<div dir="ltr">Hi,<br><br>you are using DZVP-MOLOPT-GTH basis sets instead of their SR (Short-Range) variants DZVP-MOLOPT-SR-GTH which are better suited for condensed phase simulations. I think this is the major reason for your convergence problems, because an explicit water simulation (box filled with H2O molecules) will most likely show a similar behavior.<br>You forgot to drop the EPS_* thesholds<br><br> DENSITY_CUTOFF 1.0000000000000000E-10<br> GRADIENT_CUTOFF 1.0000000000000000E-10<br> TAU_CUTOFF 1.0000000000000000E-10<br><br>in the &XC section.<br><br>You may also add in the &DFT section<br> &PRINT<br> &SCCS on<br> &EACH<br> QS_SCF 1<br> &END EACH<br> &END SCCS<br> &END PRINT<br>for a more detailed SCCS output.<br><br>Finally, I would use for MAX_SCF in the &SCF section a much smaller value and a larger threshold value for EPS_SCF, e.g.<br> &SCF<br> MAX_SCF 31<br> EPS_SCF 3.0E-7<br>The large MAX_SCF suppresses the outer SCF cycle and thus an improved preconditioning for OT.<br><br>Matthias<br><br><br>On Monday, 23 March 2015 02:27:21 UTC+1, Satish Kumar wrote:<blockquote class="gmail_quote" style="margin: 0;margin-left: 0.8ex;border-left: 1px #ccc solid;padding-left: 1ex;"><div dir="ltr">Hi <div>Thank you for your response. I tried what you suggested.</div><div><br></div><div>1) When I use default values (refer in.defaults and out.default) for all tags within SCCS section, I do see polarization energy converging to some value (which I do not if is right) but the electronic minimizations do not converge. </div><div><br></div><div>2) How does the electronic minimization work? I can see that vacuum bases electronic minimization takes place till EPS_SCF criteria within SCCS section is met. Then SCCS bases electronic minimzation occurs. After this is it just EPS_SCCS that determines when electronic minimization finishes? If not how does it work.</div><div><br></div><div>3) I have attaches two more input (in.*) and output (out.*). When I use EPS_SCF within SCCS ~E-13, I do not see any polarization energy printed (corresponding files are *.pol_energy). However, when I use EPS_SCF as default value, as above, I do see polarization enegy printed but electronic minimization doesn't converge.</div><div><br></div><div>4) Also, when I did a 1000 step electronic minimization without outer-scf loop, the energy corresponding to the convergence column in the output does not change after reaching a certain value. Corresponding seetings were: <span style="line-height:18px">EPS_SCF 1.0E-11 and within SCCS section: </span><span style="line-height:18px">EPS_SCCS </span><span style="line-height:18px"> 1.0</span><span style="line-height:18px">E-11, </span><span style="line-height:18px">EPS_SCF </span><span style="line-height:18px"> </span><span style="line-height:18px">1.01E-05. </span>I do not understand what is happening.</div>
<div><br></div><div>Thank you.<br><br>On Monday, March 16, 2015 at 1:49:03 PM UTC-4, Matthias Krack wrote:<blockquote class="gmail_quote" style="margin:0;margin-left:0.8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="auto"><div>Hi,</div><div><br></div><div>Did you check, if the inner SCCS cycle for the polarisation charge converges in each SCF iteration step? (EPS_SCCS should be tight 1E-8-1E-10)<br>My further suggestions are: </div><div>- the DERIVATIVE_METHOD FFT requires quite high cutoff values for numerical reasons, possibly 600-800 Ry. In this respect CD5 is often a better choice.