Hi, everyone.
I have a question about the shake schemes in Charmm and AMBER.
I found two shake schemes exist
X0: coordinate at t0,
X1: coordinate at t0+dt, before shake correction
X2: coordinate at t0+dt, after shake correction
V0: velocity at t0;
V1: velocity at t0+dt, with shake correction;
X1=X0+V0*dt
X2=X1+SHKDIFF(X0,X1)
V1=(X2-X0)/dt -------------------- scheme I
or
V1=V0+(X2-X1)/dt ------------------ scheme II
Scheme I is used for single time step integrator
both in Charmm and AMBER.
Scheme II is used for multi-time step integrator in Charmm.
Can anyone explain to me why different shake schemes are used?
Xiaoliang
Received on Fri Feb 23 2001 - 09:13:07 PST
