Laser welding of plastics - process selection software (October 2003)

Paper presented at ICALEO 2003 Conference, October 13-16, Jacksonville, Florida, USA.

Abstract

Lasers are now successfully applied in welding for the manufacture of plastic parts. They offer distinct advantages and performance capability that allow applications to be considered that were previously not possible with other plastics welding methods. High strength and high quality welds are possible together with rapid production rates when the methods are used appropriately. The range of techniques using lasers for welding plastics and synthetic textiles is reviewed, particularly with reference to the use of infrared absorbers to promote precise heating and weld generation. A software package is presented that allows an estimate to be made of the amount of absorber required given the material to be welded and the production requirements, and predicts the laser power and absorber dispensing equipment required for the application. In this process the software uses the laser absorption and heat flow characteristics in the plastics and absorber materials to make the required predictions. The results are discussed theoretically and compared with experimental observations.

1. Background

Since first being reported in 1985, transmission laser welding has been carried out with an infrared transmissive plastic material for the upper section and a carbon black loaded plastic for the lower layer. The carbon black absorbs and heats in the laser beam to generate a weld at the interface between the two pieces. The process is limited by the fact that one side of the component has to be black. The other side of the joint must also transmit a proportion ofthe laser energy (more that 10% is usually enough). An example can seen in Figure 1.

Fig. 1. Laser transmission weld in 4mm thick polypropylene using a 100W Nd :YAG laser at a speed of 1.6m/min. The weld is at the interface between the light and dark materials

A more general extension of the transmission laser welding process has now been developed, which allows completely clear or similarly coloured components to be welded by using an absorber material, which is clear in the visible range of the spectrum, but tailored to absorb heavily the specific wavelength of the laser beam being used. ^[1,2] This technique has been termed Clearweld ^® .

The nature of the absorber means the laser wavelength is absorbed with high efficiency, thus requiring relatively small amounts of the absorber material at the interface between the two components to be welded. Development work on the process has been carried out using polymethylmethacrylate (PMMA) test specimens, and an example of an overlap weld made by applying absorber by ink jet printing to the joint surface between two transparent sheets of PMMA can be seen in Figure 2. When the incident laser light is applied, the absorber molecules dissipate the absorbed energy principally as heat, which melts the joint only at the thin region either side of the absorber coating. In this way theouter surface of the component are unaffected.

Although the example in Figure 2 is shown with two visibly clear sheets of PMMA, an absorber applied in this way can be used to join a wide range of plastics, in moulded or sheet form, coloured or otherwise, or indeed, textiles and flexible film materials.

Fig. 2. Laser overlap weld in clear PMMA made with Clearweld ® absorber at the interface

2. Range of materials that can be joined

Laser welding is applicable to a wide range of thermoplastic materials in various forms. Polymer types that have been welded include the following:

• PMMA, PC, ABS, LDPE, HDPE, PP, PETG, PVC, PET, TPU, Nylon 6, Nylon 66, PEEK, PS.

As the process is defined by local heat absorption only at the location of the absorber material, and not by the use of vibrational or other forms of processing, the heating process can be highly controlled independently of thematerials or designs to be joined. It is also possible, within limits, to weld materials with different melting points as long as interdiffusion is not inhibited. Application is possible to a wide range of material forms andcombinations. Some examples are:

• Moulded, sheet, film, elastomeric, fibres, textiles
• PS/PC, PS/PMMA, PS/ABS, PMMA/PET, LDPE/HDPE

3. Range of absorbing materials

Development work to select the most suitable absorbers and dispensing systems has also been carried out in work at TWI and Gentex Corporation. The vast majority of near-infrared absorbers can be discounted on the grounds that theyhave a pronounced visual colour. Others are unstable or have low absorption at the laser wavelength. Examples of three absorber types which can satisfy all of the requirements are organic molecules such as cyanine, aminium/ diimoniumand squarylium. Materials such as these are used in the absorber systems for Clearweld. These absorbers have been built into a range of commercially available inks for use in welding applications. These have been designed for ink jet printing, needle deposition and spraying, with different solvents suited to different substrate materials and different concentrations suited to the process conditions that need to be used.

Further absorber types are being developed to suit different laser wavelengths in the range 800-1100nm and film forms for dry application are also feasible but not commercially available at present.