</div><div>- drop the EPS_* thresholds <a href="http://e.g.in" rel="nofollow" target="_blank" onmousedown="this.href='http://www.google.com/url?q\75http%3A%2F%2Fe.g.in\46sa\75D\46sntz\0751\46usg\75AFQjCNFo0zkkuxBpAPL_Pi2EarFGESWudg';return true;" onclick="this.href='http://www.google.com/url?q\75http%3A%2F%2Fe.g.in\46sa\75D\46sntz\0751\46usg\75AFQjCNFo0zkkuxBpAPL_Pi2EarFGESWudg';return true;">e.g.in</a> the XC section and use just EPS_DEFAULT 1.0E-14 and possibly MAP_CONSISTENT.</div><div><br></div><div>Matthias</div><div><br></div><div>Von meinem iPod gesendet</div><div><br>Am 14.03.2015 um 23:02 schrieb Satish Kumar <<a rel="nofollow">sati...@gmail.com</a>>:<br><br></div><blockquote type="cite"><div><div dir="ltr"><div>Thank you Matthias, I now use the method and FATTEBERT-GYGI works but even for just a water molecule in a cell, the calculation runs forever and does not converge. I first restarted calculation from wavefunction of gas phase calculation, then tried all the tags with the default values in SCCS section. Finally I increased EPS_SCF within SCCS to 1.0E-04 thinking it may improve convergence. But none of them helped. I am attaching the result of grep-ing "OT CG" from the output file. Till EPS_SCF reaches 1.0E-04, the convergence is fine but I think when the SCCS loop begins (I don't actually find any explicit mention with "Print medium" about when SCF loops corresponding to SCCS begins), the convergence is very oscillating. Is there anything I can do to improve convergence with SCCS.</div><div><br></div><div><br></div><div>grep "OT CG" output:</div><div><br></div><div><div> OT CG 0.64E-01 1.2 0.38492654 -16.5223537330 -1.65E+01</div><div> 4 OT CG 0.47E-01 1.4 0.33467993 -16.8712267043 -3.49E-01</div><div> 7 OT CG 0.37E-01 1.4 0.18362126 -17.1265030091 -2.55E-01</div><div> 10 OT CG 0.31E-01 1.4 0.06956912 -17.1886589747 -6.22E-02</div><div> 13 OT CG 0.74E-01 1.3 0.04840266 -17.2094960351 -2.08E-02</div><div> 16 OT CG 0.72E-01 1.4 0.01997461 -17.2194004920 -9.90E-03</div><div> 19 OT CG 0.40E-01 1.4 0.01155289 -17.2203057880 -9.05E-04</div><div> 22 OT CG 0.87E-01 1.4 0.00522995 -17.2209729877 -6.67E-04</div><div> 25 OT CG 0.61E-01 1.4 0.00301797 -17.2210686037 -9.56E-05</div><div> 28 OT CG 0.51E-01 1.4 0.00192432 -17.2210954837 -2.69E-05</div><div> 31 OT CG 0.58E-01 1.4 0.00085921 -17.2211078888 -1.24E-05</div><div> 34 OT CG 0.72E-01 1.5 0.00045713 -17.2211109595 -3.07E-06 #I think beyond this SCCS starts</div><div> 37 OT CG 0.12E+00 10.5 0.03513936 -16.9374989979 2.84E-01</div><div> 40 OT CG 0.19E+00 12.8 0.53438787 -13.0284958890 3.91E+00</div><div> 43 OT CG 0.15E+00 12.6 0.18217280 -16.4652852134 -3.44E+00</div><div> 46 OT CG 0.71E-01 11.4 0.07232781 -16.6107052961 -1.45E-01</div><div> 49 OT CG 0.67E-01 11.0 0.02449012 -16.6312618955 -2.06E-02</div><div> 52 OT CG 0.11E+00 9.7 0.00857426 -16.6349458760 -3.68E-03</div><div> 55 OT CG 0.48E+00 9.4 0.01771721 -16.6370477486 -2.10E-03</div><div> 58 OT CG 0.12E+01 11.8 0.06495594 -16.6625230341 -2.55E-02</div><div> 61 OT CG 0.10E-01 13.1 0.05789947 -16.6669359830 -4.41E-03</div><div> 64 OT CG 0.45E-01 11.0 0.03152062 -16.6795843687 -1.26E-02</div><div> 67 OT CG 0.12E+00 9.9 0.01221376 -16.6858425676 -6.26E-03</div><div> 70 OT CG 0.17E+00 9.6 0.01215730 -16.6872572020 -1.41E-03</div><div> 73 OT CG 0.53E+00 11.9 0.03221792 -16.6915824133 -4.33E-03</div><div> 76 OT CG 0.16E+00 10.5 0.04503294 -16.7005483657 -8.97E-03</div><div> 79 OT CG 0.65E+00 15.5 0.19838443 -16.6988538474 1.69E-03</div><div> 