It was 2:47 AM when the email landed in Dr. Maya Chenâs inbox. The subject line read: âcurran iso-octane mechanism download â urgent.â
If youâre reading this, youâre probably trying to simulate iso-octane combustion. Stop. The 2006 mechanism has a known error in reaction R4321 (C8H18 + OH = C8H17-4 + H2O). The rate constant is off by a factor of 2.3 above 900 K. I never published the correction. But itâs in this file, at the bottom, under the heading âTRUE_PARAMS.â curran iso-octane mechanism download
But his mechanismâthe legendary, exhaustively validated chemical kinetic model for iso-octane combustionâwas very much alive. Every PhD student simulating engine knock or trying to suppress soot in a jet fuel surrogate had begged, borrowed, or hacked together pieces of it. The full mechanism, however, was a ghost. It existed on a long-decommissioned FTP server at NUI Galway, mentioned only in footnotes of papers from 2002. Rumors said it had 1,234 species and nearly 6,000 reactions, tuned so precisely that it could predict flame speed within 0.3% error. It was 2:47 AM when the email landed in Dr
Maya needed it. Her simulation of low-temperature oxidation in a homogeneous charge compression ignition engine had been failing for six months. All because she was using a simplified mechanism that skipped the crucial third-stage ignition delay. I never published the correction
Dear Maya (or whoever found this),
Mayaâs heart hammered. She fired up her old terminal emulatorâthe one with FTP support for legacy protocols. Her fingers flew: open 131.170.168.223 , user anonymous , cd /curran/iso-octane . The directory listing appeared: a single file, mech_full_v3.4.dat , dated 2006-09-15. Size: 17.3 MB.
Also: the server is not haunted. I set up a cron job before I died. My son maintains the electricity bill. Science shouldnât disappear just because scientists do.