Can the DAE equation type be supplied via a run-time option? Shri -----Original Message----- From: barry smith <[email protected]> Date: Thursday, June 18, 2015 at 11:08 AM To: Italo Tasso <[email protected]>, "Constantinescu, Emil M." <[email protected]> Cc: "[email protected]" <[email protected]> Subject: Re: [petsc-users] arkimex rejecting all dt in petsc 3.6.0

Here is the response from Emil who made the changes to the code

If you are solving a DAE then in the new version, we introduced a new flag that distinguishes between ODEs and DAEs leading to different semantics. For DAEs this is needed b/c it may not have consistent initial conditions.

If solving DAEs, the user has to set the EquationType appropriately: e.g., ierr = TSSetEquationType(ts,TS_EQ_DAE_IMPLICIT_INDEX1);CHKERRQ(ierr);

This is documented in the doc, but I didn't add it to the changelog (Jed pointed it out to me).

Currently it only affects -ts_type arkimex. I'll work with Satish to add it to the changelog.

*If that's not the case:* let me know and we can dig deeper.

Emil

...

On Jun 18, 2015, at 5:48 AM, Italo Tasso <[email protected]> wrote:

I just upgraded to 3.6.0 and my code stopped working. All dt are rejected. I used the same configure line, same code, same everything.

With 3.5.4 I get:

0 TS dt 1e-06 time 0 0 SNES Function norm 2.549981005316e+05 1 SNES Function norm 6.107056905987e-03 2 SNES Function norm 1.483881932064e-10 3 SNES Function norm 9.122873794272e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 3 0 SNES Function norm 7.790429171165e+04 1 SNES Function norm 7.289068747803e-04 2 SNES Function norm 8.227639633330e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 2 0 SNES Function norm 1.154356516184e+05 1 SNES Function norm 2.309925413255e-03 2 SNES Function norm 6.382141981406e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 2 TSAdapt 'basic': step 0 accepted t=0 + 1.000e-06 wlte=0.000654 family='arkimex' scheme=0:'3' dt=1.000e-05

With 3.6.0 I get:

0 TS dt 1e-06 time 0 0 SNES Function norm 2.549981005316e+05 1 SNES Function norm 6.107056925316e-03 2 SNES Function norm 1.519319591792e-10 3 SNES Function norm 9.070104116945e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 3 0 SNES Function norm 7.790429171165e+04 1 SNES Function norm 6.942541792651e-04 2 SNES Function norm 8.458781909516e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 2 0 SNES Function norm 1.154356516184e+05 1 SNES Function norm 2.287202942961e-03 2 SNES Function norm 6.585201377573e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 2 TSAdapt 'basic': step 0 rejected t=0 + 1.000e-06 wlte= 324 family='arkimex' scheme=0:'3' dt=1.311e-07

Any ideas? I attached the full output.

Options I use:

-ts_view -ts_type arkimex -ts_arkimex_fully_implicit -ts_adapt_monitor -ts_monitor -snes_monitor -snes_converged_reason -ksp_type preonly -pc_type lu -pc_factor_mat_solver_package mumps -snes_rtol 0 -snes_atol 1e-10 -snes_stol 0

<petsc354.txt><petsc360.txt>

Thank you Barry and Emil. It works, but it takes twice as many steps than before. Was it wrong before? Should I have been using this option all along? I am solving the Navier-Stokes equations, full implicit and non-linear. I also noticed two things: In the 3.6.0 output, ts_monitor skips timestep #1. If I use the equation type in 3.5.4, I get segmentation fault. I attached the outputs. On Thu, Jun 18, 2015 at 1:24 PM, Emil Constantinescu <[email protected]> wrote:

No, I'm not sure if it should. It belongs to the same category as "problem_type" == TS_LINEAR / == TS_NONLINEAR.

Emil

On 6/18/15 11:15 AM, Abhyankar, Shrirang G. wrote:

...

Can the DAE equation type be supplied via a run-time option?

