Why can’t thermal camera measure sun/fire temperature?

[Logan]

Hello.
Why do fire and sun appear much colder on thermal cameras than they actually are?
Thank you.

[Vipitis]

every thermal camera comes with a specified range. In fact your normal digital camera is more suited to measure such high temperature, they are glowing in the visible range and if you calibrate it, it's possible to measure that.

[Logan]

every thermal camera comes with a specified range. In fact your normal digital camera is more suited to measure such high temperature, they are glowing in the visible range and if you calibrate it, it's possible to measure that.

Thanks you.
But they didn’t reach the range limit at all, for example, the sun reads ~150 C degrees.

[AkiTaiyo]

Thermal cameras are used to measure flame temperature, but to do it accurately you need to be measuring the correct wavelengths and know the composition of the combustion materials. 
Typically for very hot things, measuring in the 8-14 range does not make much sense and can be done using near infrared cameras instead.
Some cameras are very good at picking up soot and other bi-products, however clean burning flames emit very specific bands of radiation which cannot be picked up if you’re measuring a different wavelength. 
Water in the atmosphere can also be a problem as water is very good at absorbing quite a lot of radiation across the IR spectrum, and is also a biproduct of a lot of combustion.
Infrared cameras are used (along with nice telephoto lenses) for flare-stack monitoring, though temperature accuracy often isn’t the prime objective.
See:
https://www.ametek-land.com/-/media/ameteklandinstruments/documentation/industries/hpi/ametek_land_flare_stack_monitoring_application_note_rev_2_002_en.pdf
and

https://www.advancedenergy.com/products/temperature-measurement/thermal-imagers-and-systems/flarespection/

[JohnnyMalaria]

The cameras are only sensitive to infrared. Hotter objects emit more in the visible than cooler ones and the cameras don't detect that.

Check out the Ultraviolet Catastrophe: https://en.wikipedia.org/wiki/Ultraviolet_catastrophe

Also, the atmosphere absorbs a lot of infrared coming from the sun: https://en.wikipedia.org/wiki/Infrared_window

[Fraser]

Just think about the sun for a minute....... it is very hot, yes, but it is a long way away from us in a Vacuum and we have the Earths atmosphere in the thermal energies flight path. If it were a ‘zero loss’ path we would all be incinerated. A camera system would need to be carefully calibrated to measure the suns temperature.

As Vipitis has mentioned a LWIR camera may not be the best tool for the job. The most powerful wavelengths emitted by the Sun fall outside the LWIR passband and the ‘Sun Safe’ filtering installed in modern cameras. The camera is effectively deliberately blinded to the dominant thermal wavelength emissions from the Sun. A MWIR or SWIR camera would see more of the dominant energy so great care would be needed if using such to view the Sun in order to avoid detector damage.

The matter of measuring fire is similar. It is relatively easy to measure a surface that is being heated by a flame provided the surface temperature falls within the capabilities of the camera. If a thermal camera is used to measure actual flame temperature you need to consider the wavelength in which the flame is emitting most of its energy and whether the camera is calibrated to measure such a wavelength and temperature. There is also the previously mentioned filtering used on LWIR cameras to be considered. In MWIR cameras there is a lot of flame energy in the passband of the camera and such a camera may be used to measure flame temperatures if calibrated for such. There are specialist MWIR filters for viewing just the flames wavelengths or to exclude the flame wavelengths to view the surfaces ‘behind the flames’. Such filtering is common when carrying out preventative maintenance on open flame boiler systems.

Fraser

[Logan]

Thank you everyone!
I’ve learned a lot from you guys!

[bap2703]

Just think about the sun for a minute....... it is very hot, yes, but it is a long way away from us in a Vacuum and we have the Earths atmosphere in the thermal energies flight path. If it were a ‘zero loss’ path we would all be incinerated. A camera system would need to be carefully calibrated to measure the suns temperature.

