Illumination and luminous intensity measurements

[Assasinator_2]

Hey guys, I need to check my understanding here so I'm going to run some ideas past you to see if I'm on the right track.

The Problem

I've got to check the luminous intensity of some headlights on a train. Luminous intensity is measured in Candela (cd), and is defined as the luminous flux per solid angle, or lumen per steradian. Candela is the SI unit.

Now, because it's per an angular measurement, it should not vary with distance from source, correct? However, if one is measuring it, then the measurement shall because the detector is not defined as an angular size, but rather a Cartesian size (say, 100mm^2 to pull a number out of thin air). As such, it would need to scale for distance from the source.

If that's the case, how would one use a light meter to measure candela? One would have to measure lumens (luminous flux), and then divide by the angular distance of the sensor relative to the source, yes? Or am I misled on how the devices will work?

Furthermore, my logic tells me that if you have two sources, sufficiently far away so as to be close enough to one source, the luminous intensities would compound, as you would be effectively doubling the flux through the same unit angular area. Is this correct?

The Solution

On the assumption that I am able to scale for distance, would be the best way to measure this? I'm currently thinking that I should get some sort of matte-black opaque tube, and place it over the headlight, with the light sensor at the end. The tube will minimise reflection of light by being matte black, and being opaque will stop external light from entering (assuming the summation of luminous intensities is correct which, logic says, it should be).

So basically, I'm after a way to measure the luminous intensity (in Candela) of a light source. From what I can understand, whilst luminous intensity does not depend on distance or area measured (as it is standardised flux/unit angular area), the measurement of it would involve measuring the flux and then using an assumption of sensor area as well as distance from source to equate an angular area. Furthermore, since intensities sum, you'd need to isolate it from surrounding light and ensure as little reflection off the isolating equipment as possible (it's not feasable to simply get rid of all other light sources).

What do you lot think?

[hlokk]

It would depend how the sensor works, but even if it depends on the area of the sensor, wouldnt the sensor allready take this into account and modify the ouput accordingly? Does the sensor give luminous intensity or something else?

The light sources would add up, and the sensor would measure any light on it indiscriminant from source (however, the light may have different frequency spread which may affect the sensor differently, but it depends on the sensor).

You may be able to use a non-reflecting tube on the sensor itself which would give it a narrower field of view to isolate the websites.

Otherwise, can you take an ambient reading with the sensor in a fixed location/angle, then turn on the train lights? You could also use some kind of 'infermometry'. I.e. do different angles and different distances, then model it and play with the numbers to figure out the components (but its a bit more complicated mathematically and practically)

[Assasinator_2]

Originally Posted by hlokk

It would depend how the sensor works, but even if it depends on the area of the sensor, wouldnt the sensor allready take this into account and modify the ouput accordingly? Does the sensor give luminous intensity or something else?

That's what I'm assuming (that it's calculating for area, rather than you manually doing it), but it would not be able to guess how far it is from the source.

The light sources would add up, and the sensor would measure any light on it indiscriminant from source (however, the light may have different frequency spread which may affect the sensor differently, but it depends on the sensor).

Thought so. It should all be more or less the same, as all the light should be coming from the bulbs. I'm thinking a matte black tube should be 'close enough' to non-reflective, given we're talking a significantly high value of intensity (200k candela).

Otherwise, can you take an ambient reading with the sensor in a fixed location/angle, then turn on the train lights? You could also use some kind of 'infermometry'. I.e. do different angles and different distances, then model it and play with the numbers to figure out the components (but its a bit more complicated mathematically and practically)

That last part kills it. It's got to be done by depot staff who just want to measure it and bugger off to smoko. Making it complex will mean it won't be done properly, or at all.

Cheers for your help

[hlokk]

Originally Posted by Assasinator_2

That's what I'm assuming (that it's calculating for area, rather than you manually doing it), but it would not be able to guess how far it is from the source.

Ah, I see what you mean. Wouldnt it be more likely to be a lux meter, rather an a candela meter? I've heard of lux meters but not candela meters. Reading from a lux meter can convert to lumens easily if you know the dist, then you'd just need to tube or whatever. Putting it on the lux meter will be easier I would think (gets less hot, less reflection due to angles involved). Tape a laser pointer on top to aim it

[Assasinator_2]

Originally Posted by hlokk

Ah, I see what you mean. Wouldnt it be more likely to be a lux meter, rather an a candela meter? I've heard of lux meters but not candela meters. Reading from a lux meter can convert to lumens easily if you know the dist, then you'd just need to tube or whatever.

That's what I'm assuming. Like a strain gauge can display newtons but really measures voltage, this would measure lux but display candela through the knowledge of its sensor area + an inputted distance-from-source.

Tube on lux meter is an interesting idea, but I'm not convinced that they can isolate surrounding light well enough. I'll ponder it, to be sure. Thanks for the help!

[ernie]

I think you will find that most light meters you can buy measure Lux not Candela, and you will have to calculate the conversion.

The tube idea sounds good, but most headlights have reflectors that will change the light intensity dependent on the angle you take the measurement from, so sticking a tube over it limits the range of viewing angles, thus won't give you the full range of intensities, and their pattern, which would occur in a real world situation.

- Ernie.

[Assasinator_2]

Originally Posted by ernie

I think you will find that most light meters you can buy measure Lux not Candela, and you will have to calculate the conversion.

The tube idea sounds good, but most headlights have reflectors that will change the light intensity dependent on the angle you take the measurement from, so sticking a tube over it limits the range of viewing angles, thus won't give you the full range of intensities, and their pattern, which would occur in a real world situation.

That's not a concern. The requirement is for the intensity at a location 240m away from the train to be a set value. We can't find 240m of spare place to do it, so we need to find some way of doing it in a smaller range. However, because it's so far away, we don't really need to concern ourselves with intensity off centre, since that doesn't really come into effect if you were 240m away.

[ernie]

Originally Posted by Assasinator_2

That's not a concern. The requirement is for the intensity at a location 240m away from the train to be a set value. We can't find 240m of spare place to do it, so we need to find some way of doing it in a smaller range. However, because it's so far away, we don't really need to concern ourselves with intensity off centre, since that doesn't really come into effect if you were 240m away.

Yeah but it sure might when you are close up, unless you make a really long tube! The bulb itself can reduce the on axis measurement close up, as they usually have a reflector at the end of the bulb that shines the light backward to hit the main reflector. That effect will diminish with distance. It depends on the model of lamp without specific info I can comment much.

You can see from this link that the filaments and supporting wires will make and on axis shadow that can taint your close up readings.

http://amglo.com/locomotive_sealed_beam.html

- Ernie.