82 OT CG 0.10E+01 13.5 0.69320124 -11.0634009652 5.64E+00</div><div> 85 OT CG 0.78E+00 34.4 0.57682957 -11.1232371185 -5.98E-02</div><div> 88 OT CG 0.53E+00 34.2 0.66048561 -10.3342895296 7.89E-01</div><div> 91 OT CG 0.34E+00 12.4 0.41865566 -15.7784514964 -5.44E+00</div><div> 94 OT CG 0.10E+00 34.0 0.29143613 -13.0784219668 2.70E+00</div><div> 97 OT CG 0.88E-01 12.5 0.24158684 -16.5534274750 -3.48E+00</div><div> 100 OT CG 0.47E-01 12.2 0.05092026 -16.7441647536 -1.91E-01</div><div> 1 OT CG 0.64E-01 21.9 0.28419623 -16.7524473550 -8.28E-03</div><div> 4 OT CG 0.58E-01 10.5 0.18454413 -17.0578460723 -3.05E-01</div><div> 7 OT CG 0.18E-01 9.4 0.14071768 -17.1058023331 -4.80E-02</div><div> 10 OT CG 0.11E-01 10.0 0.11841442 -17.1118483343 -6.05E-03</div><div> 13 OT CG 0.89E-04 9.3 0.11828280 -17.1118213718 2.70E-05</div><div> 16 OT CG 0.14E-03 7.1 0.11807277 -17.1117763182 4.51E-05</div><div> 19 OT CG 0.11E-05 6.7 0.11807112 -17.1117758246 4.94E-07</div><div> 22 OT CG 0.18E-05 4.9 0.11806842 -17.1117753439 4.81E-07</div><div> 25 OT CG 0.29E-05 5.6 0.11806413 -17.1117743931 9.51E-07</div><div> 28 OT CG 0.47E-05 5.6 0.11805727 -17.1117728613 1.53E-06</div><div> 31 OT CG 0.75E-05 6.9 0.11804630 -17.1117704080 2.45E-06</div><div> 34 OT CG 0.12E-04 5.8 0.11802874 -17.1117665058 3.90E-06</div><div> 37 OT CG 0.19E-04 6.3 0.11800066 -17.1117602075 6.30E-06</div><div> 40 OT CG 0.31E-04 6.5 0.11795573 -17.1117500820 1.01E-05</div><div> 43 OT CG 0.49E-04 6.5 0.11788392 -17.1117336756 1.64E-05</div><div> 46 OT CG 0.78E-04 6.8 0.11776915 -17.1117069099 2.68E-05</div></div><div><br></div><div>After 92 optimization steps:</div><div><br></div><div><div>-------- Informations at step = 92 ------------</div><div> Optimization Method = BFGS</div><div> Total Energy = -17.2378273186</div><div> Real energy change = -0.0000433883</div><div> Predicted change in energy = -0.0000203875</div><div> Scaling factor = 0.7336561632</div><div> Step size = 0.0195987283</div><div> Trust radius = 0.4724315332</div><div> Decrease in energy = NO</div><div> Used time = 179.174</div><div><br></div><div> Convergence check :</div><div> Max. step size = 0.0195987283</div><div> Conv. limit for step size = 0.0030000000</div><div> Convergence in step size = NO</div><div> RMS step size = 0.0103012527</div><div> Conv. limit for RMS step = 0.0015000000</div><div> Convergence in RMS step = NO</div><div> Max. gradient = 0.0043764738</div><div> Conv. limit for gradients = 0.0009725000</div><div> Conv. for gradients = NO</div><div> RMS gradient = 0.0023573863</div><div> Conv. limit for RMS grad. = 0.0003000000</div><div> Conv. for gradients = NO</div><div> -----------------------------<wbr>----------------------</div></div><div><br></div><div>Input file:</div><div><br></div><div>&GLOBAL</div><div> PROJECT_NAME 4x4</div><div> RUN_TYPE GEO_OPT</div><div> &END GLOBAL</div><div> &MOTION</div><div> &GEO_OPT</div><div> OPTIMIZER BFGS</div><div> MAX_ITER 200</div><div> MAX_FORCE 9.7249999999999995E-04</div><div> STEP_START_VAL 90</div><div> &END GEO_OPT</div><div> &END MOTION</div><div> &FORCE_EVAL</div><div> METHOD QS</div><div> &DFT</div><div> BASIS_SET_FILE_NAME ./GTH_BASIS_SETS_5-12-10</div><div> POTENTIAL_FILE_NAME ./GTH_POTENTIALS_5-12-10</div><div> WFN_RESTART_FILE_NAME x.wfn</div><div> &SCF</div><div> MAX_SCF 100</div><div> EPS_SCF 1.0000000000000001E-05</div><div> SCF_GUESS ATOMIC</div><div> &OT T</div><div> MINIMIZER CG</div><div> LINESEARCH 3PNT</div><div> PRECONDITIONER