Electrofusion welding - software validation trials prove valuable

TWI Bulletin, September - October 1998

Felicity Chipperfield is a senior project engineer in the Advanced Materials and Processes Department at TWI, responsible for planning and leading projects in the field of plastics joining.

Since joining TWI in April 1995, she has managed projects aimed at joining plastics for several industries, including the medical, automotive and offshore industries. She has used a wide variety of joining techniques including microwave welding and the more common butt fusion and electrofusion processes.

As well as these shorter-term activities Felicity has managed the core research programme focussed on the structural integrity of plastic welds.

Her previous experiences include research into electrofusion welding of polyethylene pipes, and working on the Harrier Jump Jet.

Research has been carried out on the electrofusion welding process for joining polyethylene (PE) pipe. As Felicity Chipperfield reports this work has been carried out to verify an electrofusion welding simulation and analysis software package called Exefusion, which is available for use by Industrial Member companies. ^[1]

Thermoplastic pipes are widely used in the distribution and process industries for the transportation of domestic gas, water and chemicals. Low cost, durability, flexibility, corrosion resistance and resistance to chemical attack are the most important advantages that polymer materials such as polyethylene (PE) and polypropylene offer. The manufacture of polymer materials, pipes, couplers and assorted fittings, and associated process and installation equipment is a major UK and European industry.

Polyethylene pipe systems have quickly become one of the preferred material choices for the water and gas utilities. Polyethylene is experiencing a rapid growth and acceptance within these industries, as newer grades are introduced which increase the scope and pressure rating of pipeline networks. Applications are growing in Europe and world-wide, driven by the ability to install a fully welded and leakproof pipe system.

Since 1969, when PE pipe systems were first used to distribute gas in the UK, the majority of joints have been constructed using heated tool fusion techniques. The three original techniques of socket, saddle and butt fusion welding have been carried out successfully, but at the cost of a high level of operator involvement often working in a difficult working environment. These methods have also required the availability of a considerable amount of ancillary tooling and equipment. Whilst all of the methods are operationally satisfactory, an alternative technique, that of electro-fusion, has been developed; this is perceived to be less susceptible to human error and a more efficient process.

The electrofusion welding technique

This technique permits joining, of pre-assembled pipes and fittings, to be carried out with minimum equipment. It also offers a number of practical advantages to the installer. It is easy to use for repairs, where the available space and pipe movement is limited, and allows fusion of a wide range of polyethylene resins.

Design of electrofusion fittings

The electrofusion welding process involves the use of a fitting. This is basically an outer sleeve into which the two pipe ends to be connected fit. An internal stop prevents the pipe ends from meeting, see Fig.1. Fusion indicators are designed into the fitting, such that when sufficient melt pressure has been generated the indicators will protrude - giving the operator a visual indication that the welding process has been carried out successfully. If the indicators fail to protrude, then the welded fitting should be cut out from the pipeline, and a new fitting should be welded in place.

Preparation of joints

It is widely acknowledged that, in order to establish a consistent and structurally sound joint, it is necessary to follow a strict preparation procedure. If the appropriate techniques are followed, contamination and disturbance effects that might inhibit the fusion mechanism will be minimised.

For successful joining of pipes at least three important pipe preparation stages must be followed. Firstly the pipe ends must be squarely finished. This ensures that the central cold zones function to contain the melt.

Secondly, the pipe surfaces to be joined must be properly scraped to reveal uncontaminated material. With the electrofusion joining process there is little or no relative movement of the pipe to the coupler. Therefore any dirt or contamination on the pipe surface is retained at the joint interface, which can significantly weaken the strength of the joint.

Finally, the pipe and fitting should be clamped during welding effectively to eliminate relative movement. This ensures that the molten polymer is contained at the fusion interface, allowing the development of a strong joint.

Stages in electrofusion welding

The joining process during electrofusion welding can be divided into three stages; I - initial heating and fitting expansion, II - heat soaking to create the joint and finally III - joint cooling. The duration of stages I and II is commonly termed 'fusion time'.

Experimental work

Materials

All experimental investigations were undertaken using a variety of grades of PE, currently in commercial use by pipe and fitting manufacturers. The pipes were supplied by various manufacturers with a nominal outside diameter of 125mm and a standard dimension ratio (SDR, the specified OD divided by the specified minimum wall thickness of the pipe) of either 11 or 17.6. Various manufacturer supplied fittings with a nominal inside diameter of 125mm, see Table 1.

Table 1: Materials used during the trials

Equipment

Welding trials were carried out using an MCA Calder S.A.M. 6 electrofusion control box, see Fig.2.

Welding trials

'Fusion time'
In this set of trials, pipe supplied by Manufacturer B made from Fina 3802 blue resin was used. The electrofusion couplers were moulded by Manufacturer A from Fina 3802 yellow resin. The standard 'fusion time' for this coupler is 200 seconds. Eight welded joints were produced using times of between 200 and 480 seconds, with 40-second increments.

