We know, If the thermal illuminance after half an hour is 20% lower than that in the cold lamp body, the junction temperature is basically the tolerance range of the chip. It can basically judge that the heat dissipation system is unqualified.
Is there a special case? Of course, take a counterexample,
OSLON Square 1-5W The chip junction temperature increases from 25℃ to 40 ℃, the flux is Nitian rising! Then it began to decline, and the luminous flux decreased by about 10% at 120℃. So, for this chip, with half hour illumination method to assess the heat dissipation, only accept the change of 5-8%!