taken.
Spectroscopic transmission measurements
In this method, a spectrophotometer is used to measure the absorption properties of the IR absorber and substrate materials, over a range of wavelengths. This procedure may be used before and after the coating deposition to check the position and amount of absorber present. As a result of the laser heating, some or all of the dye decomposes during the welding process in to products that have a different absorption signature. Measurements made using a spectrophotometer before and after welding can therefore reveal information about the heat generated during welding.
This method is currently limited to use before and after welding for similar reasons to infrared imaging. To use this method at the time of welding, in addition to filters for the laser wavelength, processing of the recorded spectrum would have to be very rapid.
Summary of monitoring requirements
Monitoring may be applied for a number of reasons, both assessing that various stages of the process have occurred, as well as indicating the quality of the weld:
- Check infrared absorber is applied correctly to parts before welding
- View of weld during process
- Indication that parts are in contact during process
- Indication that weld heating is being carried out and control of temperature
- View of weld after completion
- Indication that parts are in contact after completion
- Indication that weld has been achieved after completion
- Indication that weld quality and strength are satisfactory
Monitoring methods can give information for most of these points with varying degrees of confidence, however it must be noted that a complete quality and weld strength assurance can only be achieved with destructive testing.
The generic configuration of the welding set up is shown in Fig.1. A computer-controlled Laserline diode laser was used for the experimental work, delivering up to 150W at 940nm wavelength. The beam was directed onto the specimens through a pressure transmitting sheet of PMMA, and then through the upper material being welded to the joint interface. Clamping was applied via a pneumatic shoe onto the PMMA sheet Specimens were cut from 3mm thick, transparent or coloured cast sheet of PMMA or polycarbonate. An infrared absorber  was deposited using a needle dispenser at the interface between the upper and lower layers of the sample to be welded in the lap configuration The laser beam is absorbed by the coating, which generates heat at the joint to melt and weld the substrates. Four monitoring methods were applied before, during or after welding.
- Visible Light Imaging. Video cameras with long focal length (200mm) and short focal length (15mm) lenses were used to record real-time welding images. Tungsten lighting was used to illuminate the weld regions.
- Infrared Thermography. Two systems with different wavelength sensitivity were used. An Agema Thermovision 900 infrared camera, with a minimum emissivity setting in the wavelength range of 8000 to 14000nm, was used to determine if the weld temperature could be measured in the absence of the PMMA pressure transmitting sheet. An infrared pyrometer, sensitive in the range 2150-2650nm, was used with the aim of measuring the weld interface temperature. The measuring wavelength of the pyrometer corresponded to a temperature range of 140 to 700°C. For black polycarbonate an emissivity of 0.8 was used, and for the clear material the lowest available setting of 0.01 was used.
- Infrared imaging. The vision system used consisted of a Cohu 4812 monochrome CCD camera, an LED backlight (wavelength 880nm) and Optimas 6.5 image processing software. The system was used to monitor the location of the coating before welding and to measure the extent of dye degradation after exposure to the laser beam. This served as an indication that heat had been generated in this region. Image processing with grey scale evaluation capability was also used to make direct comparisons between uncoated, coated and welded regions of the materials being welded.
- Spectrophotometer measurements. The equipment used in these trials was a Cary 500 double beam scanning ultra-violet/visible/near infrared spectrophotometer. Trials were carried out to compare the transmission spectra of uncoated polycarbonate (two layers), polycarbonate with infrared absorbing coating upon it (two layers with the absorber at the interface), and laser welded polymer. An area of 1cm 2 was examined in each case.
Visible light imaging
A snapshot of the video taken during welding of 3mm thick transparent PMMA to PMMA using a 5mm diameter, 15 W laser beam, is shown in Fig.4. The welded region, where the melted polymer resides and where two surfaces have been disrupted and linked together, is visible because of the change in reflectivity. A video camera focussed at the tip of the melt front showed the progress of the weld as it occurred. Regions where no weld is made are visible, as shown in Fig.4. The contrast between the welded and unwelded regions can be enhanced by suitable digital imaging manipulation.
Fig.4. Snapshot of video picture taken during welding of two PMMA sheets
The infrared camera, working at 8000-14000nm, measured the temperature at 28°C at laser welding settings that were shown to melt a layer of PEEK film (melting point of 343°C). Essentially, the temperature reading was that of the top surface of the PMMA cover sheet, well away from the heated zone.
The infrared imaging system working at 2150-2650nm gave a reading of approximately 165°C when carbon black was used as an absorber for the laser source. When infrared absorbing coating was used there was no indication of the weld temperature for clear to clear substrate welding (see Fig.5).
