[Nek5000-users] Error in Hmholtz solving

nek5000-users at lists.mcs.anl.gov nek5000-users at lists.mcs.anl.gov
Thu Jan 29 14:05:04 CST 2015


Tony,

Projection simply accelerates solution of the linear systems -- see:

http://www.mcs.anl.gov/~fischer/pubhtml/project.pdf

Dealiasing is invariably a good idea and not too expensive.  It just ensure proper
integration of the nonlinear terms and avoids some spurious eigenvalues.

J. Malm, P. Schlatter, P. F. Fischer, and D. S. Henningson. Stabilization of the spectral element
method in convection dominated flows by recovery of skew-symmetry. J. Sci. Comput., 57:254–
277, 2013.

The real question is whether to use filtering or not... I generally do, and I do so by setting
p103=0.01 and p101=0

in cases that are well-resolved.

Paul

________________________________________
From: nek5000-users-bounces at lists.mcs.anl.gov [nek5000-users-bounces at lists.mcs.anl.gov] on behalf of nek5000-users at lists.mcs.anl.gov [nek5000-users at lists.mcs.anl.gov]
Sent: Thursday, January 29, 2015 1:44 PM
To: nek5000-users at lists.mcs.anl.gov
Subject: Re: [Nek5000-users] Error in Hmholtz solving

Hi Paul,

Many thanks for your help. Yes, my simulation is scaled in convective time units. As you pointed out, the solutions are well converged.

I just have a question about the projection and dealiasing. I'm still not sure when I need to use them in my simulations. Say I'm doing simulations without any models, do I need to turn on projection and dealiasing? I actually don't know how they will affect the solutions, would projection or dealiasing (off, old or new?) change the solutions? It would be great if you could briefly explain this bit to me.

Thank you very much.

Best regards,
Tony

________________________________________
From: nek5000-users-bounces at lists.mcs.anl.gov <nek5000-users-bounces at lists.mcs.anl.gov> on behalf of nek5000-users-request at lists.mcs.anl.gov <nek5000-users-request at lists.mcs.anl.gov>
Sent: 29 January 2015 18:00
To: nek5000-users at lists.mcs.anl.gov
Subject: Nek5000-users Digest, Vol 71, Issue 16

Send Nek5000-users mailing list submissions to
        nek5000-users at lists.mcs.anl.gov

To subscribe or unsubscribe via the World Wide Web, visit
        https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
or, via email, send a message with subject or body 'help' to
        nek5000-users-request at lists.mcs.anl.gov

You can reach the person managing the list at
        nek5000-users-owner at lists.mcs.anl.gov

When replying, please edit your Subject line so it is more specific
than "Re: Contents of Nek5000-users digest..."


Today's Topics:

   1. Error in Hmholtz solving (nek5000-users at lists.mcs.anl.gov)
   2. Re: Error in Hmholtz solving (nek5000-users at lists.mcs.anl.gov)


----------------------------------------------------------------------

Message: 1
Date: Thu, 29 Jan 2015 15:50:43 +0000
From: nek5000-users at lists.mcs.anl.gov
To: "nek5000-users at lists.mcs.anl.gov"
        <nek5000-users at lists.mcs.anl.gov>
Subject: [Nek5000-users] Error in Hmholtz solving
Message-ID:
        <mailman.14256.1422546648.3424.nek5000-users at lists.mcs.anl.gov>
Content-Type: text/plain; charset="iso-8859-1"

Dear Neks,


I'm getting the error message of '**ERROR**: Failed in HMHOLTZ: ' when running my simulation. I've checked the subroutine 'cggo' in hmholtz.f where the error message is printed. I understand that when the tolerance is not satisfied, the code would output this error message. But what I'm wondering is why the maximum iteration number is only 100. In the subroutine 'cggo', niter is equal to min(maxit,maxcg) where maxit=200 and maxcg=900. I can't see other places where maxit is restricted to only 100. So why the maximum iteration number is 100 when solving the Hmholtz equation?


Also, is it right that if I increase the maximum number of iterations to certain point where the tolerance is satisfied, I won't get the error message?


Hope anyone can help me on this. Thank you very much in advance.


Best regards,

Tony


Example of the error message:

Step     52, t= 2.5600000E+01, DT= 5.0000000E-02, C=  2.689 5.6747E+02 1.0237E+01
 Solving for fluid T T T
       52 PRES gmres:   29  9.7057E-07  1.0000E-06  1.5219E-03  1.0926E+00  2.1480E+00    F
         52    Hmholtz VELX:     99   8.5954E-10   6.0898E-03   1.0000E-09
    52   100 **ERROR**: Failed in HMHOLTZ: VELY   1.7088E-09   2.5049E-02   1.0000E-09
    52   100 **ERROR**: Failed in HMHOLTZ: VELZ   1.4665E-09   2.5165E-02   1.0000E-09
               L1/L2 DIV(V)    :   2.7573E-18   9.1661E-04
               L1/L2 QTL       :   0.0000E+00   0.0000E+00
               L1/L2 DIV(V)-QTL:   2.7573E-18   9.1661E-04
         52   2.5600E+01  1.0094E+01 Fluid done
Step     53, t= 2.5650000E+01, DT= 5.0000000E-02, C=  2.697 5.7766E+02 1.0186E+01
 Solving for fluid T T T
       53 PRES gmres:   28  9.6076E-07  1.0000E-06  1.1322E-03  1.0558E+00  2.0542E+00    F
         53    Hmholtz VELX:     98   9.4891E-10   5.8783E-03   1.0000E-09
    53   100 **ERROR**: Failed in HMHOLTZ: VELY   1.6450E-09   2.4941E-02   1.0000E-09
    53   100 **ERROR**: Failed in HMHOLTZ: VELZ   1.4160E-09   2.5049E-02   1.0000E-09
               L1/L2 DIV(V)    :   2.4695E-18   1.0845E-03
               L1/L2 QTL       :   0.0000E+00   0.0000E+00
               L1/L2 DIV(V)-QTL:   2.4695E-18   1.0845E-03
         53   2.5650E+01  9.9794E+00 Fluid done
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/nek5000-users/attachments/20150129/e21e4255/attachment-0001.html>