As Vipitis has mentioned a LWIR camera may not be the best tool for the job. The most powerful wavelengths emitted by the Sun fall outside the LWIR passband and the ‘Sun Safe’ filtering installed in modern cameras. The camera is effectively deliberately blinded to the dominant thermal wavelength emissions from the Sun. A MWIR or SWIR camera would see more of the dominant energy so great care would be needed if using such to view the Sun in order to avoid detector damage.

The matter of measuring fire is similar. It is relatively easy to measure a surface that is being heated by a flame provided the surface temperature falls within the capabilities of the camera. If a thermal camera is used to measure actual flame temperature you need to consider the wavelength in which the flame is emitting most of its energy and whether the camera is calibrated to measure such a wavelength and temperature. There is also the previously mentioned filtering used on LWIR cameras to be considered. In MWIR cameras there is a lot of flame energy in the passband of the camera and such a camera may be used to measure flame temperatures if calibrated for such. There are specialist MWIR filters for viewing just the flames wavelengths or to exclude the flame wavelengths to view the surfaces ‘behind the flames’. Such filtering is common when carrying out preventative maintenance on open flame boiler systems.

Fraser

That's not totally correct. An unfiltered sun wouldn't incinerate anything. It's just a mere  1.5 more power than what we would get below the atmosphere (look for AM0 vs AM1 spectrum). But it's UV content would be bad.

The sun surface is very close to a blackbody: it's an optically thick plasma where photons can radiate at every possible energy, and you don't see the effect of what's behind it. That makes it a good candidate for temperature measurement through radiation. The issue isn't even about the band in which is system is working : you just calibrate your sensor in that band and NOT against the whole radiated power.

The issue with high flux target is technological/commercial :
- it's so far above what you will usually measure that it's not worth calibrating there for only a few users.
- it's damaging the sensor: actually sensors need to be designed to withstand it because you can inadvertently point it to the sun. But it's expensive to attain that spec: I mean you're probably sacrificing something somewhere to be able to just look at the sun a short time. The more ability you want to look at the sun, the more you are sacrificing from your main usage.

A thermal camera measuring the sun's temperature is totally doable.


Now for fire : you're now looking at a completely different object. It's a slightly ionised plasma = it's more a gaz where only a low amount of chemical reaction emit light directly and heat that particular gaz.
That can be far from a blackbody.
Look for example at rocket engine exhausts: H2 + O2 engines burn a transparent flame... it's still very hot

It's still doable to measure the temperature of non blackbody objects: you just need to not do the assumption that it's a blackbody :p

[Fraser]

Note I said ‘Zero Loss path’ and not ‘unfiltered’    ‘Zero path Loss’ was referring to the combination of distance and atmosphere and I think you will, agree, if a ‘no path loss’ situation occurred between the Suns surface temperature and the observer.... there would be the distinct smell of burnt flesh

As I said a thermal camera can be calibrated to take account of path losses. No argument there but you need to know what you are doing to achieve reasonable accuracy as the Atmospheric influences are not a constant. It all depends on what the measurement accuracy requirement is.

Your comments on measuring the temperature of a flame rather than a surface heated by a flame are very interesting. I have often wondered what the flame actually ‘is’ and how it behaves with respect to emissivity. Thank you for that insight.

[Bill W]

The cameras are only sensitive to infrared. Hotter objects emit more in the visible than cooler ones and the cameras don't detect that.

No, that can, and is, calibrated out.  A camera set up to measure a 1000°C black body really does measure 1000°C (or thereabouts) when looking at it.  It is simply a rescaling of 'received energy' by the factor you say, emitted energy / passband from the Planck curves.

What the OP is seeing in 'they all say 150°C' is the result of sloppy and misleading software that should at the very least be saying '>150°C' or '+++°C' having detected signals above the limited calibrated level.  [FLIR Bosons just wrap round and show cold though - design faulty / crippled]

At some source temperature the detector will begin to saturate on the more responsive pixels, maybe around 180°C if set to range up to 150°C.  That is where 'high temperature' ranges are used to lower the sensor gain.


Bill

[JohnnyMalaria]

The cameras are only sensitive to infrared. Hotter objects emit more in the visible than cooler ones and the cameras don't detect that.

