999999: mpt_accept: error: accept failed: No error

- February 20, 2021 at 7:26 am
  
  Anadi_Mondal
  Subscriber
  
  Hello,nGreetings! I am experiencing same error since last 3 weeks. I also posted earlier about my problem. As I am getting error during initialization, one of you told me I have problem with DEFINE_DPM_INJECTION_INIT . Other recommended me to allocate UDMI before compiling UDF and to use double precision . I followed last two recommendations but still same error. However, someone told me to set counter so that I can isolate the error. Here is my DEFINE_DPM_INJECTION_INIT ,would you tell me how can I set counter here or inside any other udf?.DEFINE_DPM_INJECTION_INIT(entrainment_inj,I)n{nParticle *p;ncell_t c;nThread *tc;nnloop(p,I->p_init)n{nc = P_CELL(p);ntc = P_CELL_THREAD(p);nP_POS(p)[0]=C_UDMI(c,tc,6);nP_POS(p)[1]=C_UDMI(c,tc,7);nP_POS(p)[2]=C_UDMI(c,tc,;nP_FLOW_RATE(p)= C_UDMI(c,tc,5);nP_DIAM(p)=p_diam;nP_VEL(p)[0]=C_U(c,tc);nP_VEL(p)[1]=C_V(c,tc);nP_VEL(p)[2]=C_W(c,tc);nC_UDMI(c,tc,9)=C_U(c,tc);nC_UDMI(c,tc,10)=C_V(c,tc);nC_UDMI(c,tc,11)=C_W(c,tc);n}n}nAnd my error is same at the time of initialization,n999999: mpt_accept: error: accept failed: No errorn999999: mpt_accept: error: accept failed: No errorn==============================================================================Node 0: Process 18384: Received signal SIGSEGV.nn==============================================================================nn999999: mpt_accept: error: accept failed: No errornn999999: mpt_accept: error: accept failed: No errorn999999: mpt_accept: error: accept failed: MPI Application rank 0 exited before MPI_Finalize() with status 2nThe fl process could not be started.nnThank younAnadin
- February 21, 2021 at 12:10 am
  
  YasserSelima
  Subscriber
  
  I think the problem is in the line c = P_CELL(p); But let's make sure first.n let the code print message after each line and see where will it stop ... use the function like thisnnDEFINE_DPM_INJECTION_INIT(entrainment_inj,I)n{nParticle *p;ncell_t c;nThread *tc;nint i=0;n#if RP_HOSTnMessage(
  i = %d, i);n#endifnloop(p,I->p_init)n{ni++n#if RP_HOSTn Message(
  i = %d, i);n#endifnc = P_CELL(p);n#if RP_HOSTn Message(
  passed line 1, i = %d, i);n#endifntc = P_CELL_THREAD(p);n#if RP_HOSTn Message(
  passed line 2, i = %d, i);n#endifnP_POS(p)[0]=C_UDMI(c,tc,6);nP_POS(p)[1]=C_UDMI(c,tc,7);nP_POS(p)[2]=C_UDMI(c,tc,;n#if RP_HOSTn Message(
  passed line 3, i = %d, i);n#endifnP_FLOW_RATE(p)= C_UDMI(c,tc,5);nP_DIAM(p)=p_diam;nP_VEL(p)[0]=C_U(c,tc);nP_VEL(p)[1]=C_V(c,tc);nP_VEL(p)[2]=C_W(c,tc);n#if RP_HOSTn Message(
  passed line 4, i = %d, i);n#endifnC_UDMI(c,tc,9)=C_U(c,tc);nC_UDMI(c,tc,10)=C_V(c,tc);nC_UDMI(c,tc,11)=C_W(c,tc);n}n}n
- February 21, 2021 at 7:07 am
  
  Anadi_Mondal
  Subscriber
  
  Thank you YasserSelima ! I will use this code tomorrow. But will you check the attached image. In the ansys udf manual I see that DEFINE_DPM_INJECTION_INIT can be only used with Linux system. But I am working on Windows machine. Can I still use DEFINE_DPM_INJECTION_INIT on windows system?nn
- February 21, 2021 at 10:43 am
  
  YasserSelima
  Subscriber
  
  I am not sure. We have nothing to loose. Let's debug and see where it stops.nIf it gives you an error in the lineni++; nreplace it withni=i+=1;nn
- February 22, 2021 at 1:21 am
  
