[Nek5000-users] **ERROR**: Failed in HMHOLTZ

nek5000-users at lists.mcs.anl.gov nek5000-users at lists.mcs.anl.gov
Thu Nov 6 05:28:23 CST 2014


Hi Wei,

Your Helmholtz tolerance of 1e-16 is too high (p22) -- velocity components start an interation from 5...4.e-2 residuals and converge to 4...2.e-16.

Once you lower it, make sure the Divergence tolerance (p21) is still an order or 2 magnitude higher.

Aleks
________________________________
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, November 06, 2014 5:21 AM
To: nek5000-users at lists.mcs.anl.gov
Subject: [Nek5000-users] **ERROR**: Failed in HMHOLTZ

Hi Neks

when I setup a profile 2D flow, there are informations like this, but the simulation is not crash,

Step    800, t= 8.0000000E-02, DT= 1.0000000E-04, C=  0.431 9.2444E+03 1.0117E+01
 Solving for fluid F T T
   800   100 **ERROR**: Failed in HMHOLTZ: VELX   3.7727E-16   5.3129E-02   1.0000E-16
   800   100 **ERROR**: Failed in HMHOLTZ: VELY   2.4683E-16   3.5181E-02   1.0000E-16
       800 U-PRES gmres:     23  5.6827E-15  1.0000E-14  9.6473E-07  9.5640E-01  2.4052E+00
        800   8.0000E-02  1.0044E+01 Fluid done
     800  8.00000000000E-02  6.70022578679E-01  1.99103625323E-01  4.70918953356E-01      1dragx
     800  8.00000000000E-02  4.12612202513E-01  5.10664067531E-01 -9.80518650173E-02      1dragy
 dump history points
 done :: dump history points


The CFL number is below than 0.5 already, the polynomial order is 7 in this case. when I choose a lower order p, this information will be disappeared after several  iteration steps.  I have checked the mesh looks fine, the parameters setting are copied from “ext_cyl”  with a smaller values of p21 to p25, I  don’t know what is wrong, should I ignore ?

thanks,

Wei

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/nek5000-users/attachments/20141106/1d452d9c/attachment.html>


More information about the Nek5000-users mailing list