------------------------------

Message: 2
Date: Thu, 29 Jan 2015 16:00:45 +0000
From: nek5000-users at lists.mcs.anl.gov
To: "nek5000-users at lists.mcs.anl.gov"
        <nek5000-users at lists.mcs.anl.gov>
Subject: Re: [Nek5000-users] Error in Hmholtz solving
Message-ID:
        <mailman.14259.1422547425.3424.nek5000-users at lists.mcs.anl.gov>
Content-Type: text/plain; charset="iso-8859-1"


Tony,

It looks like your Helmholtz tolerance is too tight.

If your problem is scaled in convective time units with characteristic length scale 1, timescale 1, and visc ~ 1/Re,
then typical Helmholtz tolerances are around 1.e-7.

You might also try turning on projection ---

set p93=20 (and lmxprev=20 in SIZE), p94=5 and p95=5 in the rea file.

No guarantees at this point because there seems to be something flaky in the projection routines
these days under certain runtime combinations (e.g., I think splitting + heat transfer +
velocity projection does not work)...  It's on my todo list to clean these up.

Note that your helmholtz solves are pretty well converged... residual is 1.7e-9, etc.
Projection should easily put you over the top

Paul

________________________________
From: nek5000-users-bounces at lists.mcs.anl.gov [nek5000-users-bounces at lists.mcs.anl.gov] on behalf of nek5000-users at lists.mcs.anl.gov [nek5000-users at lists.mcs.anl.gov]
Sent: Thursday, January 29, 2015 9:50 AM
To: nek5000-users at lists.mcs.anl.gov
Subject: [Nek5000-users] Error in Hmholtz solving


Dear Neks,


I'm getting the error message of '**ERROR**: Failed in HMHOLTZ: ' when running my simulation. I've checked the subroutine 'cggo' in hmholtz.f where the error message is printed. I understand that when the tolerance is not satisfied, the code would output this error message. But what I'm wondering is why the maximum iteration number is only 100. In the subroutine 'cggo', niter is equal to min(maxit,maxcg) where maxit=200 and maxcg=900. I can't see other places where maxit is restricted to only 100. So why the maximum iteration number is 100 when solving the Hmholtz equation?


Also, is it right that if I increase the maximum number of iterations to certain point where the tolerance is satisfied, I won't get the error message?


Hope anyone can help me on this. Thank you very much in advance.


Best regards,

Tony


Example of the error message:

Step     52, t= 2.5600000E+01, DT= 5.0000000E-02, C=  2.689 5.6747E+02 1.0237E+01
 Solving for fluid T T T
       52 PRES gmres:   29  9.7057E-07  1.0000E-06  1.5219E-03  1.0926E+00  2.1480E+00    F
         52    Hmholtz VELX:     99   8.5954E-10   6.0898E-03   1.0000E-09
    52   100 **ERROR**: Failed in HMHOLTZ: VELY   1.7088E-09   2.5049E-02   1.0000E-09
    52   100 **ERROR**: Failed in HMHOLTZ: VELZ   1.4665E-09   2.5165E-02   1.0000E-09
               L1/L2 DIV(V)    :   2.7573E-18   9.1661E-04
               L1/L2 QTL       :   0.0000E+00   0.0000E+00
               L1/L2 DIV(V)-QTL:   2.7573E-18   9.1661E-04
         52   2.5600E+01  1.0094E+01 Fluid done
Step     53, t= 2.5650000E+01, DT= 5.0000000E-02, C=  2.697 5.7766E+02 1.0186E+01
 Solving for fluid T T T
       53 PRES gmres:   28  9.6076E-07  1.0000E-06  1.1322E-03  1.0558E+00  2.0542E+00    F
         53    Hmholtz VELX:     98   9.4891E-10   5.8783E-03   1.0000E-09
    53   100 **ERROR**: Failed in HMHOLTZ: VELY   1.6450E-09   2.4941E-02   1.0000E-09
    53   100 **ERROR**: Failed in HMHOLTZ: VELZ   1.4160E-09   2.5049E-02   1.0000E-09
               L1/L2 DIV(V)    :   2.4695E-18   1.0845E-03
               L1/L2 QTL       :   0.0000E+00   0.0000E+00
               L1/L2 DIV(V)-QTL:   2.4695E-18   1.0845E-03
         53   2.5650E+01  9.9794E+00 Fluid done
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/nek5000-users/attachments/20150129/b14c95c6/attachment-0001.html>

------------------------------

_______________________________________________
Nek5000-users mailing list
Nek5000-users at lists.mcs.anl.gov
https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users


End of Nek5000-users Digest, Vol 71, Issue 16
*********************************************
_______________________________________________
Nek5000-users mailing list
Nek5000-users at lists.mcs.anl.gov
https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users


More information about the Nek5000-users mailing list