Resistance seam welding of exhausts

Exhaust gas cleaning devices on road vehicles are increasingly popular. Road users have been made more aware of the environmental dangers of pumping lead and other harmful combustion products into the atmosphere. European countries are introducing laws requiring all new vehicles to have catalytic converters.

Converters resemble traditional exhaust silencers in configuration but contain a honeycomb brick structure that uses platinum to catalyse removal of the harmful constituents of exhaust fumes.

TWI has been involved from design to completed production of such units. It has largely investigated the feasibility of using robots to assemble the components but has also been involved with design, laser applications, resistance welding, arc welding and even use of plastics. Most recently, fatigue design and non-destructive testing have also come under study.

Recent work at TWI for two Industrial Member company exhaust suppliers has explored efficient ways of sealing stainless steel skins by resistance seam welding. TWI has also studied the robotic welding of circular pipe and thick flanges to the ends of the boxes.

Seam welding, however, does not always provide the ideal solution. The process requires a flange about 15mm wide on which to make the weld. Historically, the exhaust system is one of the last parts of a vehicle to be designed and it frequently has to fit around or inside the previously designed floor pan. It can prove more economical on space to use a peripheral MIG weld.

The converters are usually designed for a five-year life. Beyond this time the platinum and stainless steel may well be in good order, but fatigue of such a component is the limiting factor.

Much of TWI's work has involved studying manufacture of these units. Are they designed correctly? Can you robot weld them? Can you jig them? Can you fixture them? ...and when they are jigged and fixtured, can you make them fast enough? If you can, is the quality good enough and how do you check it? TWI staff can advise on a long list of variables.