Welding bosses to polypropylene pipework

TWI Bulletin, May/June 1990

After obtaining a degree in Materials Science, and a PhD for work on the metallurgy of submerged-arc weld metals in constructional steels, Martin joined TWI in 1979. Since then he has been responsible for many developments in laser welding (particularly for joining thick sections) and plastics and polymer composites joining. He has been closely involved in process selection and research projects in industries such as power generation, structural engineering, process plant, aerospace and mass production.

Martin is currently Head of the new Plastics Joining Department, responsible for all work carried out at the Institute on development of new and improved joining techniques for plastics and polymer composite materials. The department maintains facilities and a level of expertise sufficient to enable member companies to identify, test, develop and implement the correct joining process for their applications and to help solve production problems.

Martin Watson describes an investigation undertaken on behalf of an Industrial Member company to optimise the welding procedures for joining bosses on to industrial pipework.

Polypropylene pipework is widely used in industrial plant, particularly for water at low temperatures (<100°C). A wide range of pipe diameters for several different pressure ratings is available, with fittings such as bends, branches, size reducers, valves and so on. Joining is usually by mechanical means if later disassembly may be needed, or by welding if a permanent joint is acceptable. The usual welding technique is heated tool, employed either for socket joints or straight butt joints.

Industrial pipework is often complex and sometimes the optimum design cannot be achieved using commercially-available fittings, in which case special-purpose fittings must be fabricated. A requirement to do this arose recently in the cooling water pipework for some large electrical systems at GEC Electrical Projects Ltd, and TWI was asked to help implement the non-standard fittings.

The requirement was for a number of 20mm OD branches to be welded to a 63mm OD supply pipe. Commercial fittings were available, but their use was precluded by the close spacing required between the branches.

Fabrication approach

Relatively few assemblies were needed, so the approach had to be as simple as possible. It was decided that the branches could be made by welding 20mm OD bosses directly into holes drilled in the 63mm OD pipe. Welding could be achieved by the heated tool socket welding process in which a shaped heated tool is used to heat the outside of the pipe and the inside of the socket. In this case, however, the socket would be provided by the hole in the 63mm OD pipe.

In view of the intended working conditions it was decided that a short-term pressure test to 13.6 bar at room temperature would indicate acceptable joint properties. The primary aim of the investigation undertaken at TWI was to establish a welding procedure for the heated tool process capable of meeting this pressure test. Trial welds were made at different conditions ( eg temperature and time) and were pressure tested. Care was taken to ensure that the procedures could be easily reproduced in the production environment and did not require expensive or complex equipment.

Experimental

The pipe was manufactured by the George Fischer company to DIN 8077/78 in Dekaprop polypropylene type 2. Dimensions were nominally 63mm OD and 6.8mm wall thickness. Two kinds of boss were supplied as-machined from solid rod: the first had parallel sides and was 20mm diameter; the second had a 7mm long parallel portion 20mm diameter ending in a step which was intended to provide a positive location on the outside of the pipe. Bosses with blind holes were used in the trials to make pressure testing easier.

The welding equipment developed in the project was based on a standard manual heating tool for socket fusion of polyethylene pipes, modified to allow different temperatures to be used. The standard heating element had male and female tool faces bolted on to it to heat the outside of the boss and the inside surface of the hole in the 63mm OD pipe.

Initial trials showed that fully manual operation did not provide accurate enough alignment between the boss and the hole, as molten material was squeezed out unevenly around the joint circumference, resulting in a poor weld. So it was decided that a simple jig - in practice, an electric drill stand - should be used to maintain alignment. In operation, the 63mm pipe was clamped in shell clamps mounted on the drill stand bed. The hole was drilled using an electric drill. Then the drill was removed, and the boss mounted in the collar of the drill stand using a purpose-made adaptor ring. This ensured accurate alignment between the boss and the hole.

Using this simple equipment, the welding pressure was applied manually via the drill stand and the heating tool was inserted and removed by hand. The welding equipment and a typical sequence of operations are illustrated in Figures 1-4.

All the welds were examined visually to determine whether heating was uniform around the weld and whether an adequate bead of squeezed-out material had formed. The appearance of a typical sound weld is illustrated in Figure 5.

