Can material temperature change be used to predict mechanical failure?

  • 2 Replies

0 Members and 1 Guest are viewing this topic.


Paul Anderson

  • Guest
Paul Anderson  asked the Naked Scientists:
Hi Chris and team,

If I break a twig in half by bending it, can I assume the surface of the inner side of the bend is slightly hotter than the surface of the outer bend? Could the size of the twig and the temperature differential be used in some way to predict when the twig would snap?

In Auckland they are currently debating about the life of the Auckland Harbour Bridge and the Nippon clip-ons. If stress in steel can be measured, could this be applied to the surface of the earth and be used to predict earthquakes?

I am trying to think of some way of using satellite surveillance to predict earthquakes.


What do you think?


Offline DoctorBeaver

  • Naked Science Forum GOD!
  • *******
  • 12656
  • A stitch in time would have confused Einstein.
    • View Profile
In a previous life I was a consultant/analyst for a maintenance management software company. Temperature difference was 1 of the methods we employed to monitor potential component failure.

I haven't a clue about applying it to earthquakes.
Fledgling science site at needs members and original articles. If you can help, please join.


Offline graham.d

  • Neilep Level Member
  • ******
  • 2208
    • View Profile
Some failure mechanisms occur more rapidly at higher temperatures. Often these mechanisms can be ultimately responsible for the end-of-life failure of a component. Testing components at elevated temperatures is therefore often used as a way of predicting the life of a component. It means that testing a component, for say 6 weeks, at a particular high temperature, can be used as a way of predicting the life of the component which may be (say) 20 years. It requires a lot of understanding of the failure mechanisms that are likely to occur and how they accelerate with temperature (look up "activation energy" for example or "electromigration" - a major failure mechanism in electronic components).

Having written this, I don't think it really addresses your problem though. I suspect it is, maybe, related to the sort of analysis that is done on the bridge. It is quite common to cover bridges in strain gauges nowadays. When steel fails under stress must have been studied and analysed for many years.

I suspect that predicting eathquakes from visual satellite surveillance alone may be not be possible, but combining this with information about surface movements from gps positioned sensors on the ground and knowledge of the geology (even just knowing where to look) would be promising. Is it known whether eathquakes are associated with temperature changes on the surface?