LM-80 is a standardised method of measuring lumen depreciation and chromaticity shift in LEDs over time, and at different operating temperatures. LM-80 was written by the Illuminating Engineering Society of North America (IESNA, or IES more commonly) and it requires representative samples of LEDs to be assesed for a suitable period of time at at least two different temperatures (one of which must be either 55°C or 85°C). Data are typically recorded for several thousand hours and are then projected into the future using the methods and formulae defined in TM-21.
The exact duration of the assessment and the frequency of data sampling are decided by the LED manufacturer and will depend on the intended use and rated life of the LEDs. Note that LM-80 is not a test in the sense that there is a pass/fail result at the end; rather, LM-80 is a standard method of measurement so that lighting professionals can judge the relative merits of different LEDs for a particular application.
A lumen maintenance graph as will be found in almost any LM80 report. This shows the actual lumens recorded at different temperatures over time.
A graph of projected lumen maintenance. The actual lumen maintenance is projected forward using the methods defined in TM21.
In an LM-80 report Luminous flux is expressed as a percentage, with the initial light output, deemed to be 100%.
When chromaticity shift is quantified in an LM80 report it is done with reference to the CIE 1976 diagram.
Chromaticity is determined by plotting the initial light colour on the CIE 1976 colour space diagram.
In addition to the gradual shift of LEDs towards the red end of the spectrum, other colour changes can occur over the life of LEDs, especially if manufacturing defects are present. Two common such defects are Curling and Lamination.