Shri

-----Original Message----- From: barry smith <[email protected]> Date: Thursday, June 18, 2015 at 11:08 AM To: Italo Tasso <[email protected]>, "Constantinescu, Emil M." <[email protected]> Cc: "[email protected]" <[email protected]> Subject: Re: [petsc-users] arkimex rejecting all dt in petsc 3.6.0

...

Here is the response from Emil who made the changes to the code

If you are solving a DAE then in the new version, we introduced a new flag that distinguishes between ODEs and DAEs leading to different semantics. For DAEs this is needed b/c it may not have consistent initial conditions.

If solving DAEs, the user has to set the EquationType appropriately: e.g., ierr = TSSetEquationType(ts,TS_EQ_DAE_IMPLICIT_INDEX1);CHKERRQ(ierr);

This is documented in the doc, but I didn't add it to the changelog (Jed pointed it out to me).

Currently it only affects -ts_type arkimex. I'll work with Satish to add it to the changelog.

*If that's not the case:* let me know and we can dig deeper.

Emil

On Jun 18, 2015, at 5:48 AM, Italo Tasso <[email protected]> wrote:

...

I just upgraded to 3.6.0 and my code stopped working. All dt are rejected. I used the same configure line, same code, same everything.

With 3.5.4 I get:

0 TS dt 1e-06 time 0 0 SNES Function norm 2.549981005316e+05 1 SNES Function norm 6.107056905987e-03 2 SNES Function norm 1.483881932064e-10 3 SNES Function norm 9.122873794272e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 3 0 SNES Function norm 7.790429171165e+04 1 SNES Function norm 7.289068747803e-04 2 SNES Function norm 8.227639633330e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 2 0 SNES Function norm 1.154356516184e+05 1 SNES Function norm 2.309925413255e-03 2 SNES Function norm 6.382141981406e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 2 TSAdapt 'basic': step 0 accepted t=0 + 1.000e-06 wlte=0.000654 family='arkimex' scheme=0:'3' dt=1.000e-05

With 3.6.0 I get:

0 TS dt 1e-06 time 0 0 SNES Function norm 2.549981005316e+05 1 SNES Function norm 6.107056925316e-03 2 SNES Function norm 1.519319591792e-10 3 SNES Function norm 9.070104116945e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 3 0 SNES Function norm 7.790429171165e+04 1 SNES Function norm 6.942541792651e-04 2 SNES Function norm 8.458781909516e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 2 0 SNES Function norm 1.154356516184e+05 1 SNES Function norm 2.287202942961e-03 2 SNES Function norm 6.585201377573e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 2 TSAdapt 'basic': step 0 rejected t=0 + 1.000e-06 wlte= 324 family='arkimex' scheme=0:'3' dt=1.311e-07

Any ideas? I attached the full output.

Options I use:

-ts_view -ts_type arkimex -ts_arkimex_fully_implicit -ts_adapt_monitor -ts_monitor -snes_monitor -snes_converged_reason -ksp_type preonly -pc_type lu -pc_factor_mat_solver_package mumps -snes_rtol 0 -snes_atol 1e-10 -snes_stol 0

<petsc354.txt><petsc360.txt>

I understand. Thanks again. On Thu, Jun 18, 2015 at 10:58 PM, Emil Constantinescu <[email protected]> wrote:

The results enclosed for 3.5.4 (354.txt) are not correct. I can tell that by looking at the first step. Most implicit schemes in arkimex have the first stage explicit and therefore need to be "started" somehow when dealing with algebraic constraints. In this case, a different scheme is used to start the integration.

Note that in 3.6 and 3.5.4_SEGV the first step uses scheme 1bee (scheme=0:'1bee') and then switches to '3'. This is the correct sequence otherwise inconsistent initial conditions are inadvertently used or the solver thinks it solves an ODE. I'm not sure why it's crashing in 3.5.4_SEGV. Configuring with --with-debugging=yes may reveal more.

Emil

On 6/18/15 4:10 PM, Italo Tasso wrote:

...

Thank you Barry and Emil.

It works, but it takes twice as many steps than before.

Was it wrong before? Should I have been using this option all along?

