
[Sponsors] 
October 21, 2013, 05:02 
Floating point exception: Zero divide

#1 
Member
liladhar
Join Date: May 2012
Posts: 35
Rep Power: 6 
Hi everybody
I am trying to simulate hullpropeller interaction to find the incoming flow properties at the propeller. I am using ANSYS CFX v.14 for this analysis. my system has ram of 8 GB. I did meshing in ICEM CFD of good quality. When i am trying to run simulation in CFX i am getting Floating point exception: Zero divide error. I modelled hull and propeller separately and meshed separately and combined them in CFX. I am using frozen rotor approach for this simulation. if anybody having idea of how to model hull and propeller in CFX please help me. Or can anybody tell me whats wrong in my physics settings? thanks in advance.... 

October 21, 2013, 05:14 

#2 
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 12,412
Rep Power: 97 
Floating point error is an FAQ: http://www.cfdonline.com/Wiki/Ansys...do_about_it.3F
Your approach of modelling the prop with frozen rotor sounds OK, providing the transient stuff does not matter. You will probably have to model a few different prop angles so you can get a prop torque curve against rotation angle. 

October 21, 2013, 05:23 

#3 
Member
liladhar
Join Date: May 2012
Posts: 35
Rep Power: 6 
Thanks ghorrocks..
i already got that link which u have sent me. i gone through the literature there. my mesh quality is good except in the region where hull and propeller interacts. there quality is greater than 0.1. i am running for steady analysis only. actually i am not modelling real propeller instead, i am taking a disk and giving it a rotation rate. i have two doimains 1) stationary domain containing hull and a disk 2) rotating domain having a propeller disk i created an interface between theses two disks. Did i made a mistake in this? please suggest.. thanks 

October 21, 2013, 05:31 

#4 
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 12,412
Rep Power: 97 
Your setup sounds fine  except for the bit about using a disk for a prop. If fluid cannot go through a prop then how is that in any way representative? The flow is going to have to divert around the prop and that is not realistic at all.
Also  if you have a rotating disk then this can be modelled using a tangential wall velocity on a normal wall boundary. No need for rotating frames of reference or GGI interfaces. Much simpler. But if you have no idea what your prop looks like then why not model it simply as a momentum source? Then you easily can give it the thrust and torque which is realistic, you do not need rotating frame of reference and you do not need GGIs. This will be a miuch simpler simulation. 

October 21, 2013, 05:42 
Floating point exception: Zero divide

#5 
Member
liladhar
Join Date: May 2012
Posts: 35
Rep Power: 6 
dear ghorrocks
i wanted to measure the wake field at the propeller plane. When the flow approaches the propeller, beacause of suction effect it accelerates in the vicinity of the propeller. That flow i wanted to study. The propeller meshing is difficult so i have choosen a simple rotating disk which can act as a propeller although flow will not pass through it. i really dont know how to model a propeller as a momentum source and how to give tangential wall velocity. please suggest.... 

October 21, 2013, 05:57 

#6 
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 12,412
Rep Power: 97 
As my previous post said, your disk will not suck flow into it, it will push it around it. Not realistic at all.
It sounds like the momentum source approach is what you want. Read the documentation and a CFD text book about momentum sources. 

October 24, 2013, 02:06 
Floating point exception: Zero divide

#7 
Member
liladhar
Join Date: May 2012
Posts: 35
Rep Power: 6 
hi everyone
i am thankfull to all who helped me till now. i have one more doubt. as my previous post of modeling a propeller behind the ship hull Mr. ghrocks gave me an idea to model the disc as a momentum source. i followed that. 1) i have given the force per unit volume as 1500 kg/m2s2 in x dir and rest is zero. The diameter of the disc is 1m and the thickness is 0.245m. hence applying the area and volume formula for a cylinder i got the values for are and volume as 0.7854m2 and 0.1924m3 respectively. So i got the force value as 288.63 kgm/s2. Using this force value and pressure force relation i am getting the momentum source coefficient as 367.68. Is this value correct? Did i made the mistake in finding the momentum source coefficient? 2) If i give inlet velocity as 1m/s my solution diverges badly and ending up in floating point error. If i give 3m/s is running good but u and w momentum are not converging as good as v momentum. can anybody reason why? thanks in advance 

October 24, 2013, 05:40 

#8 
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 12,412
Rep Power: 97 
I do not have time to check your maths. If you want to check it then check the flow it generates is correct, or the total momentum is correct of some other check.
Did you read the FAQ I linked to in post #2? That explains the floating point error. 

