Looking 10 minutes on the web i found a portable thermal camera with the temperature range from 0 to 2800 F. Below is the insert:
http://www.ir55.com/MikroScan.html
Ah, no you didn't.
I assume you're referring to the following specification:
Sierra Pacific Innovations (??) said:
Spectral Range: 8-12 microns for models 570 and up (LWIR)
Detector: 320 x 240 PtSi focal plane array (550 model) / 320x240 MicroBolometer for models 570 and greater.
Standard Field of View (FOV): 24° x 18°
Instantaneous FOV: 1.1 mrad
Multiple colour palattes
Swedish, French, Spanish and english user selectable menus
Accuracy: +/- 2% or 2°C
Thermal measurement Range: 1500c (*570 E cameras) email
rk@x20.org for additional info
Now, what this marketeering site
doesn't tell you is that in order to get measurements in the 1500
oC range, you need to set up the camera very, very carefully, and even then the accuracy will be abysmal. You can't just pull it out of the case, turn a dial, and get that result. You need to know
precisely what the material is made of. You need to set up proper filters for a particular temperature range. And you need to control the geometry of the problem. You would need to place a calibrated target basically right next to the object you wanted to measure, and shoot from only that location. Changes in angle, smoke, and sunlight will require a recalibration.
None of this is possible from a helicopter, period.
The reason for this is in the detector. Note this is an LWIR device, with peak sensitivity between 8 and 12 microns. This is a great range for measuring the temperature of something where the peak emission wavelength is between 8 and 12 microns, or, in other words, 241 K to 362 K, or about -30
oC to 90
oC, according to
Wien's Law. This covers the range of most daily temperatures and most living things, and this is the range nearly all LWIR cameras operate in. This includes the FLIR mounted on DEA helicopters, which I imagine is what the images referred to actually came from.
The way the camera measures temperature is one of two ways, either measuring photons (focal-plane array) or measuring the total amount of energy deposited in a given pixel (bolometer). The effect is pretty much the same, though the physics is different. The FPA approach is electrical in nature, sort of like a CCD, whereas the bolometer measures tiny increases in temperature caused by absorption.
Both approaches have their strengths and weaknesses. The FPA is preferred in nearly all portable applications, because it allows a lighter device. Bolometers need active cooling and large heat sinks, often using liquid nitrogen, and are expensive and messy particularly in an aviation context. On the other hand, the FPA is much harder to calibrate, and every single pixel requires its own calibration because tiny variations in the array change its characteristics; most FPAs have a built-in sequence to do this but accuracy is limited. A good bolometer can be made with great precision, and as a result can be pushed further outside its normal envelope by a clever researcher.
Whichever device is used, it measures temperature by measuing how much energy arrives at a pixel in some given length of time. When the sensitivity window is comparable to the peak wavelength of the emitted spectrum, the detector captures a large fraction of the total energy. This energy varies with the
fourth power of temperature. In our typical example above, going from the low end to the high end -- 241 to 362 K -- means a five-fold increase in energy. So our detector needs to respond to a maximum gain of five. If you want to resolve temperature differences of 1 K, you need about eight bits of resolution, because the difference from 241 to 242 K is about 1/245th of the difference from 241 to 362 K. No problem.
Aha, but now you want to point that detector at a hot fire, up to 1500
oC, or 1773 K. Big, big difference.
First of all, the peak frequency is now well outside the sensor's range -- it's at 1.6 microns.
Second of all, the total power output has raised by a factor of nearly
three thousand.
Both cause you problems. Obviously, the energy output is so much higher that you need some way to either block it (like putting on sunglasses) or else you need a whole lot more than eight bits of range. But you also need a totally new analysis approach. Before, you could assume that your detector captured a good fraction, say 25% or more, of the total energy. Now, that is no longer true. Your window of 8-12 microns is now not even a few percent. So your original assumption -- and the equation your camera uses to estimate temperature -- goes out the window.
The other problem is that your detector is
not only sensitive to the 8-12 micron band. It is
most sensitive there, and there is some rudimentary filtering applied (filtering that is as lightweight as possible, since it too emits thermal energy and will confuse your readings), but it will be partially sensitive to radiation well outside this range. The actual response curve will be a spiky, jaggy thing over the entire continuum. So, suppose your detector is 100% efficient on photons in the 8-12 micron range, and a mere 5% efficient in the 1-2 micron range. Well, with a temperature this high, the contribution from 1-2 microns may actually swamp that of the intended range. Unless you've tested this, you cannot predict how it will behave. And as the temperature goes up, more and more of the shorter wavelengths come into play.
