Polypropylene welds - orientation evaluation

TWI Bulletin, July/August 1992

Sheila Stevens is a Principal Research Chemist in TWI's Chemical Laboratory.

Her work at TWI has involved analysis of steel weld metal, fluxes, wire and wrought products, specialising in modern instrumental methods. This has extended to study of theoretical aspects of optical emission spectroscopy.

Sheila has also worked on projects involving sampling and analysis of welding fume and gases, including local extraction ventilation, determination of free nitrogen by hydrogen hot extraction, and extraction and analysis of gases from pores in weld metal. The last five years have seen a change of emphasis in her work which is now directed towards use of chemical and physical techniques for characterisation of structural properties in weld areas in thermoplastics.

Polarised Fourier transform infrared-microspectrometry can be used to determine molecular orientation in polypropylene welds, and thus gain an understanding of effects of welding on mechanical properties. Sheila Stevens reports.

Polypropylene (PP) is a semicrystalline thermoplastic that is widely used for water distribution pipes, chemical plant, and injection moulded parts for a number of industries, as a result of its attractive service properties. The mechanical properties of semi-crystalline polymers are influenced by a number of factors, including crystallinity and orientation of the polymer chains. It has been reported that in PP, tensile strength increases with increasing crystallinity ^[1] and orientation ^[2] while impact strength decreases. ^[1]

Evaluation of molecular changes induced in a polymer by the welding process is a relatively new research area, as the often small size of the welds necessitates highly specialised analytical techniques. One such is Fourier transform infrared (FTIR)-microspectrometry, which makes possible analysis of samples as small as 10 x 10µm square. This article describes use of polarised FTIR microspectrometry to determine the orientation of the polymer chain in PP welds, and simultaneously to measure the helical content, which can be correlated with crystallinity.

Experimental

Hot plate welds were made using George Fischer Dekaprop Type 2 PP pipe, which is randomly copolymerised with ethylene. The pipe was 180mm diameter with a wall thickness of nominally 10mm. Welding conditions are given in the Table: weld HP5 was made using the conditions recommended by the material manufacturer, while HP8 was welded with a very short heating time to produce a 'poor' joint. Photomicrographs of the welds are shown in Fig.l.

Hot plate welding conditions

Orientation measurements

The FTIR technique uses the principle that, when IR radiation is passed through a polymer sample, some frequencies (wave numbers) are absorbed and others transmitted. The transitions involved in IR absorption are associated with vibrational changes within the molecule, different bonds absorbing different frequencies. The detailed theory behind polarised FTIR measurements has been discussed previously, ^[3] and only the necessary equations are given below:

The dichroic ratio, D, can be regarded as a measure of the degree of preferred orientation in polymers. It is given by:

where

and

The orientation function, f, gives the fraction of perfectly orientated (with respect to the draw direction) chains and for uniaxial orientation is related to D by:

where  

and α is the angle between the transition moment of the vibration of the group and the axis of the polymer chain.

A value of zero for f denotes random orientation, and values of +1 and -0.5 mean that all the polymer chains are orientated parallel or perpendicular to the draw direction, respectively. If separate absorbance bands are unavailable for the crystalline and amorphous phases, the amorphous orientation function, f _am, can be calculated from the crystalline, f _c and average, f _av, orientation functions:

where X _c is the fraction of the crystalline phase.

For PP, the 998cm ^-1 crystalline band, and the 973 ^-1 band (which arises from both the amorphous and crystalline phases) can be used. A value of α of 18° has been assigned to both bands, and f _am can be calculated from equation [4] after determining the helical fraction, H, which equates with X _c, as follows:

where A ₉₉₈ and A ₉₇₃ are the absorbances of the 998cm ^-1 and 973cm ^-1 bands, respectively.

As absorbance readings can be affected by the presence of orientation in the sample, the values used in equation [5] are 'normalised':

Analytical tests

Sampling of weldments

A bright rotary retracting microtome with a tungsten carbide tipped steel blade, blade angle 25°, was used to obtain 50µm thick transverse sections from the weldments. A diagram of the sampling scheme is given in Fig.2.

FTIR measurements

Spectra were recorded parallel and perpendicular to the draw direction (defined in Fig.2) using a Mattson Polaris FTIR spectrometer with FIRST software, a Spectra Tech IR-PLAN microscope with a dedicated 0.25mm mercury cadmium telluride detector, and a wire grid polariser consisting of aluminium deposited on a KRS-5 substrate.

Reproducibility and precision of the technique were examined by appropriate repeat measurements and found to be suitable for quantitative measurements of orientation and helical content: details are given in Ref. ^[3] .