4. Process parameters

Thus, in order to have the weld occur, the absorber material must be absent from the upper plastic material and must be localised at least at the surface of the lower plastic. The welding occurs as the heat generated in the absorber. Typically the weld depth is of the order ≈ 0.1mm of the polymer material. The heat generation at the interface is controlled by the absorption coefficient of the absorber layer and the processing parameters. The main welding parameters are laser power, beam spot size, and the welding speed. Intimate contact and clamping is also required for the welding to effective. Experimental procedures are described below leading to the generation of an absorber selection software package.

5. Experimental procedures

Diode laser welding of lap joints in 3mm thick cast polymethylmethacrylate (PMMA) was carried out using three methods to apply the liquid absorber: Ink jet, needle and spray, and absorber applied in film form. A diode laser manufactured by Laserline was used to carry out the welding. It had an output wavelength of 940nm with a maximum power of 150W. The spot produced was approximately rectangular in shape with a minimum focused size of 0.5x0.6mm. In thebulk of this work the beam was defocused to a size of 5x5mm. Weld samples were manipulated beneath the stationary laser system on a two-axis CNC table. They were clamped using a pneumatic actuated ring clamp of 15mm diameter, through which the laser beam was passed. The clamp was used to give an applied load to the overlap region of the sample of 400N The dispensing methods are compared in Table 1 and Fig.3. A range of deposit densities were applied to the PMMA joint interface using the different techniques. These samples were then used for laser absorption measurements and definition of the welding process requirements.

To measure the infra-red absorption of the samples with and without absorber deposits at the laser beam wavelength of 940nm, a laser calorimeter (manufactured by Ophir) was used to compare the energy from a short laser pulse with and without the polymer sample.

Three tensile samples, 25mm wide were cut from each weld sample, avoiding the weld ends. These were mounted in grips, with spacers of the same material thickness to promote pure shear at the start of the test, at a gauge length of120mm. The samples were tested following standard EN 12814-2:1998 at a cross-head speed of 5mm/min. The location of failure was noted in addition to the peak tensile load.

Sections containing approximately 5mm weld length were cut from the weld and mounted in clear epoxy resin. The weld section was polished and viewed between crossed polars with an optical microscope to view the extent of the Heat Affected Zone (HAZ), i.e. that material with its temperature raised above the glass transition temperature during the welding process and cooled rapidly to leave behind a birefringence zone.

Table 1. Characteristics of the liquid dispensing systems

	Ink Jet	Needle	Spray
Line width, mm	0.5-280*	1-5	5-25
Weld geometrical complexity	any pattern	line	line
Rate of flow, ml/min	0.01-12.0	0.4-3.8	0.4-6.0
Deposition of liquid on substrate, nl/mm 2	2-20 (1 pass) 2-80 (up to 4 passes)	50-100 1 pass only	10-75 (1 pass) 10-300 (up to 4 passes)
Deposition of absorber # , dimensionless	2-288	50-1000	10-3000
Some solvents applicable	diacetone alcohol/ methoxy propanol	acetone, methyl ethyl ketone, ethanol, methanol	acetone, methyl ethyl ketone, ethanol, methanol
* A 280mm wide track is achieved by using 4 heads linked together. # The volume of absorber is determined by a commercial calculation based on solvent type and concentration. The figures quoted allow direct comparisons between the different application methods.

The fraction of laser energy absorbed by the deposits was compared with the applied laser energy per unit area required to make a high strength weld. The limiting amounts of applied laser energy required to give parent material failure in tensile tests were defined to compare the different absorber deposition techniques techniques. Selected samples were sectioned and the heat affected zone (HAZ) depth was measured.

Software was written to take information about the material, weld complexity and the processing conditions used to select the type of absorber, the amount of it and its most appropriate deposition method.

Fig. 3. Comparison of the amounts of absorber applied by the different deposition methods

6. Results

All the methods gave welds that remained intact beyond their failure in the parent material adjacent to the weld in PMMA, given that a suitable laser heat input was applied. The laser absorption generated by the different deposition techniques is shown in Fig.4. For each different absorber quantity used (three for each deposition method), the applied laser energy required to generate a parent material failure was defined. These are plotted in Fig.5, which indicates that for similar use of absorber, welds using film require the lowest heat input. The low deposit amounts delivered by ink jet require high applied energy from the laser compared with the higher deposits from needle and spray methods.

