Checking lighting in inspection dept.

Hello, there was a guy who left where I work that used to check the brightness of the lights in the inspection area. He left and I can't find any procedure or specs to follow. We are an AS9100 shop and I was hoping someone can point me in the direction of how many Lumens or candlepower we should have. I tried holding the meter to get a reading but they were all over the place when I held it at different distances away. Thanks for any help.

rich p said:

I tried holding the meter to get a reading but they were all over the place when I held it at different distances away.

Click to expand...

What are you trying to test? If you had a previous inspector, and if none of the bulbs are burned out, you should have pretty much the same illumination today as when the inspector was working. As for different readings, how different are they when you hold the meter horizontally (i.e. with the sensor pointed up), at arm's length away from you so your body isn't blocking any light coming from the side, and move it around at the height of the inspection table? Or, if the area is for inspecting large 3D objects, how much do the readings change when moving the meter side-to-side at the highest point to be inspected, the lowest point, and up/down from the highest to the lowest point? Also, what meter are you using for this?

Light energy (usually measured in lumens/sq meter) follows the inverse square law, so the reading will always change with distance. Most light meters have a limited angle of sensitivity so what direction you are pointing the sensor matters.

I’m wondering why this matters. Any burned out bulbs? Replace them.
Is this something to do with ISO? How would you adjust it anyway?

Sounds like this guy was trying to BS someone to justify his job.

Likely some form that has to be filled out for compliance and consistency. Not uncommon in the visual inspection world.
Light value should be checked at the part surface being inspected.
Bob

Solarbell said:

Light energy (usually measured in lumens/sq meter) follows the inverse square law, so the reading will always change with distance. Most light meters have a limited angle of sensitivity so what direction you are pointing the sensor matters.

Click to expand...

not really. Photons are not electromagnetic waves.

Larry Dickman said:

not really. Photons are not electromagnetic waves.

Click to expand...

Light is both.

Solarbell said:

Light energy (usually measured in lumens/sq meter) follows the inverse square law, so the reading will always change with distance

Click to expand...

That’s only true for a point source, and even at 10 ft. a single 60 W bulb in a room is a poor approximation. A room with a number of lights in a, say, 12 ft. ceiling will have essentially the same light intensity at the floor as 6 ft. above the floor, not 1/4 the intensity.

I guess we got written up a few years ago from an auditor but I wasn't working there at the time. I was told we had to upgrade our lighting in the inspection department. They have a meter I send out to calibrate once a year but I can't find the actual procedure. I thought it was AS9100 but it might have been a supplier audit, like Pratt and Whitney.

A quick google suggests something between 1000 and 1500 lux, even and shadow free, will be a reasonable level. For example :- Lighting Level for Visual Inspection | Lighting | Rendering (Computer Graphics) and Inspection Lighting Standard for Class A automotive inspection area.

As the first link makes clear there are standards for this sort of thing, albeit more aimed at the design side than mandatory compliance inspection. My experience suggests that even, shadow free lighting at a comfortable colour temperature is more important than absolute brightness once you have basically "enough" light. Many folk find cold white from florescent or lED lights gives a sharper appearance to things but becomes uncomfortable after a while. Unsurprising as these are pseudo colour temperatures built up from various narrow band spikes rather than the smooth near black body curve of sunlight which our eyes evolved to exploit. Neutral colours on walls and equipment helps too. Spent a lot of my time in blacked out optical labs. Predominately black decor didn't help with precise work even though there was plenty of light when needed.

High level of lighting is only needed in the actual working area. Getting the right balance between general room light level and work area is important for comfort. Really helps if you have the right contrast level between room and workspace so you can look away from the bright working area into the main room to rest your eyes. Counterproductive if the difference is too great tho'.

Clive

Booze Daily said:

I’m wondering why this matters. Any burned out bulbs? Replace them.
Is this something to do with ISO? How would you adjust it anyway?

Sounds like this guy was trying to BS someone to justify his job.

Click to expand...

Does anyone still use bulbs in shops nowadays? Maybe buying and installing decent lighting is too easy a solution?

If it's even half as important as it sounds then mandatory eye testing should be done annually.

Clive603 said:

A quick google suggests something between 1000 and 1500 lux, even and shadow free,

Click to expand...

Even and shadow free, how is this done?
In the machine vision world you may see this as called "cloudy day illumination". A lot of light power will not end up in the inspection area.
Bob

CarbideBob said:

Even and shadow free, how is this done?
In the machine vision world you may see this as called "cloudy day illumination". A lot of light power will not end up in the inspection area.
Bob

Click to expand...

Been a long time since I was actively involved in optics and illumination but I imagine the sophisticated luminaire design computer programs can come up with a pretty decent design solution these days. As I recall things it was around the millennium when various suppliers started jumping up and down about how wonderful their illumination analysis program was as compared to crappy competitor Brand X. I guess desktop computers finally got powerful enough to do a half decent run on the calculations overnight. I was just pleased that a I could get a multi-element lens design properly optimised by the next day. Lens design is, of course, much simpler than illumination analysis.

Back in the day we'd just stick a few tubes behind thick opal glass, maybe with diffuser in the way, and call it good. Pretty much your "cloudy day illumination" I suspect. Less than efficient but did the deed. Albeit for optical labs not inspection so not as bright.

Clive

CarbideBob said:

Even and shadow free, how is this done?
In the machine vision world you may see this as called "cloudy day illumination". A lot of light power will not end up in the inspection area.
Bob

Click to expand...

Thats quite open specification to interpret but the idea is probably to have more than one lightning source. Fill the ceiling with luminaries, use light-colored ceiling and walls and maybe add some indirect light reflected via ceiling and walls. (Tho in some cases more contrast creating point source seems better, like surface irregularities and shapes)

Indirect lightning reflected via ceiling and walls gets close to ”cloudy day” but is less efficient.

I am in the API world. We are required to have 500 lux (50 foot candles) where inspection is performed. However, I would suggest get it up to at least 1000 lux.