I am solving the Navier-Stokes equations, full implicit and non-linear.

I also noticed two things:

In the 3.6.0 output, ts_monitor skips timestep #1.

If I use the equation type in 3.5.4, I get segmentation fault.

I attached the outputs.

On Thu, Jun 18, 2015 at 1:24 PM, Emil Constantinescu <[email protected] <mailto:[email protected]>> wrote:

No, I'm not sure if it should. It belongs to the same category as "problem_type" == TS_LINEAR / == TS_NONLINEAR.

Emil

On 6/18/15 11:15 AM, Abhyankar, Shrirang G. wrote:

Can the DAE equation type be supplied via a run-time option?

Shri

-----Original Message----- From: barry smith <[email protected] <mailto:[email protected]

...

...

Date: Thursday, June 18, 2015 at 11:08 AM To: Italo Tasso <[email protected] <mailto:[email protected]>>, "Constantinescu, Emil M." <[email protected] <mailto:[email protected]>> Cc: "[email protected] <mailto:[email protected]>" <[email protected] <mailto:[email protected]>> Subject: Re: [petsc-users] arkimex rejecting all dt in petsc 3.6.0

Here is the response from Emil who made the changes to the code

If you are solving a DAE then in the new version, we introduced a new flag that distinguishes between ODEs and DAEs leading to different semantics. For DAEs this is needed b/c it may not have consistent initial conditions.

If solving DAEs, the user has to set the EquationType appropriately: e.g., ierr = TSSetEquationType(ts,TS_EQ_DAE_IMPLICIT_INDEX1);CHKERRQ(ierr);

This is documented in the doc, but I didn't add it to the changelog (Jed pointed it out to me).

Currently it only affects -ts_type arkimex. I'll work with Satish to add it to the changelog.

*If that's not the case:* let me know and we can dig deeper.

Emil

On Jun 18, 2015, at 5:48 AM, Italo Tasso <[email protected] <mailto:[email protected]>> wrote:

I just upgraded to 3.6.0 and my code stopped working. All dt are rejected. I used the same configure line, same code, same everything.

With 3.5.4 I get:

0 TS dt 1e-06 time 0 0 SNES Function norm 2.549981005316e+05 1 SNES Function norm 6.107056905987e-03 2 SNES Function norm 1.483881932064e-10 3 SNES Function norm 9.122873794272e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 3 0 SNES Function norm 7.790429171165e+04 1 SNES Function norm 7.289068747803e-04 2 SNES Function norm 8.227639633330e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 2 0 SNES Function norm 1.154356516184e+05 1 SNES Function norm 2.309925413255e-03 2 SNES Function norm 6.382141981406e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 2 TSAdapt 'basic': step 0 accepted t=0 + 1.000e-06 wlte=0.000654 family='arkimex' scheme=0:'3' dt=1.000e-05

With 3.6.0 I get:

0 TS dt 1e-06 time 0 0 SNES Function norm 2.549981005316e+05 1 SNES Function norm 6.107056925316e-03 2 SNES Function norm 1.519319591792e-10 3 SNES Function norm 9.070104116945e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 3 0 SNES Function norm 7.790429171165e+04 1 SNES Function norm 6.942541792651e-04 2 SNES Function norm 8.458781909516e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 2 0 SNES Function norm 1.154356516184e+05 1 SNES Function norm 2.287202942961e-03 2 SNES Function norm 6.585201377573e-11 Nonlinear solve converged due to CONVERGED_FNORM_ABS iterations 2 TSAdapt 'basic': step 0 rejected t=0 + 1.000e-06 wlte= 324 family='arkimex' scheme=0:'3' dt=1.311e-07

Any ideas? I attached the full output.

Options I use:

-ts_view -ts_type arkimex -ts_arkimex_fully_implicit -ts_adapt_monitor -ts_monitor -snes_monitor -snes_converged_reason -ksp_type preonly -pc_type lu -pc_factor_mat_solver_package mumps -snes_rtol 0 -snes_atol 1e-10 -snes_stol 0

<petsc354.txt><petsc360.txt>