Fig. 4. Absorption of incident radiation at 940nm wavelength with respect to the amount of absorber applied using different application techniques

Fig. 5. Comparison of the applied laser energy required to give a weld that fails in parent material on tensile lap shear testing. Any applied energy above the lines will ensure a strong weld

The HAZ depths decrease when lower applied energy (higher welding speed and/or lower laser power) and lower amounts of absorber are used (see Table 2). This is due to lower heat generated per unit volume of material under such conditions. The welds made with ink jet deposited absorber required low welding speeds as well as high applied energy and so tended to give large HAZ values at conditions to give a strong weld. At a welding speed of 1.0m/min, HAZ depths of below approx. 0.1mm were associated with failure at the weld interface. With the other deposition techniques, higher welding speeds and lower powers could be used resulting in small HAZ values at conditions to give a strong weld. Using needle, spray or film deposition, at a welding speed of 6.0m/min or more, HAZ depths of below approx. 0.06mm were associated with failure at the weld interface. Typical weld sections are shown in Figure 6.

Table 2 Measured HAZ depth

Method	Absorber deposited, dimensionless	Abs. %	Applied power W	Weld speed m/min	Adjusted laser energy # J/mm 2	Failure locations 2 or 3 samples	HAZ depth mm
Ink jet	70	8	150	0.2	0.72	PPP	0.52
	70	8	150	0.6	0.24	PPP	0.21
	70	8	150	1	0.14	WWW	0.08
Needle	275	37	100	3	0.15	PPP	0.13
	275	37	100	6	0.07	PPP	0.08
	275	37	100	15	0.03	PWP	0.03
Spray	120	27	100	3	0.11	PPP	0.11
	120	27	100	6	0.05	PPP	0.08
	120	27	100	15	0.02	WW	<0.01
Film	166	68	40	3	0.11	PPP	0.09
	166	68	40	6	0.05	PPP	0.06
	166	68	40	15	0.02	WWW	0.05
P = failure in parent material adjacent to weld, W = failure in weld interface # = adjusted taking account of the % absorbed by the absorber

Fig. 6. Selected weld sections from welds listed in Table 2 showing the difference in the size of the HAZ for different deposition procedures and applied laser energy used 

a) Ink Jet, 9.0J/mm 2

b) Needle, 0.20J/mm 2

c) Spray, 0.20J/mm 2

7. Software development

Experimental data and results of theoretical temperature and melt depth calculations for cast PMMA were used to specify a weld temperature that gives sufficient weld strength to generate failure in a lap shear tensile test in the parent material. Less comprehensive results were generated and used for polycarbonate (PC).

From these results and knowledge of other thermoplastic polymers and their welding characteristics, it has been possible to postulate a melt temperature that is needed for each polymer, to achieve a sufficient weld strength using the Clearweld process.

The absorption properties of the absorber and the applied energy from the laser are used to increase the weld temperature of the material. Knowing this weld temperature, with an assumption on accuracy, enabled the design of an early software tool that will use the material type, weld temperature and applied energy from the laser to indicate how much absorber is required to give a weld which fails in the parent material.

In addition to defining the amount of absorber required, a software tool can also use information about the weld width and complexity to select the most appropriate deposition method for the absorber. At this stage the use of a film absorber has been excluded because it is more appropriate to a different range of applications than the liquid deposition techniques, and these film applications are still in development. The other liquid applications are commercially available.

The software takes the following inputs:

• Material Data:
  - Type by selection from list provided in a drop down window.
  - Thickness between 0.005 to 10 mm
  - Colour selection from a drop down list

• Weld Data:
  - Geometrical complexity from a drop down list
  - Weld width between 0 to 50 mm
  - Weld speed between 0.1 to 500m/min

• Laser Data:
  - Laser power
  - IR transmission with a choice of a default value

From these inputs the program then carries out the following selection and filtration procedures:

1. From the weld width input and information from Table 1, the software excludes absorbers and methods that cannot deliver the track width of absorber required.

1. From the weld complexity input and information from Table 1, the software excludes absorbers and methods that cannot deposit upon the joint line because of its shape, i.e.
   
