Skip to content
Skip to main content

About this free course

Download this course

Share this free course

Engineering: The challenge of temperature
Engineering: The challenge of temperature

Start this free course now. Just create an account and sign in. Enrol and complete the course for a free statement of participation or digital badge if available.

6.2 Refining the specification

The ideas for the boiler cut-out switch can now be based on some real knowledge about temperature effects. You are now ready to tackle the next exercise.

Exercise 7

List four temperature-dependent changes in material properties that could be exploited in the automatic cut-out switch for an electric water-boiler.


Here are my suggestions:

  • Expansion of a solid

  • Expansion of a gas

  • Phase change in a solid

  • Phase change in the water

  • Change in electrical conductivity of a metal

  • Change in electrical conductivity of a thermistor.

Table 10 Examples of gradual, accelerating and sudden changes
ServiceOpportunityElectrical resistance thermometerHeat-curable gluesOnset of ferro-electric properties
Bimetal switchesEnhanced performance of detergents in hotter waterCan leave cooking to simmer
ChallengeCrazed glaze on crockeryHigh-temperature limit on semiconductor performanceSome permanent magnet materials demagnetise at modest temperatures
Need to compensate timer in clocks and watchesExcessive corrosion of high-temperature componentsWater pipes burst through freezing
ProcessOpportunityDensity variation provides convective mixing in castingsTemperature control of workability of glassChemical reaction selection by use of critical temperature
Metal tyres and bearing sleeves can be shrink-fitted to wheels/shaftsSintering of powders to continuous solid (a route to near net shape forming)Melting allows casting processes
ChallengeContinuous welded railway lines may buckle in extreme heat or crack in extreme coldOver-ageing of precipitation-hardened alloysUpset metallurgy in the heat-affected zones of welds
Continued diffusion of previously implanted dopants in subsequent processing of microcircuit chipsCracks in porcelain due to crystal transition in quartz

In the specific case of the water-boiler we have been looking at, what is wanted is a cut-out switch (not a thermostat), so our specification is for the switch to operate at a temperature close to the boiling point of water, and to be re-settable manually once triggered. This overall cut-out function requires a kind of irreversibility to be built into the device mechanism. What I mean is that, having reached the operating temperature, the device has to flip into a state where it has caused the current to be switched off, such that, without a manual reset, it will not start to pass current again once the water has started to cool.

All the thermal effects listed in my answer to Exercise 7 are fully reversible – as we would want them to be because the switch is to be reused many times over. For example, when a metal is allowed to cool again after being heated to a given temperature, its electrical resistivity will be the same as it was at that temperature before it was heated (otherwise a platinum resistance thermometer would be pretty useless). This means that to make the device that is wanted, some kind of mechanism has to be introduced to do the flipping, that will not flip back by itself until helped by the user. Another good reason this is needed in this case, where a flow of electric current is being interrupted, is that the switch-off should be done as sharply as possible to avoid arcs and sparks.

Exercise 8

For the four thermal effects you selected in the last exercise, describe how each could be used in a thermal cut-out switch operating at or near the boiling temperature of water.


Here are my suggestions, based on my answer to Exercise 7. They are not necessarily the best ways of using each effect. The fun is in trying to think up a really neat way of getting the result you want.

  • Expansion of a solid: Two metals with different expansion coefficients are joined together to form a bimetallic strip. This bends as the temperature changes. At the right temperature, the strip has bent far enough to act on an over-centre latch to open a switch.

  • Expansion of a gas: A sealed capsule of gas exerts a temperature-dependent pressure on a diaphragm, which has been formed to be domed slightly inwards. At the right temperature this dome snaps outwards, opening the switch.

  • Phase change in a solid (shape memory alloy): This works in the same way as the bimetallic strip, except that the deformation of the shape memory is caused by a phase change, not a linear expansion, and so will happen all at once at a sharply defined temperature.

  • Phase change in the water: This could be a microphone picking up the sound of the steam bubbles as the water boils. This is detected in an electronic circuit that opens either a transistor switch or an electromechanical relay.

  • Change in electrical conductivity of a metal: This could be detected by an electronic circuit which then operates a switch, as in the microphone system above.

  • Change in electrical conductivity of a thermistor: This would be similar to the resistor-based system, but with the difference that the sensor output would be an exponential function of temperature.

The next step of selecting and modelling the most suitable solution has to be taken with reference to the specification and other considerations, such as whether the company wants to manufacture the devices itself, perhaps making use of equipment it already has and techniques it already knows.

One part of the functional specification that could be of importance to the user is the repeatability with which the device trips at the set-point temperature. This might be wanted where the water is to be used in a temperature-dependent process, and product consistency is a requirement, or if the water has to have reached a temperature very close to boiling. On the other hand, there may be a market for switches that can be set to trigger at any temperature chosen by the user.


Identify a type of thermal effect (gradual, accelerating, or sudden) likely to give:

  • (a) a repeatable switching temperature;

  • (b) an adjustable switching temperature.

For each, explain why.


  • Sudden: These effects produce dramatic changes in material properties over very small temperature ranges, determined by the physical process itself.

  • Gradual: These effects are linear or nearly linear changes in the properties. The device can be arranged to be adjustable by having a movable set-point.

Exercise 9

Table 11 lists the candidate solutions from my answer to the previous exercise, against some of the criteria for the water-boiler listed in the example at the beginning of this course. If you had some candidate solutions of your own, add these to the table. Devise a rating or scoring system, and obtain a rank-ordering of the candidate solutions.

Table 11
Gas capsule
Shape memory
Metal resistance
Thermistor resistance


Here is a simple rating scheme in which I have given scores of 1, 2 or 3 and simply added together the scores for each candidate. It takes no account of the relative importance of each criterion. If it had, the rating scheme would have looked more like the example in OpenLearn course T207_1 The engineer as problem-solver: the nature of problems of choosing a TV set, where each score against a particular criterion is multiplied by an 'importance factor'. So the answer here rates shape memory as the winner, ahead of bimetallic, whereas the bimetallic solution is in fact the commonest. This is probably because in most real-life applications, accuracy is less important than cost.

Table 12
AccuracyReliabilityRuggednessCheapnessScore (higher=better)
Gas capsule12227
Shape memory333211
Metal resistance322310
Thermistor resistance322310

You will have realised when trying the last exercise that much of the information you supplied to perform the ranking task was largely guesswork – for instance, off the top of your head you can't be really sure whether the gas capsule approach will be cheaper or more expensive than the shape memory solution. Another example is the comparison of the accuracy of the thermistor and the metal. At or near the boiling point of water, the percentage change of resistance of the metal per degree centigrade is the same as at lower temperatures – at least that's what our linear model presumes. But for the thermistor, the change is an accelerating one, and its rate of change at a given temperature depends both on the value of the activation energy of the electrical conduction process, Ea, and on the temperature itself. Without doing the calculations for specific materials, you don't know which approach will provide the greater accuracy.

Therefore we can say that to make a good decision, more work needs to be done. But at least now, the tasks are very specific fact-finding or mathematical modelling exercises. It leads me on to an observation that is true of any solution-finding undertaking. Whatever the details of the decision-making process used to select the preferred solution from the candidates presented, none of it is of much value unless the information that informs that process is of the highest possible quality. For the engineer, that information is:

  • the identification and expression of the need

  • the detailed specification

  • good-quality ideas, based on a sound understanding of the underlying principles

  • facts and figures (accurate and reliable data) relevant to the selection criteria.