# Would switching a bulb on and off each second extend it's life?

09 May 2011

## Question

Would switching a bulb on and off each second extend it's life?

Dave - There's at least two different failure modes for a light bulb filament. One of them is essentially evaporation. You've got a very thin coil of wire sitting there at 2,000 degrees centigrade, and although tungsten has a very high melting point, some of it is still going to be evaporating away all the time. There is a little bit of gas in there which slows it down, but eventually the filament is going to evaporate away to the point that there's a gap. That gap gets hotter and hotter and evaporates quicker and quicker until eventually it fails. So if you're turning the light bulb on and off, then that mode of failure is going to be a lot slower because it's only on half the time and most of that time it's either heating up or cooling down. So that mode of failure will probably take at least two or three times as long before it fails on average if you turn the light bulb on and off. But there's another mode of failure, which is just the wire breaking. If you're repeatedly heating something up and then cooling it down again then you produce all sorts of physical stresses and strains on it; as it heats up it expands and stretches, then as it cools down again it shrinks, so you'll get metal fatigue. I think the far bigger effect from turning the bulb on and off will be the wire breaking due to metal fatigue and repeated stretching, and it will break far sooner than it would do normally.

Chris - Can I ask you a spin-off question about this, Dave? If you are using AC current, where it goes 'plus', for the sake of argument, then down through zero into 'minus', and then back up into 'plus', it means that the bulb is effectively flicking on and off. So with 50 cycles per second, that's 50Hz AC, it means the bulb is flicking on and off 100 times a second. So does that knacker a bulb more than if you ran a steady current like DC through it?

Dave - It will have a slight effect, but the time it takes for the bulb filament to cool down is a lot more than 1/50th of a second, it's probably around 1/10th of a second. So the actual temperature of the bulb isn't fluctuating hugely due to the AC, so it's approximately the same as DC. Because the filament is effectively an electromagnet there will be some small extra forces acting with AC, though. As the electricity goes through the coil it's going to produce a magnetic field, and that will apply some slight forces and vibration on it, so probably AC is slightly worse for it but not by very much, because the heating up and cooling down time is so much longer than the AC time.