More recent results have shown that it is possible to take temperature measurements from welds made with infrared absorber when sensors of other wavelength ranges are used.
Fig.5. Results of infrared thermography of welds made in polycarbonate to compare welding of clear-clear samples with clear-black samples. The minimum temperature reading for the equipment was 140°C, so the clear-clear sample did not provide a reliable reading
Images of the infrared absorber coated samples, taken prior to welding ( Fig.6) clearly show variation in coating deposits on the surfaces to be welded. When welded, the samples with good coating coverage had higher weld strengths.
Fig.6. Infrared images of weld face areas for polycarbonate samples made after coating and before welding, indicating variations in coating concentration a) poor deposit b) good deposit
Near infrared imaging was also used to verify that a weld had been produced. Figure 7 shows an image with deliberate welded and unwelded areas. As shown, the image processing software can detect the difference in grey values between welded and unwelded sections.
Fig.7. Near infrared image of weld in polycarbonate showing welded and unwelded regions
The spectrophotometer was used to compare the transmission of substrates, substrates plus infrared absorber prior to welding and the substrates after welding ( Fig.8). As shown, prior to coating, the substrate shows transmission of approximately 85% in the visible and near infrared range. After coating the IR absorber, a clear reduction in transmission is detected in the near infrared range due to the presence of the infrared absorber. After welding the near infrared absorption has almost disappeared and the transmission in the visible region has increased to 90%. The transmission increase in the visible region as a result of the formation of a weld between the two surfaces, and therefore the disappearance of two reflecting surfaces, is apparent. The increase would also be apparent in the IR region, but is masked by the absorber.
Fig.8. Spectra displaying pre- and post-weld transmission measurements made using a spectrophotometer over a 1cm 2 area
Results from this work show that when using infrared absorber to weld transparent polymers, spectrophotometry and infrared imaging provide useful information about the joint before, and after, making the weld. Visible imaging and thermography can provide information during welding. The use of infrared absorber in a coating provides extra opportunities for weld monitoring purposes.
Infrared thermography is successfully applied in laser welding of plastics where carbon black is used as an absorber and a strong signal of thermal radiation is produced. These methods are used routinely in plastics laser welding equipment for plastics. However it is more difficult, though not impossible to monitor the temperature when using low visible colour infrared absorber coating as the laser absorbing medium. The difficulties are possibly explained by the low emissivity of the heated region, compounded with the fact that plastics do not transmit the radiation wavelengths that are most useful for temperature measurement. The measurement is further limited if the materials aresemi-crystalline polymers and hence scattering to the emitted thermal radiation. However, IR thermography is useful as a weld monitoring method and provides measurements rapidly enough for real-time weld temperature control. When the IR absorber dye is used in the polymer resin, IR thermography can also be used in a similar way to its use with carbon black absorber.
Near infrared imaging can be used as a monitoring method to detect variations in coating concentration along the weld line before welding. It can also detect thermal decomposition of the dye after welding, and hence identify both the location and extent of the weld made. It did indeed prove to be a successful technique for verifying that the absorber coating had been applied in the correct location and, by virtue of the changing transmission properties of the dye upon heating, it could be used to monitor changes between welded and unwelded regions.
Infrared imaging is very effective as a monitoring method before and after welding. A number of factors need to be considered before it can also be used at the time of welding: These include:
- Reducing background radiation
- Preventing saturation of the detector by the laser source
- Rapid image processing and interpretation
When the top polymer layer is transparent or translucent to visible light, video monitoring can be used to view the welding process as it is proceeding. The images can be collected quickly and the cameras are less prone to saturation by the laser source than IR devices. The weld region can be seen because the reflection from the two surfaces at the joint is lost when they weld together. The potential of this method as a diagnostic technique for laser welding has been described to identify the difference between welded, unwelded and blackened upper surfaces.  Although only a destructive test can measure the weld strength, this monitoring method, showing that the surfaces have been brought into contact, combined with control of the material surfaces before welding and monitoring of the welding machine parameters, will provide high assurance that a good weld has been made. An automated imaging system should be able to define the regions of welded and unwelded material, hence providing a viable monitoring method. In our previous work it has been shown that this method is also applicable to laser welding when carbon black filled material is used in the lower layer.
Other monitoring methods can be used in combination with the imaging methods discussed here. Where melting causes a change in dimension of the part, displacement control can be applied as an indication of weld completion. This does not guarantee that all gaps are closed, but in combination with a surface reflection imaging method even more confidence is achieved.