No, that can, and is, calibrated out.  A camera set up to measure a 1000°C black body really does measure 1000°C (or thereabouts) when looking at it.  It is simply a rescaling of 'received energy' by the factor you say, emitted energy / passband from the Planck curves.

What the OP is seeing in 'they all say 150°C' is the result of sloppy and misleading software that should at the very least be saying '>150°C' or '+++°C' having detected signals above the limited calibrated level.  [FLIR Bosons just wrap round and show cold though - design faulty / crippled]

At some source temperature the detector will begin to saturate on the more responsive pixels, maybe around 180°C if set to range up to 150°C.  That is where 'high temperature' ranges are used to lower the sensor gain.


Bill

I need some convincing Most cameras detect in the 3 - 12 micron wavelength range. A white hot body emits a lot of energy across the visible spectrum, too. Hence, a camera that is not sensitive to visible wavelengths cannot measure the temperature of a body emitting in the visible range.

[AkiTaiyo]

I need some convincing Most cameras detect in the 3 - 12 micron wavelength range. A white hot body emits a lot of energy across the visible spectrum, too. Hence, a camera that is not sensitive to visible wavelengths cannot measure the temperature of a body emitting in the visible range.

E = ((2hc^2)/(λ^5))(1/(e^(hc/λkt)-1)

If your infrared source is a blackbody or close approximate, then using the Planck formula for just one wavelength in the typical 8-14 thermal camera range, the value of emitted energy increases with temperature.  Since the emitted energy at that wavelength is increasing, you can detect that and convert it back to a temperature.

[Bill W]

The key is 'black body' ie one that has no wavelength dependence on radiant energy beyond the Planck curve.  However most things are at worst 'grey' so it does not go far wrong on 'lumps of hot stuff' of any temperature

It can, and indeed must, go wrong when considering line output sources and the like.

Attached a graph of energy vs temperature for various LWIR passbands.  As you see hotter is always higher, even though the curves are not the whole black body energy output

Bill

[bap2703]

I have often wondered what the flame actually ‘is’ and how it behaves with respect to emissivity. Thank you for that insight.

A flame is a gaseous chemical reactor with plenty of energy.
Many process can lead to direct emission of light that is not related to our usual thermal emission.

For example two papers that show what can happen in hydrocarbon flames.
https://www.researchgate.net/publication/263198897_Model_based_method_to_obtain_lean_turbulent_flame_heat_release_rate_from_chemiluminescence
https://www.researchgate.net/publication/257401843_CO2_chemiluminescence_study_at_low_and_elevated_pressures

Thermal emission is still there but not that useful for thermography (more useful for your CO2 heatseaking missile  )

[Fraser]

Thank you bap2703

One of the reasons that I love working with thermal cameras is that it is a voyage of discovery with new experiences and knowledge ‘around every corner’. 

Thank you for taking the time to explain 

Fraser

[Max Planck]

Hello.
Why do fire and sun appear much colder on thermal cameras than they actually are?
Thank you.

In fact it is possibble to measure flame temperature using an IR camera. An interference filter is required. From memory a 4.26um one. A 3.9um filter is usually applied for through flame measurements. Thus, an MWIR camera is needed

Max

[Max Planck]

I need some convincing Most cameras detect in the 3 - 12 micron wavelength range. A white hot body emits a lot of energy across the visible spectrum, too. Hence, a camera that is not sensitive to visible wavelengths cannot measure the temperature of a body emitting in the visible range.

E = ((2hc^2)/(λ^5))(1/(e^(hc/λkt)-1)

If your infrared source is a blackbody or close approximate, then using the Planck formula for just one wavelength in the typical 8-14 thermal camera range, the value of emitted energy increases with temperature.  Since the emitted energy at that wavelength is increasing, you can detect that and convert it back to a temperature.

The hot gases contained inside a flame are neither a black or a gray body. They are selective emitters radiating IR at wavelengths one can find looking for their absorption spectra.

Max