Wait, What? If you put 10,000 J into a metal block, the block does not necessarily gain 10,000 J.
That sounds wrong at first. The power supply really does transfer electrical energy into the heater. But the heater is not the block, and the block is not isolated from the room. Some energy warms the heater, thermometer or probe. Some leaks into the bench and air. Some leaves by convection and radiation while the experiment is still running.
This is why the specific heat capacity practical is not mainly about substituting into E = mcΔT. It is an experiment in energy accounting: deciding what energy entered the system, what part of the system you are modelling, which losses you are ignoring, and whether the resulting estimate of c deserves to be trusted.
The scientific job
Specific heat capacity c tells us how much energy is required to raise the temperature of 1 kg of a substance by 1 K (or 1 °C temperature difference). In the simplest school model:
E = mcΔT
where E is energy transferred to the material, m is its mass and ΔT is its temperature change.
For electrical heating, the electrical input is often estimated using:
Eelectrical = VIt
or with a joulemeter where available. Combining the idealised relations gives:
c ≈ VIt / (mΔT)
The approximation symbol matters. It reminds you that the experimental model assumes most of the measured electrical energy ends up as thermal energy in the block.
Why this practical often gives a value that is too high
Suppose you calculate c from all the electrical energy supplied, but some of that energy escaped into the room. The block warms less than the ideal model predicts, so ΔT is too small relative to the measured electrical input. Because c is calculated by dividing by ΔT, the estimated c can come out too large.
This is a powerful general lesson: the direction of an experimental error should be reasoned from the equation and the physical mechanism, not memorised.
The apparatus is part of the thermal system
A typical setup uses a metal block, immersion heater, thermometer or temperature probe, insulation, ammeter, voltmeter, power supply and stopwatch. Each item has a role, and some also create limitations.
The heater transfers energy into the block by thermal contact. The thermometer estimates the temperature near its sensing region. Insulation reduces unwanted energy transfer to the surroundings. The electrical meters estimate the power supplied. None of these measurements is automatically perfect.
The Institute of Physics recommends exactly this kind of setup and notes that a cooling correction can improve the result. It also highlights thermal contact, insulation and power-supply behaviour as practical issues. See the IOPSpark specific thermal capacity practical.
Thermal contact: where is the temperature actually measured?
A temperature probe does not measure “the block temperature” in an abstract sense. It measures the temperature at its sensing location after exchanging energy with nearby material. If there is poor contact between probe and block, the probe may lag behind the bulk block temperature.
Similarly, if the heater does not fit well, some of its energy may heat air gaps instead of transferring efficiently into the metal. Appropriate contact material or fitted holes can reduce this problem where the apparatus is designed for it.
Why insulation helps but does not make the experiment adiabatic
Wrapping the block in insulation reduces heat transfer to the surroundings, especially by convection and conduction. But no classroom insulation is perfect. Heat still escapes through exposed surfaces, wires and supports.
Therefore “use insulation” is a good improvement only if you explain the mechanism: it reduces the energy leaving the block during heating, so the observed ΔT represents a larger fraction of the electrical energy supplied.
A stronger method: temperature against time
Instead of recording only an initial and final temperature, take regular temperature readings while heating. Plot temperature against time.
If electrical power is approximately constant and heat loss is small initially, the early slope of the graph reflects heating rate. As the block becomes hotter than the surroundings, heat loss grows and the slope may decrease.
This means the graph itself can reveal a violation of the simplest model. A curved temperature-time graph is not “bad data”; it may be evidence that the rate of heat loss changed as the temperature difference from the room increased.
Cooling correction: using later data to estimate earlier loss
A more advanced school method continues taking readings after the heater is switched off. The cooling trend can be used to estimate how much temperature rise was lost to the surroundings during the heating interval.
This is conceptually important because it turns “heat loss” from a vague complaint into measurable evidence. Instead of merely stating that heat escaped, you use the cooling behaviour to estimate its effect.
Quantitative window: a full estimate
Suppose an aluminium block has mass 0.95 kg. The heater is supplied with 8.0 V and 3.0 A for 300 s. The temperature rises from 21.0 °C to 29.0 °C.
Electrical input:
E = VIt = 8.0 × 3.0 × 300 = 7200 J
Temperature rise:
ΔT = 8.0 °C
Estimated specific heat capacity:
c ≈ 7200 / (0.95 × 8.0) ≈ 947 J kg⁻¹ K⁻¹
This is reasonably close to a typical room-temperature value for aluminium, but agreement alone does not prove the method was valid. A wrong method can occasionally produce a plausible answer through compensating errors.