So the smart thing to do, obviously, is pick a different detector that is most sensitive around the new peak wavelength. Why don't we do that? Well, we do, in fact, with a device called an
optical pyrometer. What this does is to either intercept
all of the incoming radiation, across all bands, or it seeks the peak wavelength, and from either measurement extrapolates temperature. These are pretty common.
Unfortunately, they don't work in this case. A lot of the energy is in the visible or near-IR spectra. That's bad because now there are lots of new sources of interference. Smoke will block it, sunshine will add to it, and now the emissivity of the material becomes of crucial importance. An optical pyrometer only works well if you can test it in the shade, and if you can get pretty close to the object you're studying. Neither of these work well in this case.
The reason the DEA FLIR is still somewhat relevant is the 8-12 micron range is pretty good at cutting through smoke, and is not greatly affected by daylight. So it's a good one to use from a helicopter. But, because you're operating in the entirely wrong part of the spectrum, it gives you terrible performance in temperature estimation. I even believe the FLIR may have given them a 2800
oF estimate, but that number can't be trusted -- it would be +/- 1000
oF or more. I've actually observed this effect with LWIR cameras. The ones with better software throw you an "UNCAL" flag to warn you that the number is totally bogus.
Now, back to the specification. You can, if you're careful, set up a good LWIR camera just like an optical pyrometer. But to do this requires repeatability. Nobody knows what the camera will do if you point it at a 2800
oF object under certain conditions of lighting, distance, surface condition, and so on, so what you do is set up a test where the temperature is known. Then you add filters and adjust the gain and fiddle with the camera until that known object shows up in the middle of the camera's operating range. Then, so long as everything stays the same, you can use it to estimate temperature of a similar, unknown object. It's doable in the laboratory, and absolutely ridiculous on the Pile from a helicopter.
But there is one other way to do it, one that doesn't require all the babysitting described above. The other way is, rather than estimate temperature from the
total energy received by the detector, is to look at the
shape of the blackbody curve. Even if you're restricted to a suboptimal frequency band like the 8-12 micron range, you can still make a pretty good guess about the temperature from how steep the blackbody curve is in that region.
To do this, you need a detector with many different, narrow bands of sensitivity, or one that measures not just the total number of photons, but the actual energy (the "color," if you will) of each one. Such a device is known as a spectrometer. This is what AVIRIS does, and that's why it produces reliable temperature estimates, though the processing is complicated. See the papers I already linked you to, where they describe the shape-fitting process to work out the actual temperature.
But this can't be done with a thermal imager, period. Those have no capacity for spectroscopy.
In summary, there is no IR camera that will work in this case. Can't be done. You would either need (a) a near-IR camera, in which case smoke and daylight and materials will interfere so much that it's not worth doing; (b) some way to brute-force calibrate a LWIR camera that accurately describes the conditions of the Pile, which is not possible period; or (c) give up and hire a spectrometer.
The officials in charge chose option (c), and the instrument determined the temperatures were nowhere near 2800
oF. Maximum observed was under 1400
oF. That's all there is to it.
What evidence do you have that they bought the AVIRIS flights? Why would they need to if they had helicopters? I thought we agreed that there were DEA helicopters and the NASA flights on separate missions?
The evidence I have is that there are several papers presenting the AVIRIS data. Obviously that instrument was there, is capable of measuring the temperature, and proved it was much lower than the other article (with no support and no credible way of getting its measurement) claimed.
Would you support the analyze of the meteorite to establish empirically and to end all speculation that it is indeed – (as the expert suggests) – a fused element of steel and concrete through heat?
Now assuming it is determined to be a fused element by heat and we know that to melt steel takes 2800 F. We can conclude that absolutely ‘something’ must have generated this heat. Any ideas?
No. There is no reason to assume it
was fused by heat. During the collapses, it would have experience pressures in excess of 100,000 PSI. That's more than enough to fuse steel and concrete. Even at room temperature, if you hit steel hard enough, it will weld.
Anyway, regarding the camera specification, it's an honest mistake you made and I hope my explanation helps to clarify. Unfortunately, there is a huge difference between finding something that sounds like it could be right, and actually being right. This is why merely Googling around is no substitute for actual training and expertise, and this is why the Truth Movement ever existed to begin with.