Surface conditions
This section of work was carried out in two parts. In the first study, various standard pipe-scraping tools were used to prepare the pipe surface. In the second study, the pipes were scraped and then left for a recorded period before welding (a minimum of 24 hours delay). This latter study allowed airborne particles to contaminate the surface of the pipes and may have also caused a limited degree of oxidation to the PE.

No welds were carried out immediately after scraping, although this is recommended in standard practice. The reason for the delay before welding, in this research programme, was to allow the pipe to stress relieve after scraping.

Electrofusion control box output voltage
This investigation was carried out using electrofusion couplers made from Finathene 3802 yellow resin by Manufacturer C, and Neste 2420 yellow pipe extruded by Manufacturer D. Three welded joints were produced using the standard voltage of 39.5V (as specified in British Gas specification for Electrofusion Control Boxes - where the actual output voltage should be between 39 and 40V). Two non-standard joints were produced using voltages of 37 and 42V.

Manufacturers fitting design
This investigation was carried out to determine whether the various designs of electrofusion fittings had an effect on the welded joint strength. Five different designs were investigated.

The materials for the resistance heating wires for 125mm ID electrofusion fittings were either copper or steel, and the dimensions varied from 0.63 to 1.63mm in diameter.

Four out of the five electrofusion fitting designs had wires that were either fully embedded in the coupler or coated with PE. The other fitting system incorporated bare wires that were visible at the bore of the fitting.

Mechanical testing
Samples were taken from the welded joints and mechanically tested. A double cantilever beam type specimen was used as specified, by the Water Research Council, in standard WIS 4-32-14. ^[2] Testing was carried out, at room temperature, on a tensile testing machine with a cross head speed of 20mm/min. Eight specimens from each half of the pipe/fitting joint were taken for each set of welding conditions. Specimens were peel tested from both the outer cold zone (labelled A) and from the inner cold zone (labelled B).

The values of peak load were read directly from the testing machine chart recorder. 'Peel energy' values were calculated using a weighing technique. The resulting fracture surfaces were classified as either ductile or brittle failure.

Results and discussion

'Fusion time'

For the shortest 'fusion time' of 200 seconds an interface temperature of 145°C was achieved, and the cumulated energy produced was 4.9MJ/m². The interface temperature was measured by embedding a thermocouple at the extremes of the fusion zone within the fitting. The cumulated energy was calculated from monitoring the voltage and current during the actual welding process.

For the intermediate 'fusion time' of 360 seconds, the interface temperature increased to 218°C producing a cumulated energy of 8.3MJ/m². However an interface temperature of 274°C was obtained by the sample welded with the longest 'fusion time', and the cumulated energy in this joint was 10.6MJ/m².

Surface conditions

Manual pipe preparation tools
All the samples failed in a ductile manner when mechanically tested. In 59% of all samples tested, failure occurred through the position of the resistance heating wires. In 39% of samples, the failure site was at the hole drilled in the fitting to facilitate testing.

It was observed that greater values of joint strength were achieved when using the Universal Scraper, and that the lowest values were gained by using either a Harris or Skarsten type scraper. One reason for this maybe that the Universal Scraper has a wide serrated blade, which is rotated around the circumference of the pipe by means of a manually operated ratchet. The depth of pipe surface to be removed can be set, and a uniform layer is removed. However, for the other types of scraper investigated, the process is entirely manual, and therefore quality is operator dependent.

Contamination
There was a significant difference between samples that were scraped and welded with only a delay of 24 hours compared with samples that experienced a delay of 650 hours. Samples that experienced the longer delay had weld strengths of approximately 60% of the joints welded after only a short delay.

Electrofusion control box output voltage

Table 2: A summary of results from the investigation into control box output voltage

Voltage, V	Peel side	Average peak load, N	Peak load	Average peel energy, J	Peel energy
37.0	A	1519	281	147	91
37.0	B	1213	310	112	53
39.5	A	1411	215	112	39
39.5	B	1268	138	128	48
42.0	A	1498	226	163	70
42.0	B	1381	333	159	40

Manufacturers fitting design

The fusion indicators rose satisfactorily on each fitting after welding. In this particular study the greatest peel load and energy values were gained by the fitting design which had wires that were positioned at the bore of the fitting, and were not coated with polymer.

Conclusions

• This practical piece of work was able to validate the simulation and analysis results from the Exefusion software package.
• In this brief study the most important variable that characterised the electrofusion welding process was that of initial surface condition, ie scraping and contamination.
• Increasing the 'fusion time' leads to an improvement in joint strength up to a point, where thermal degradation then takes over.
• Different types of manual scrapers affect mechanical properties. The Universal Scraper was the best out of those tested.
• The time between scraping and welding should be kept to a minimum.
• Electrofusion welded samples, when peel tested from the outer cold zone, give greater joint strengths than samples peel tested from the central cold zone.

Acknowledgements

Bradford University wrote the Exefusion software package, in conjunction with Brunel University. The project was sponsored by TWI, British Gas, WRc, BP Chemicals, Fina Research, Fusion Group plc, Glynwed Plastic International and Stewart & Lloyds Plastics (now merged and trading under the name Durapipe - S&LP), Uponor Aldyl (now trading as Uponor Limited) and Wavin Industrial Products.

References