   
   • Straight lines of even width - all three processes
   • Lines and curves of even width and radii >5mm - all three processes
   • Lines and curves of even width and radii <5mm - ink jet and needle
   • Detail of uneven width and radii down to 0.1mm - ink jet only
2. The software contains data on the solvents used with the absorbers that are commercially available and their compatibility with different polymers. Incompatible combinations are excluded.

1. The fraction of power available for welding is calculated using the material type and thickness with the following equation:

   1 T = (1-R).exp(-a.t)     (1)

   T = transmitted fraction
   R = Surface reflected fraction, assumed to be 0.05 for all materials
   a = absorption coefficient for upper material, mm ^-1
   t = thickness of upper material in mm.

   
   

   Alternatively, the user may input a known transmission value for his material. T is then multiplied by the applied power for the calculations that follow.

2. The proportion of laser energy reaching the joint interface required to make a weld is then calculated using the required weld temperature. This is described as the required surface absorption. If the surface absorption required to make the weld is greater than 75% it is assumed that this will be difficult to achieve using the absorber delivery methods and the user is asked to input either a higher laser power or a lower required welding speed for his procedure.
3. The amount of ink required at the surface is calculated from the required surface absorption using the following equation:

    

abs = IR absorption required to achieve a strong weld taken from section 5 above,

a _w = dye absorptivity (4.2465 m ² /g for A194 at 940 nm)
C = ink concentration (g/l)
T = transmission of laser by upper material from section 4 above

Note that this takes no account of any changes in the efficacy of the ink which may occur as a result of the application methods used, or any changes that may occur in the absorption of the ink during the welding process. These might be added at a later development stage.

Using the result of section 6 and the information from Table 1, absorbers are excluded that cannot be delivered to the material surface at the required rate in nl/mm ² .

The software delivers an output and a copy of data used that describes:

• The laser power used.
• A list of the absorber type and delivery methods in order of the volume of absorber required.
• The input parameters, including the calculated loss of energy in the upper material.

8. Software verification

The software calculations were verified in the following ways:

• comparison of the size of the HAZ formed in welds with a calculated HAZ ( Fig.7)
• calculation of the HAZ size related to weld strengths generated ( Fig.8)
• calculation of the amount of absorber required at the joint to make a strong weld ( Table 3).

Fig. 7. Comparison of measured and calculated HAZ sizes for different absorber application methods

The agreement between predicted and measured HAZ for different deposition methods was shown to be good. The amount of absorber required to achieve parent material failure has been calculated for welding PMMA. This was then compared with the amount deposited in the experimental work described above with the three different liquid deposition techniques, which again compared well (see Table 3).

Fig. 8. Experimental weld strength (failure load) plotted against calculated HAZ depth for PMMA, using ink-jet deposition method for the absorber. This suggests that a HAZ depth in excess of 0.1mm is sufficient toensure a high weld strength, the same as that suggested experimentally in Table 2. The software calculates the conditions required to achieve a weld interface temperature which safely gives the required HAZ size and then gives the amount of absorber required for a selected set of processconditions.

Table 3 Measured and calculated amount of absorber required for a high strength weld in PMMA.

Method	Ink	Amount of absorber deposited nl/mm 2	Laser power W	Max speed giving parent failure m/min	Calculated amount of absorber required nl/mm 2
Ink jet	IJ111E	5	150	0.43	6
	IJ111E	10	150	0.72	8
	IJ111E	19	150	0.90	13
Needle	LD120A	85	100	4.40	88
	LD120B	85	100	20.00*	83
	LD120C	85	100	20.00*	76
Spray	LD120A	60	100	6.70	39
	LD120A	120	100	12.00	180
	LD120A	240	100	20.00*	243
* Maximum speed used in tests, parent failure may still be achieved at higher speeds. NB the software does not include calculations on thin film methods, so they are not included here.

9. Discussion 

The selection of the technique to be used for deposition is based mainly on the practicalities of the methods in relation to the proposed application. Many of the considerations are listed in Table 1, such as the track width that can be deposited, the printing complexity available and the limits of deposition at the joint surface that can be reasonably achieved. In addition, film application can be considered where liquids are not preferred or where additional polymer is required at the joint to fill gaps or to provide a tie layer that is melt compatible with two different polymers to be joined.