Pressure testing was carried out using the rig illustrated in Figure 6. The ends of the pipe sample were sealed by plates containing inset O rings and drawn together with lengths of studding. The sample was internally pressurised with water at room temperature (20-23°C), either at a constant pressure of 13.6bar (200psi) (in which case pressure was held for up to 30min), or at increasing pressure until failure.

Selected welds were cut out and sectioned by microtome for viewing with an optical microscope with crossed polars.

Welding trials

Initially, a small number of welds were made to investigate the operation of the equipment and the testing procedures. A number of factors relevant to the satisfactory operation of the process were also investigated.

1. It was difficult to avoid inserting the parallel-sided bosses too far into the pipe wall. This resulted in poor welds because cold material in the boss was adjacent to the pipe wall. To overcome this a mechanical stop in the drill stand was used. The problem did not arise with the stepped bosses.
2. Firm pressure was required at the beginning of the heating time to ensure that the heating tool was completely home in the pipe wall and the boss was being heated to the correct depth. Once this was achieved, only gentle pressure was required.
3. In the welding (consolidation) part of the operation, firm pressure was required. This was held for 30s, by which time the joint was sufficiently cool for the welded sample to be removed.
4. The change over time, ie the time between the end of the heating phase and the welding phase during which the heating tool is being removed, must be kept short. Two operators can be used to achieve the shortest time, but this was felt to be inappropriate. With one operator a change-over time of approximately 5s was possible and all the work reported here was with a 5s change over time.

A series of trials was made to optimise the welding parameters: specifically the heating phase parameters were investigated as these are critical to the achievement of sound joints in heated tool welding. All the lessons learnt in the initial investigation were incorporated in these trials.

The welding temperature recommended for polypropylene is usually in the range 190-220°C, depending on wall thickness. Two temperatures, 205°C and 220°C, were investigated here. The standard temperature for pipe of the size in question is 205°C, and this is also widely used for polyethylene pipe, hence equipment for operation at 205°C is readily available.

Heating times were varied between 2 and 30s. Welds were evaluated in terms of their physical appearance and on the results of the 13.6bar (200psi) pressure test.

For the parallel sided boss (see Table 1) there appeared to be a very wide range of acceptable welding conditions when a tool temperature of 205°C was used. Heating times of 5s or more gave satisfactory strength. Times of 25s or greater gave rise to considerable volumes of molten material expelled from the weld and were therefore rejected on grounds of appearance, although properties were not reduced. With a tool temperature of 220°C the melt was more difficult to control and an acceptable condition with both good strength and satisfactory appearance was not found.

Table 1: Results of optimisation trials on the parallel-sided boss pressurised at 13.6bar (200psi) for 30min

For the stepped boss (see Table 2) the results were very similar. Again there was a wide range of acceptable heating times at 205°C with an optimum of 15s. Samples of each boss were pressure tested to fracture and the stepped boss failed at 40% higher pressure.

Table 2: Results of optimisation trials on the stepped boss pressurised at 13.6bar (200psi) for 30 min.

Based on the above, ten welds were made at the best welding conditions studied, ie 205°C temperature, 15s heating time, and with the stepped boss. Five were tested at 13.6bar for 30min. None of these failed. The others were tested to failure. All failed at pressures in the range 26.9-34.5bar (390-500psi).

Microtome sections from a weld made with a parallel sided boss are shown in Figure 7. A sound weld of even width throughout the wall thickness was formed.

Overall there was less weld area with the parallel sided boss than the stepped boss, which probably accounted for the slightly lower joint strength achieved with this geometry.

Recommendations

This project assessed methods for welding bosses to polypropylene pipe with a non-standard joint geometry. In view of the numbers required, the heated tool welding process was thought to be the most appropriate technique for joining. Trials on this process have shown that:

• The heated-tool welding technique can produce acceptable joints with a good tolerance to welding condition variations using inexpensive, manually operated equipment.
• A stepped boss is preferred as it results in a stronger joint and position stops in the welding equipment are not needed.
• Care must be taken to ensure good alignment between the hole in the pipe and the boss. This was achieved by mounting the boss to be welded in a drill stand and using this to provide the welding pressure.
• Satisfactory welding conditions were found to be a heating tool temperature of 205°C, a heating time of 15s and a cooling time under pressure of 30s. Change-over time must be kept short; typically, 5s.

Acknowledgments

Thanks are due to Ian Froment and Steve Willis, who conducted the experimental work, and to GEC Electrical Projects Ltd for permission to publish the results in this article.