October 24, 2013, 23:23 
Floating point exemption error

#9 
Member
liladhar
Join Date: May 2012
Posts: 35
Rep Power: 6 
After the convergence reached i am getting following message
CFD Solver finished: Thu Oct 24 19:41:01 2013 CFD Solver wall clock seconds: 2.8020E+03 ================================================== ==================== Termination and Interrupt Condition Summary ================================================== ==================== CFD Solver: All target criteria reached (Equation residuals) ================================================== ==================== Boundary Flow and Total Source Term Summary ================================================== ==================== ++  UMomBulk  ++ Boundary : hull 6.0959E+12 Boundary : inlet 5.0978E+17 Boundary : outlet 4.9922E+17 Boundary : slip walls 1.3809E+15 Boundary : symmetry 1.3863E+15 Boundary : top 7.9038E+10 SubDomain : Subdomain 1 1.1523E+12 Domain Interface : Domain Interface 1 (Side 1) 2.3761E+11 Domain Interface : Domain Interface 1 (Side 2) 2.3712E+11  Domain Imbalance : 1.0561E+16 Domain Imbalance, in %: 0.0745 % ++  VMomBulk  ++ Boundary : hull 1.7420E+13 Boundary : inlet 4.8804E+11 Boundary : outlet 2.3332E+11 Boundary : slip walls 4.2680E+18 Boundary : symmetry 4.2789E+18 Boundary : top 6.1874E+08 Domain Interface : Domain Interface 1 (Side 1) 4.0946E+09 Domain Interface : Domain Interface 1 (Side 2) 4.2418E+09  Domain Imbalance : 1.0891E+16 Domain Imbalance, in %: 0.0768 % ++  WMomBulk  ++ Boundary : hull 2.3860E+15 Boundary : inlet 7.2921E+11 Boundary : outlet 3.6711E+13 Boundary : slip walls 1.4140E+19 Boundary : symmetry 4.3586E+15 Boundary : top 2.5582E+11 Domain Src (Neg) : Domain 1 1.4175E+19 Domain Interface : Domain Interface 1 (Side 1) 2.5574E+11 Domain Interface : Domain Interface 1 (Side 2) 2.5393E+11  Domain Imbalance : 2.8210E+16 Domain Imbalance, in %: 0.1990 % ++  PVol  ++ Boundary : inlet 7.7341E+12 Boundary : outlet 7.6603E+12 Boundary : top 1.9741E+10 Domain Interface : Domain Interface 1 (Side 1) 1.2149E+08 Domain Interface : Domain Interface 1 (Side 2) 1.2493E+08  Domain Imbalance : 9.3550E+10 Domain Imbalance, in %: 1.2096 % ++  Masswater  ++ Boundary : inlet 7.7317E+12 Boundary : outlet 7.6578E+12 Boundary : top 1.8184E02 Domain Interface : Domain Interface 1 (Side 1) 2.1728E+08 Domain Interface : Domain Interface 1 (Side 2) 2.1728E+08  Domain Imbalance : 7.3861E+10 Domain Imbalance, in %: 0.9553 % ================================================== ==================== Wall Force and Moment Summary ================================================== ==================== Notes: 1. Pressure integrals include the reference pressure. To exclude it, set the expert parameter 'include pref in forces = f'. ++  Pressure Force On Walls  ++ XComp. YComp. ZComp. Domain Group: Domain 1 hull 6.0956E+12 1.9076E+14 2.0211E+15 slip walls 1.3844E+15 4.2789E+18 1.4165E+19    Domain Group Totals : 1.3905E+15 4.2787E+18 1.4167E+19 ++  Viscous Force On Walls  ++ XComp. YComp. ZComp. Domain Group: Domain 1 hull 3.4003E+08 2.2151E+08 7.4606E+08 slip walls 6.1402E01 1.1486E+03 7.4831E+05    Domain Group Totals : 3.4003E+08 2.2151E+08 7.4531E+08 ++  ERROR #001100279 has occurred in subroutine ErrAction.   Message:   Floating point exception: Overflow            ++ ++  ERROR #001100279 has occurred in subroutine ErrAction.   Message:   Stopped in routine FPX: C_FPX_HANDLER            ++ ++  An error has occurred in cfx5solve:     The ANSYS CFX solver exited with return code 1. No results file   has been created.  ++ End of solution stage. ++  The following transient and backup files written by the ANSYS CFX   solver have been saved in the directory D:\Liladhar\morring   analysis\icem file\231013\cfx\try 1_023:     14_full.bak  ++ ++  The following user files have been saved in the directory   D:\Liladhar\morring analysis\icem file\231013\cfx\try 1_023:     mon  ++ ++  Warning!     After waiting for 60 seconds, 1 solver manager process(es) appear   not to have noticed that this run has ended. You may get errors   removing some files if they are still open in the solver manager.  ++ This run of the ANSYS CFX Solver has finished. Can anybody tell me why this error is encountering? I have give velocity inlet and specified the volume fraction of air and water. Outlet as a pressure outlet and top as a opening where only air is present. hull surface as a no slip wall and other walls as slip walls. Are my boundary conditions wrong? please help thanks in advance. 

October 25, 2013, 04:54 

#10 
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 12,412
Rep Power: 97 
You have a lot of E+12's happening there. The simulation has diverged. The link I posted in #2 is still what has gone wrong and what to do about it.


December 16, 2013, 03:05 
z

#11 
New Member
Mausam Shresha
Join Date: Aug 2012
Posts: 13
Rep Power: 5 
PRoblem
it is for the simulation of Cross flow turbine. the interface used is Frozen rotor. The inlet mass flow rate and Outlet atmospheric pressure is set. The default wall is set for rotor. 

December 16, 2013, 05:07 

#12 
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 12,412
Rep Power: 97 
FAQ: http://www.cfdonline.com/Wiki/Ansys...do_about_it.3F
That is the forth time I have referred to it in this thread. 

Tags 
cfx, floating point error 
Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
simpleFoam Floating point exception error help  sudhasran  OpenFOAM Running, Solving & CFD  3  March 12, 2012 17:23 
Pipe flow in settlingFoam floating point exception  jochemvandenbosch  OpenFOAM Running, Solving & CFD  4  February 16, 2012 04:24 
blockstructured mesh for tjunction  Robert@cfd  ANSYS Meshing & Geometry  20  November 11, 2011 05:59 
Finished simulation doesn't start: floating point exception [Divide by zero]  MaxCFD  STARCCM+  3  June 26, 2011 10:31 
Floating point error: divide by zero  Chris  FLUENT  2  April 15, 2005 02:33 