# How to measure temperature

Some theory about temperature measuring technique.

Measuring temperature is actually more difficult than you can expect. Everybody has sometimes experienced that different thermometers do not show exactly the same value. This can depend on the measuring accuracy of the thermometer, but it can also depend on the way the temperature probe is placed. In the following we are trying to explain the theory.

What is temperature?
Temperature is actually a measure of how much energy a body contains. Physically it is a measure of how fast the atoms are moving in the material, the warmer the material is the faster the atoms are moving. In other words temperature is a measure of the kinetic energy of the atoms. If it is cold enough, the movement of the atoms stops completely, what is called the absolute zero point (-273.15°C).

Heat exchange
Nature strives to level temperature differences between two bodies (materials) getting in contact with each other. Heat is always transferred from a warm body to a cold. You will notice this when you open the hot oven and the heat flows out towards you. In the same way the heat will flow into an open freezer (even if we experience it as if the cold flows out, but it depends on that we feel the heat leaving the body).
The exchange can take place in one or more of the following ways:

• Conduction, for instance heat conduction in a metal bar.
• Convection, for instance thermal circulation in gases and fluids.

Heat conductors and isolators
Different materials have different conductivity. Example of good heat conductors are metals. Of course, different metals do not conduct heat equally good. A silver spoon rapidly gets so hot that you can not hold it when you dip it into hot water. A corresponding spoon of stainless steel does not get hot equally fast, and a plastic spoon is hardly warm at all. Silver is consequently a much better heat conductor than steel, which is a much better heat conductor than plastic. Plastic is generally a poor heat conductor and is therefore called an isolator. A very good isolator is static air, that is why we insulate with cellular plastic (containing air bubbles).
Fluids, for instance water, conduct heat better than air.

To measure with temperature probes
Thermometers always measure the temperature in the sensor!
A good temperature probe must as fast as possible acquire the same temperature as the measured object. Here are the factors involved:

• The mass of the probe.
• The heat conductivity in the active part of the probe.
• The heat conductivity in the measured object.
• The contact surface between probe and measured object (heat conduction between them).
• Possible convection in the measured object, that is how the heat varies and moves within the object.

The first three factors are hard to change, they are depending on the probe design and the type of object that is measured. A fast temperature probe must of course have the smallest possible mass.
But you can try to increase the contact surface between object and probe. If you are measuring a package with frozen goods, the active part of the probe shall be placed so that it is in direct contact with the goods. If there is air between the probe and the goods (also possible air in the package) the heat conduction will be decreased since the air is insulating. This will cause an incorrect reading or it will take longer time before the correct temperature is measured!
You can improve the convection by, for instance, whipping around the fluid or waving the probe in the air.
Do not forget that a wet surface has a better heat conduction than a dry!

IR measurement
Measuring with IR sensors (pyrometers) has become popular, in particular because the reading is very fast. But this technique is affected with a great inaccuracy in the measuring result, which will be explained here:
A pyrometer measures how much heat is radiated from a body and hence its temperature. In theory this technique is optimal, because pyrometer does not touch the object and therefore really measures the temperature of the object and not in the sensor. Hence there is no problem with the sensor conducting heat to or from the object.
But in reality it does not work that well! The IR sensor measures all the heat radiation coming from a certain direction (hopefully only from the measured object). This heat radiation has three different origins:

• Emitted – heat radiated from the measured object.
• Reflected – heat radiated from surrounding objects and reflected by the surface of the measured object.
• Transmitted – heat radiated from objects behind the measured object and transmitted through the measured object (only in certain cases).

The pyrometer measures all these three components, but it is only the emitted radiation we want to measure. A common problem is that the body heat of the person measuring is reflected by the measured object and hence increases the measured temperature.
Another factor is the surface of the measured object. Matte surfaces radiates different from glossy, and dark surfaces different from light. It is obvious that the inaccuracy in measuring on different more or less colourful packages will be huge! To be added: the temperature of the pyrometer also affects the result!
All these affecting factors together could give a measuring inaccuracy of 4–5°C!
We are talking about deviations so large that you could as well estimate the temperature by sensing with your fingers!

Regarding what have been said above, it seems that the IR technique is completely impossible to use and it is partially true. This technique requires a great knowledge from the person measuring!
When can you use the IR technique?

• When you need to compare temperatures in different parts of the same area, for instance on the walls in a refrigerator. But you cannot compare the readings of the wall surface with packed goods, because of the different surfaces.
• When you have the instrument mounted at a conveyor belt where you check the temperature of the passing objects, where you measure on the same type of surface and where the surrounding radiators are the same. In this case you can adjust the instrument to read correctly compared to a reference thermometer.

When it comes to temperature measurements in food, the Swedish National Food Administration recommends not to use IR instruments with this motivation:

"The trouble is that the measured value will be influenced by circumstances which can be hard to take in, for instance type of packing material or the temperature of surrounding surfaces."

Measuring accuracy
Modern thermometers are often digital, they show the temperature as figures on a display. Most instruments show the value with a resolution of 0.1°. This has nothing to do with the real accuracy of the instrument. In fact it can show a value several degrees wrong, compared to a calibrated reference instrument. It is important to remember this fact!

The manufacturer of the instrument usually indicate its accuracy within a specified measuring range, for instance accuracy ±0.5°C between -30°C and +70°C. An accuracy of ±1.0°C means that the read value differs maximum ±1.0°C from the real value. A reading of 4.6°C indicates that the real temperature is somewhere between 3.6°C and 5.6°C. (The real value can be determined with a reference instrument with considerably better accuracy, for instance ±0.1°C).

Calibration
In order to know how much wrong an instrument can read, it has to be calibrated. It means that the read value is compared to a reference instrument, which is traceably calibrated against one of the national measurement standards for temperature. In other words, calibration is a comparison. When you calibrate, the deviations will be documented in a calibration certificate.

### Eltex of Sweden AB

Post:    Box 24
SE-283 21  OSBY
Street: Södra Portgatan 19
SE-283 50  OSBY
Sweden
Phone:+46 479 53 63 00
Fax:    +46 479 53 63 99
Email: info@eltex.se

### For inquiries in USA, Canada & Mexico

Eltex U.S., Inc.
Greer, SC 29651, USA
Phone: +1 864 879 2131
Fax: +1 864 879 3734
Email: sales@eltexus.com

### Other countries

For contact details of a