BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Waves Group (DAMTP)
SUMMARY:Computational AeroAcoustics: how they got it wrong
(repeatedly) - Dr Ed Brambley\, University of War
wick\, UK
DTSTART;TZID=Europe/London:20181023T103000
DTEND;TZID=Europe/London:20181023T113000
UID:TALK113140AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/113140
DESCRIPTION:Despite the best efforts of mathematicians\, aircr
aft still do not use infinite straight engine-duct
s with thin zero-angle-of-attack blades\, and so u
nfortunately good Computational AeroAcoustics (CAA
) simulations remain necessary in order to calcula
te the sound produced in the actual geometry and C
FD-computed flows within aircraft engines. CAA si
mulations push the limits of current computing pow
er\, and so try to be as efficient as possible\, w
hile also being precariously balanced on the knife
edge between being overly dissipative and being u
nstable.\n\nMost CAA simulations today use optimiz
ed Dispersion Relation Preserving (DRP) spatial de
rivatives\, which claim to be accurate down to 5 p
oints per wavelength. They aren't.\n\nMost CAA re
searchers are aware there are problems with numeri
cal instability near boundaries due to the use of
asymmetric stencils. There are a number of papers
proposing optimized spatial derivatives near bound
aries that are supposedly stable. They aren't.\n\
nMost CAA simulations use optimized Runge-Kutta ti
me stepping to evolve the solution forwards in tim
e\, (most commonly LDDRK56\, Low Dispersion and Di
ssipation Runge-Kutta 5stage/6stage)\, which are c
laimed to be optimized for wave propagation. They
aren't.\n\nThere is some theory of finite differe
nce schemes that are provably stable\, and which d
o work near boundaries. These are called Summatio
n By Parts (SBP) schemes\, with the important prop
erty that they have a discretized version of Integ
ration By Parts (IBP)\, which could be used for CA
A simulations. They aren't.\n\nThis talk will exp
lain some of the maths behind the claims above\, a
nd some recent work to bring adapt recent best pra
ctice (SBP) finite differences to CAA simulations.
LOCATION:CMS\, MR15
CONTACT:Matthew Priddin
END:VEVENT
END:VCALENDAR