It is worth mentioning that computer models [10,11] could be used to map out the relationship between power and process speed as part of a monitoring and control technique. These models can also be used as a means to assist in the optimisation of the processing conditions prior to actual welding operations and as part of a monitoring and feedback system.
Based on the results of this and earlier work, the features of the monitoring methods, assessed for the applicability to laser welding particularly using low colour infrared absorbers are summarised in Table 2.
Table 2 Summary of the features of the monitoring methods
|Monitoring method||Suitable for||Advantages||Limitations||Availability|
|Visible light imaging
||Viewing the weld during and after the welding process. For both absorber resin and coating methods with IR and carbon black absorbers.
||Easy to set-up and use. Indicates weld region where the two joint surfaces have merged.
||Only suitable for transparent and translucent upper materials.
||Temperature measurement and heating control during the welding process. For IR and carbon black resin based systems. Care needed in use with IR absorber coatings.
||Provides rapid temperature reading of use in control of the laser power in real-time.
||Difficult to use with low emissivity materials and where upper material transmission is poor.
||Available at relatively low cost.
||Verifying location of absorber coating before and after welding, showing where heating has occured. Also gives image of heated zones just after welding.
||Proven laboratory technique suitable for checking coatings on parts before welding and the extent of the weld afterwards.
||Interference from the laser wavelength if used at the time of welding. Not useful for weld checking of absorber filled resins after welding.
||Might also be used at the time of welding with further development.
||Measuring the coating absorption before and after the weld, and potentially during the weld. Verification of an intimate surface connection.
||Checks position and concentration of the coating before welding and that remaining after welding.
||Difficult method to use at the time of welding. Not useful for weld checking of absorber filled resins after welding.
||Might also be used at the time of welding with further development.
Monitoring methods for transmission laser welding using low visible colour infrared absorbers have been explored for use before, during and after welding.
- When the top layer of the joint is sufficiently transmissive, visible light video imaging can be used to view the polymers being melted under the laser beam. Contrast between welded and unwelded regions is visible because of the change in reflection from the two joint surfaces when they are melted together.
- Near infrared imaging and spectrophotometry are suitable for pre- and post-welding analysis of an infrared absorbing coating as means of inferring weld formation and location of the weld from changes in the IR dye present during heating. Both methods can be used to check the presence of infrared absorber on the parts before welding, and both can show that the absorber has been used and generated heat in the weld. Spectrophotometry can also be used to monitor the reduction in reflection when the two material surfaces at the joint are lost as a weld occurs. These methods are only applicable when processing with infrared absorbing coatings and not with carbon black absorbers or for IR absorber in resin systems.
- Due to the low and undefined infrared emissivity of material at the joint region, established infrared thermography methods are more difficult to use with low visible colour infrared absorbing coatings than with carbon black absorbers. However, there is evidence that measurements are possible for all laser welding applications with suitable thermography equipment.
- The imaging methods are only useful where the weld is visible through the upper material at the joint. The upper material layer needs to be translucent or transparent to the visible or IR radiation being used. Some materials transmit IR better than visible radiation.
This work was carried out within the Core Research Programme of TWI, which is funded by the Industrial Members of TWI. Gentex Corporation is thanked particularly for work on spectrometry and IR imaging.
- Jones I A and Wise R J: 'Welding Method'. Patent WO 00/20157, 1 October 1998.
- For example see: www.landinst.com and www.meta-mvs.com
- Sun A, Kennatey-Asibu E and Gartner M: 'Sensor Systems for real-time monitoring of laser weld quality'. Journal of Laser Applications, volume 11, number 4, 1999,pp 153-168.
- Smith J S and Lucas W: 'Vision Based Control of Arc Welding Processes'. Welding in the World, vol.46, Special Issue. July 2002. pp.251-262.
- Hierl S and Hofmann A: 'Innovative solutions for laser plastics welding', Proc 2nd Int Conf on Lasers in Manufacturing, Munich, June 2003, pp 385-389.
- Private communication with a representative from the National Physical Laboratory, Aug 2004.
- Private communication with a representative from the Land Infrared company.
- Russek U A: 'Innovative trends in laser beam welding of thermoplastics', Proc 2nd Int Conf on Lasers in Manufacturing, Munich, June 2003, pp 105-111.
- Hilton P A, Jones I A and Kennish Y: 'Transmission laser welding of plastics'. International Congress on Laser Advanced Materials Processing (LAMP 2002), May 2002.
- Rostami S and Jones I: 'Process Guidance and Software for Clearweld ®'. TWI Report 772/2003, August 2003.
- Jones I A and Olden E: 'A thermal model for transmission laser welding of thermoplastic polymers'. TWI members report 708/2000, July 2000.