Measurements on weldments were carried out along an axis perpendicular to the weld centreline at the centre of the parent material thickness using an analysis area of 188 X 50µm (see Fig.2).

Results

The results of measurements on the weldments are given in Fig.3: the vertical solid lines denote the edges of the weld, and the broken vertical lines denote the outer edges of the heat affected zones (HAZs). For HP8 the weld edge and the HAZ are virtually coincident, as shown by the micrograph ( Fig.1b).

Orientation

The results in Fig.3 show that the trend is for the orientation functions to go from negative in the pipe to positive within the weld, so that the alignment of the molecules is changing from the axis of the pipe to the flow direction of the weld. The behaviour of the molecules in the amorphous regions is similar for both welds, but this is not true for molecules in the crystalline regions.

In the 'good' weld HP5, f _am is around zero within the HAZs (indicating no preferred orientation) and rises to 0.02 in the weld, with a minimum of zero at the centreline. Within weld HP8, f _am is typically zero, with a minimum of -0.02 at the centreline, which is the same level as in the parent pipe.

In the crystalline regions in HP5, f _c increases to 0.02 to 0.04 in the HAZs, and to 0.10 in the weld, but drops at the weld centreline to 0.02. In weld HP8, made with a shorter heating time, f _c exhibits quite different behaviour; it only rises to 0.01-0.03 and drops to -0.08 at the centreline, close to the pipe figure.

Helical content

For both welds, the parent pipe helical content increases to around 59% just outside the HAZs (this effect does not correlate with any visible microstructural changes) and then falls within the welds to 57.5-59% in HP5, and 56.5-58% in HP8. At the weld centreline, corresponding to the central dark lines, the helical contents are 59.2% and 57.5% for HP5 and HP8, respectively. The lower helical content in HP8 indicates that the weld thermal cycle produced less favourable crystallisation conditions, perhaps because of reduced heating time (23 sec compared with 200 sec for HP5) or of a faster cooling rate.

Discussion

It is apparent that molecular orientation in the amorphous regions is much less affected by the weld thermal cycle than in the crystalline regions, presumably as a result of the higher degree of disorder in the amorphous phase. It is also noteworthy that the maxi mum orientation does not occur within the HAZs, where optical microscopy indicates flow lines ( Fig.1a and b). However, the maximum flow would be expected to occur in the weld, where the material has been fully molten and pushed outwards, with simultaneous cooling and crystallisation with the polymer chains aligned with the flow direction. On this basis, the visible HAZ flow lines are merely deformed spherulites containing fewer aligned polymer chains. Neither does the maximum orientation coincide with the maximum helical contents (see below) just outside the HAZs, suggesting that the latter result from an annealing effect rather than orientation induced crystallisation.

In this context, it is worth noting that the maximum helical content at the weld centreline in HP5 coincides with a minimum orientation function. However, the orientation in the rest of the weld region could aid crystallisation since lower helical contents and orientation levels are found in HP8.

The dramatic difference in centreline f _c for the two welds suggests that in HP8 the material at the interface did not flow, perhaps because it was too cold, and was not squeezed out (although the material behind it was); certainly the weld bead is much smaller than in HP5. If the material had not been squeezed out, it means that the weld line would incorporate any contaminants originally on the pipe surface, or oxidation products.

It is desirable to be able to correlate the measured molecular changes with mechanical property data, but the tensile strengths of HP5 (24.9MPa) and HP8 (22.4MPa) were similar to that of the parent pipe (22.0MPa). But, in the field, bead (or flash) size is used as a visual indication of weld quality and weld HP8 would almost certainly have been rejected as a 'poor' weld. Thus, in this instance, the weld orientation behaviour might be a better indicator of weld integrity than tensile strength.

Conclusions

• Changes in molecular orientation were dependent on the welding conditions and tended to be parallel to the flow direction of the weld.
• The crystalline phase was more easily orientated as a result of welding than the amorphous phase.
• Measured orientation changes could not be related to visible microstructural changes. Also, there was no change in visible microstructure immediately adjacent to the welds, despite the local increase in helical content.

FTIR spectrometry is a very powerful technique, especially when combined with an IR microscope. The application described in this article is one of many potential uses, and some of these are detailed below. The FTIR facility is available as a service to TWI Industrial Members, and anyone interested is invited to contact the author.

Applications of FTIR

Typical applications of FTIR include the following:

Polymers or welds

• Composition
• Molecular structure
• Molecular degradation mechanisms
• Molecular orientation
• Crystallinity
• Inclusion analysis

Other materials

• Adhesives
• Polymer coatings on metals
• Metal oxide films
• Corrosion products on metals
• Contaminants in microjoining
• Painted surfaces

References