Merge heads
Add intfTypes in elasticAnisotropicDiscrInterface, to allow for nonconforming interfaces.
Modify elasticAnisoSupergridMixedDiscr to make it possible to run odd order D+D-.
Add intfTypes in elasticAnisotropicSupergridMixedDiscr. Add example for curved interface with disontinuous material parameters.
Add dispersion analysis for the narrow scheme for the scalar acoustic equation for easy comparison. Turns out to be identical to staggered applied to the elastic equation with mu=0.
Add helper function that splits a four tensor into a cell array for a multiblock grid.
Add option for non-fully compatible in AnistropicDiscrCurveMultiblock
Add support for non-fully compatible scheme in elasticAnisotropicDiscrInterface.
Fix bug dc -> L in aging law mms. Add CFL to default dynamic rupture input file.
Add solver struct to dynamic rupture input struct.
Move mexOps struct to ops.mex
Update sinusoidalFault and sinusoidalFaultSG input files.
added input filed used for the initial parameter study that discovered the low normal stresss issue
created branch axel
changes made long ago, possibly for test fzero; have to commit this or hg complains.
add a function that uses fzero to solve for slip velocity
temp fix to not output failed timesteps -quasiStatic
deleted output fields that were used for debug on local machine
added t as a field to OUT
added poroelastic factor B
added a parameter to bound sigma
changed value for viscous parameter
changed viscosity
changed input parameters for a larger fault and trapezoidal shaped parameter 'a' profile
changed function for rate and state parameter profile
make VWLength an input variable
added linear viscous logic for tensile normal stress
Add timeStepper struct to input files in friction.input
Change return value of +friction input files from several structs to one struct with previous structs as fields.
Prepare to refactor dynamicRupture main
Fix sign mistake in rhsFullyDynamic.
update parameter names in some dynamic rupture input files
Add a bp3qd-like input file for Axel to use.
Merge heads
Add elastic-acoustic supergrid discr.
Bugfix prestress sign for thrust fauling in new BC input files.
Add input file for same BC as original config (traction bottom, displacement sides), but with slip data on sides.
Generalize tetonic struct to multiple motions. Add input file for BP3 with other sets of BC.
Add helper function for number of seconds per year.
Change plotOutput to plot log10 of slip velocity profile
Decrease faultFieldStride in BP3 default input file.
Add input files for smaller time-stepping tolerance, larger domain, and gentler stretching, to determine what is most important.
Add functions for psi-theta conversion. Add theta in plotOutput.
Add thrust fault option in input file
Add dipAngle as argument in BP3 input file
Lower PCG tolerance from 1-9 to 1e-10.
Clean upp saving and plotting of fields. Add option to save/plot normal stress.
Add saving and plotting of surface displacement
Lower pcg tolerance to 1e-9 and add input file for bigger domain.
Default pcg tolerance 1e-6 does not yield accurate results. Set it to 1e-8.
Add option to run use mex file for interior stencil in viscoelastic solver.
Optimize mex operator further. All left D1 operators are now applied in the same mex file.
Start optimizing the stencil-based elastic operator by essentially combining mex files to reduce number of reads and writes. Significant speedup observed.
Make GPU implementation run. Computes correctly. Unfortunately didnt manage to make preconditioning efficient.
Add openMP implementation of stencil-based elastic operator.
Implement first attempt at doing PCG on the GPU
Add option to use pcg instead of lu. PCG uses two previous time levels of full displacement field to extrapolate initial guess.
Merge heads
Add input file with 20x stretching
Make frictionCoefficient return a function handle that is used after one setup-call, for efficiency. Significant performance improvement.
Add timing of LU factorization
Replace fzero by a vectorized bisection method in solve for slip velocity. Huge performance improvement.
Allow for tectonic loading on interfaces as well as boundaries. Fix incorrect loading in SEAS_BP3QS input file.
Add red and blue slip curve plots in plotOutput
Change input files and rhs.quasiStatic to write to disk every x:th time step instead of at specified time points.
Fix plotting for vertical or near-vertical fault
Add input file for SEAS BP3 QS
Change sign convention and friction in quasi-static viscoelastic mms input so that it runs again. Expected convergence rates observed.
Include grid stretching in springSlider input file.
Fix plotting mistake that showed stressesmagnified by 1e6.
Change input file convention: add outputDir and MODE (plot or simulation) as input args. Some older input files are currently incompatible. Add SEAS-inspired input files.
Remove theta as state variable and clean up friction input options.
Speed up slip velocity solve by using value from neighbor grid point as initial guess.
Add input files for the spring-slider-like test setup.
Add SAT to friction law, making the semidiscrete problem energy stable.
Add sign change for shear stress so that slip velocity and shear stress have the same sign.
Set viscofriction default time-stepping tolerance to relative instead of absolute.
Include Erics sign change in the frictional force balance in rhsFullyDynamic
Include Erics comments in frictionCoefficient.m. Should change srtings here.
Include Erics change in default input, where L replaces dc.
Include Erics sign change in residual in force balance in frictionLaw.m
Use Erics comments in stateEvolution, where Lapusta2000 now means theta and Rice2001 means psi.
Include Erics bracketing + some additional changes for efficiency in slipVelocity.m
Some changes in ExampleForEric input file.
Precompute friction forcing outside friction law to save time.
Add output saving to quasi-static rhs. Add Lapusta2000 friction law.
Factor out slip velocity computation into viscofriction.slipVelocity
Add mms setup for viscoelastic quasistatic with friction (in package viscofriction)
Change basic plotting time unit to years
Allow etaInv to be a function handle in viscoElastic mms forcing computation
Merge heads
Add tectonic plate loading
Allow forcing in viscoElastic discr to be of multiblock type
Complete frictionLaw function
Add radiation damping
Remove some out-commented code. Switch to discontinuous viscous strains in mms without significant change in eyeball-norm.
Remove extra absolute value in frictionCoeff.m
Move viscofriction timeStepper options to hte input files.
Add non-mms viscoelastic input file with friction
Add Lapusta friction coeff