FULL_SINGLE_INVERSE</div><div> &END OT</div><div> &OUTER_SCF T</div><div> EPS_SCF 1.0000000000000001E-05</div><div> MAX_SCF 20</div><div> &END OUTER_SCF</div><div> &END SCF</div><div> &QS</div><div> EXTRAPOLATION ASPC</div><div> EXTRAPOLATION_ORDER 3</div><div> &END QS</div><div> &MGRID</div><div> NGRIDS 5</div><div> CUTOFF 4.8000000000000000E+02</div><div> &END MGRID</div><div> &XC</div><div> DENSITY_CUTOFF 1.0000000000000000E-10</div><div> GRADIENT_CUTOFF 1.0000000000000000E-10</div><div> TAU_CUTOFF 1.0000000000000000E-10</div><div> &XC_FUNCTIONAL NO_SHORTCUT</div><div> &PBE T</div><div> &END PBE</div><div> &END XC_FUNCTIONAL</div><div> &END XC</div><div> &PRINT</div><div> &END PRINT</div><div> &SCCS T</div><div> DERIVATIVE_METHOD FFT</div><div> DIELECTRIC_CONSTANT 7.8400000000000006E+01</div><div> EPS_SCCS 9.9999999999999995E-07</div><div> EPS_SCF 5.0000000000000001E-04</div><div> MAX_ITER 100</div><div> METHOD FATTEBERT-GYGI</div><div> MIXING 5.9999999999999998E-01</div><div> &FATTEBERT-GYGI</div><div> BETA 1.7000000000000000E+00</div><div> RHO_ZERO 5.9999999999999995E-04</div><div> &END FATTEBERT-GYGI</div><div> &END SCCS</div><div> &END DFT</div><div> &SUBSYS</div><div> &CELL</div><div> A 1.2000000000000000E+01 0.0000000000000000E+00 0.0000000000000000E+00</div><div> B 0.0000000000000000E+00 1.3000000000000000E+01 0.0000000000000000E+00</div><div> C 0.0000000000000000E+00 0.0000000000000000E+00 1.4000000000000000E+01</div><div> MULTIPLE_UNIT_CELL 1 1 1</div><div> &END CELL</div><div> &COORD</div><div>O -7.5733907062610586E-02 8.9770405110133449E-03 1.2904624124972142E-01</div><div>H 6.5401889552783654E-01 -5.0597734685847906E-01 5.1647932522805295E-01</div><div>H -8.1191163881804929E-01 -9.1336870679231297E-02 7.6138190132564698E-01</div><div> &END COORD</div><div>{</div><div><br></div><div>basis and potential data</div><div><br></div><div>}</div><div> &TOPOLOGY</div><div> NUMBER_OF_ATOMS 3</div><div> MULTIPLE_UNIT_CELL 1 1 1</div><div> &END TOPOLOGY</div><div> &END SUBSYS</div><div> &END FORCE_EVAL</div><br>On Friday, March 13, 2015 at 2:17:54 AM UTC-4, Matthias Krack wrote:<blockquote class="gmail_quote" style="margin:0;margin-left:0.8ex;border-left:1px #ccc solid;padding-left:1ex">Hi,
<br>
<br>Maybe you requested the calculation of the stress tensor, e.g. you specified the keyword STRESS_TENSOR.
<br>The presence of the SCCS method section is not sufficient. Did you select the right method using the METHOD keyword?
<br>
<br>Matthias</blockquote></div>
<p></p>
-- <br>
You received this message because you are subscribed to the Google Groups "cp2k" group.<br>
To unsubscribe from this group and stop receiving emails from it, send an email to <a rel="nofollow">cp2k+...@googlegroups.com</a>.<br>
To post to this group, send email to <a rel="nofollow">cp...@googlegroups.com</a>.<br>
Visit this group at <a href="http://groups.google.com/group/cp2k" rel="nofollow" target="_blank" onmousedown="this.href='http://groups.google.com/group/cp2k';return true;" onclick="this.href='http://groups.google.com/group/cp2k';return true;">http://groups.google.com/<wbr>group/cp2k</a>.<br>
For more options, visit <a href="https://groups.google.com/d/optout" rel="nofollow" target="_blank" onmousedown="this.href='https://groups.google.com/d/optout';return true;" onclick="this.href='https://groups.google.com/d/optout';return true;">https://groups.google.com/d/<wbr>optout</a>.<br>
</div></blockquote></div></blockquote></div></div></blockquote></div>