In terms of the performance, all the methods can give welds that fail in the parent material adjacent to the weld in PMMA, provided that a suitable laser heat input was applied. The different absorber deposition methods led to different laser absorption and welding performance. As shown by Fig.4, the needle, spray and inkjet deposits gave generally overlapping laser absorption for a given amount of absorber applied, ranging from absorption of about 5% for an applied absorber amount of 60units per surface area upto 73% absorption for 850units. The trend does not appear linear. Ink jet printing at the parameters used was only capable of low deposition rates compared to the other processes, so much higher applied laser energies were required toachieve high strength joints. For a similar deposition of absorber, the film deposits tended to absorb a greater proportion of the applied laser energy than the deposits from an ink solution. This is possibly due to the fact that the absorber molecules are encapsulated inside the film and may be less prone to a rapid degradation by the laser beam because of the thermal protection of the surrounding matrix. ^[3] In the film the absorber molecules would also be more evenly dispersed than in the in deposits. The absorber deposited from liquid inks will form crystal particulates containing an agglomeration of molecules. Only the outer molecules of the agglomerate are likely to be effective in laser absorption and local degradation of the absorber is more likely, as there would be less thermal protection from surrounding material.

The procedures used in the deposition process selector software are successful at predicting the amount of absorber required to give a successful weld in cast PMMA. Application to other polymers has been studied, but only over a few welds in some cases, and then not with the aim of finding the limiting processing conditions for parent material failure. The results are compatible with the limited information that is available on laser welding of those materials, but the software is expected to be less accurate in use for materials other than cast PMMA.

10. Conclusions

This paper has compared four different applicator methods to deposit absorber for the Clearweld process when welding PMMA. From an understanding of the effects of absorber deposition method and processing conditions, a software model has been developed and the following conclusions are drawn:

• When comparing the absorber deposition methods (ink jet, needle, spray and thin film), all were suitable for the Clearweld process to make strong lap welds that failed in the parent material. The limited flow rates available for ink jet deposits result in high applied laser energy requirements compared to spray and needle deposits. The different dispersion of the absorber in film leads to greater efficiencies in energy dissipation compared to the deposits from liquids. Other practical considerations define the selection of the deposition technique.
• Processing windows have been established experimentally in terms of the absorber amount, welding conditions, ink type and deposition methods. It has been shown that weld strength has some correlation with HAZ depth, making HAZ depth an indicator of weld quality.
• A computer model has been developed to predict the amount of absorber required to make a strong weld using any of the fluid deposition methods. The model has been verified with the experimentally available data and found to fit well. This has been developed into software to give guidance on process procedure and absorber requirements given inputs of material type, weld dimensions and process speed. Verification of the model has been successful in a number of test methods.

11. References

1. Jones I A, Taylor N S, Sallavanti R, Griffiths J: 'Use of Infrared Dyes for Transmission Laser Welding of Plastics'. proc. SPE ANTEC 2000 conference, May 7-11 2000, Orlando, Florida. Vol. 1, pp1166-1170.
2. Jones I A, Wise R J: 'Welding Method', Patent WO 00/20157, 1 Oct 1998.
3. King T A: 'Lasers and Ultra-structure Processing', Chemical Processing of Advanced Materials, Ch.90 pp997-1019 ed. Hench L L and West J K, pub. John Wiley and Sons, 1992.

Meet the authors:

Ian Jones joined TWI in 1989 after a degree in Materials Science at the University of Cambridge. He has worked on a wide range of laser materials processing applications, including high power laser welding of metalsand non-metals. In particular, laser welding of aluminium for light-weight vehicles, guidelines for laser welding structural steel for ship construction and a fundamental study into transmission laser welding of plastics. Most recentlyhe has invented and developed the Clearweld ^® process, which can be used to join plastic parts and fabrics without altering the colour or appearance of the product.

Sam Rostami has over 18 years experience in physics and physical chemistry of polymers, polymer surfaces, interfaces and adhesion. He joined TWI in 2001. Previously he worked for ICI Advanced Materials and AcrylicsR&Dmp;D Departments on all aspects of polymers, blends and composites, He has wealth of experienced in fundamental research, new product development and characterisation of multi-component materials. He is co-editor and author of abook on 'Multi-component Polymeric Systems', author five chapters in technical books, over fifty published technical papers and eight patents.