Thanks for the link. I'll need to read it more closely.

Warning: Some of the graphs are badly mis-drawn. The first one on the glass phase has an axis that runs 50 to 300, and it should be 0 to 250 or 50 to 250 in order to agree with the text. The ones showing LTLT and HTST roasts both have a line for bean center temperature lines which is not mentioned in the text. The LTLT one is physically impossible, since the center gets hotter than the environment and does so in 30 seconds. In any case, the article's editor must have been sniffing something other than coffee, so be on the lookout for further errors.

Page 454. The graph shows that the bean center hits 225C in 100 seconds with the air blowing in at 228C. Then the bean center stays constant for another 600 seconds. 225C is 437F, roughly the second crack. So according to the graph, the roast hits the second crack at 1 minute 40 seconds, and then stays there for another 10 minutes. Either there's a mistake here (for instance, the solid line is the supply air temperature, not the bean temperature) or I really don't want to try this coffee.