Uncertainty: which measurement matters most?
Students often focus on the stopwatch because time appears explicitly in the formula. But if timing is 300 s to the nearest 1 s, its fractional uncertainty is small. If ΔT is only 2 °C measured to the nearest 1 °C, temperature uncertainty may dominate.
This suggests an experimental design principle: where safe and appropriate, choose a temperature rise large enough that thermometer resolution is a small fraction of ΔT—but not so large that heat loss becomes overwhelming.
Observation versus inference
Observation: “The block temperature rose from 21.0 °C to 29.0 °C in 300 s.”
Inference: “The block absorbed thermal energy.”
Stronger inference: “Most of the measured electrical energy was transferred into the block and associated apparatus, though some was lost to the surroundings.”
Keeping those layers separate prevents students from treating the ideal equation as a directly observed fact.
Failure modes that cap practical standards
- Tiny temperature rise: thermometer uncertainty becomes a large percentage of ΔT.
- Poor insulation: too much electrical energy leaves the block.
- Poor probe contact: measured temperature may lag or misrepresent the block.
- Changing voltage/current: assuming constant power becomes weak.
- No waiting for equilibrium: initial temperature may not be uniform.
- Using a value because it is “close to the textbook”: agreement is not evidence that the method was sound.
Unfamiliar transfer: water instead of a metal block
Suppose the same method heats water in a beaker. What changes? The water needs stirring to reduce temperature gradients. The beaker has its own heat capacity. Evaporation may remove energy. The heater may warm water unevenly. The energy-accounting logic remains the same, but the dominant error mechanisms can change.
This is what transfer looks like: preserve the reasoning structure while adapting the error model to the new apparatus.
Secondary → JC → deeper Physics
Secondary: use E = mcΔT, measure voltage/current/time and mass, identify heat loss, choose insulation and calculate c.
JC: analyse gradients, estimate uncertainty, use cooling corrections, discuss thermal contact and recognise that the apparatus contributes heat capacity.
Deeper Physics: calorimetric measurement extends into thermal modelling, lumped-capacitance assumptions, heat-transfer coefficients, data logging, calibration and formal uncertainty budgets.
Checkpoint 1: which improvement actually targets the problem?
A student obtains a value of c much larger than the accepted value. She suggests: “Use a stopwatch that measures to 0.01 s.” The heating time was 600 s, but ΔT was only 3 °C.
What is likely the more useful improvement, and why?
Checkpoint 2: interpret the graph
Temperature rises quickly at first but then more slowly even though V and I remain approximately constant. What does this suggest about the energy balance?
Answer key and WHY reasoning
Checkpoint 1: improve temperature measurement and/or increase the safe temperature rise, while reducing heat loss with better insulation. A 0.01 s stopwatch barely improves a 600 s measurement; the fractional timing uncertainty is already tiny. A 3 °C rise makes temperature resolution and heat loss much more significant.
Checkpoint 2: as the block becomes hotter relative to the room, the rate of heat loss increases. With roughly constant electrical power, a larger fraction of input power is leaving the block, so the net heating rate decreases.
How to study this practical to a high standard
Do not revise by memorising “heater, thermometer, block.” Instead, redraw the experiment as an energy-flow diagram. Then annotate every measured quantity with its instrument, uncertainty and hidden assumption. Finally, practise predicting the direction of each error using the equation.
Your target is to be able to answer four questions without notes: What energy did we measure? Where else could it go? Which measurement dominates uncertainty? What change would most improve the evidence?
Evidence boundaries
A classroom experiment can estimate the specific heat capacity of a sample under its experimental conditions. It does not establish that c is perfectly constant at every temperature, phase or composition. Nor does a close result prove every assumption was valid. State conclusions at the resolution the experiment actually supports.
Authoritative next steps
- Institute of Physics: Specific thermal capacity of aluminium more accurately
- Institute of Physics: supporting A-Level practical requirements
- SEAB O-Level syllabus directory
- SEAB A-Level syllabus directory
Teaching Guide
For teachers and parents: ask students to explain the experiment as an energy ledger rather than a formula exercise. Give them one wrong result and require a directional diagnosis: “Which mechanism would push c too high? Which would push it too low?” Then ask them to rank improvements by expected impact. This develops practical judgement, not merely procedural recall.