research linking Black-Scholes computer algorithms & "flash crashes"?

i'm trying to research the role of computer modeling in the market crashes of 1987 & 2010, and i was wondering if there was anywhere to look into the mathematical deficiencies of the programmers' models.

this link LINK gives a good overview of a VPIN measure to explain 2010

this LINK shows how to extend VIX measures to before its 1993 launch.

but where can i look for places that identify the flaws in some of the specific models used by quants??