  Anadi_Mondal
  Subscriber
  
  Hi YasserSelima!nI have tried using both i++; and i=i+=1; with both serial and parallel processing options. But, in all cases I got same error message as follows:n999999: mpt_accept: error: accept failed: No errornn999999: mpt_accept: error: accept failed: No errornn999999: mpt_accept: error: accept failed: No errorn==============================================================================Node 0: Process 22844: Received signal SIGSEGV.nn==============================================================================nn999999: mpt_accept: error: accept failed: No errornn999999: mpt_accept: error: accept failed: No errorn999999: mpt_accept: error: accept failed: No errornMPI Application rank 0 exited before MPI_Finalize() with status 2n The fl process could not be started.nnJust for more information ,DEFINE_DPM_INJECTION_INIT is one of the udf out of 7 in a file. I am also using liquid and vapor mass, momentum sources( 2 mass sources, 4 momentum sources and one DEFINE_DPM_INJECTION_INIT) in the same udf.nThis all errors are happening at the time of solution initialization.nnThank You,nAnadin
- February 22, 2021 at 1:55 am
  
  YasserSelima
  Subscriber
  
  I know, but the error happens when you initialize ... This means the problem occurs in the initialization function. Other functions might have problems but they are not recalled at initialization.nBefore you got this error, did you get a message ni=0ni=1 n... and so onnif yes, please copy and paste the printed line. We can know from these lines where did the code stopn
- February 22, 2021 at 2:09 am
  
  YasserSelima
  Subscriber
  
  A questionnDo you compile, or interpret the UDF?n
- February 22, 2021 at 3:09 am
  
  Anadi_Mondal
  Subscriber
  
  Hi YasserSelima!nHere I have attached the images of console output from uploading case file to starting of error.nI am compiling the UDF.nActually for liquid it's a boundary condition . This is why define_profile is being used.n n
- February 22, 2021 at 3:51 am
  
  YasserSelima
  Subscriber
  
  It didn't print any message!!nLet's try something else ... nAdd this line at the top of all functions.Message(
  
  This is function :function_name);nIn this line, replace the function_name with the actual function name.nnThen Delete the libudf folder and recompile ... After initialisation you will receive the same error ... but every function will print a message when recalled. So, we know the last recalled function, n
- February 22, 2021 at 4:33 am
  
  Anadi_Mondal
  Subscriber
  
  I am using above format at the top of all function but getting the following error for all cases:n..\..\src\liquid_film.c(17): error C2143: syntax error: missing ')' before 'string'n..\..\src\liquid_film.c(17): error C2143: syntax error: missing '{' before 'string'n..\..\src\liquid_film.c(17): error C2059: syntax error: 'string'n..\..\src\liquid_film.c(17): error C2059: syntax error: ')'n
- February 22, 2021 at 11:47 am
  
  YasserSelima
  Subscriber
  
  This is the top of of the function ... inside the function, before any calculations.nnn
- February 23, 2021 at 2:25 am
  
  Anadi_Mondal
  Subscriber
  
  HelloYasserSelima!nI am using '' message'' as fellows:nDEFINE_PROFILE(film_mass_source,tf,i)n{nreal film_thickness, film_vel, vap_vel, Jf, Jg, fi, Ref, area_face, area[ND_ND], pos[ND_ND], r1, angle, x_inj, y_inj, z_inj, evap_mass_source, evap_mom_source, ent_mass_source;nreal flow_time = CURRENT_TIME;nface_t f;ncell_t c;nThread *tc;nMessage(
  
  This is function :film_mass_source);nbegin_f_loop(f,tf)n{nfilm_thickness = F_EFILM_HEIGHT(f,tf);n...............................................................nn And starting of error is like these-nnAvailable material properties or methods have changed.nPlease confirm the property values before continuing.nnn==============================================================================nnNode 1:Process 1804:Received signal SIGSEGV.nn==============================================================================nn999999:mpt_accept:error:accept failed:No errornn999999:mpt_accept:error:accept failed:No errorn.......................................................................................n999999:mpt_accept:error:accept failed:No errornn999999:mpt_accept:error:accept failed:No errorMPI Application rank 0 exited before MPI_Finalize() with status 2n The fl process could not be startednnStill it's not printing any output in console windows
- February 23, 2021 at 3:18 am
  
  YasserSelima
  Subscriber
  
  Do you get this error when you run the simulation without UDF?nAvailable material properties or methods have changed.nPlease confirm the property values before continuing.n
- February 23, 2021 at 3:36 am
  
  YasserSelima
  Subscriber
  
  nn?if(!Data_Valid_P()) n return;nnn
- February 23, 2021 at 4:05 am
  
  Anadi_Mondal
  Subscriber
  
  Hi YasserSelima !nI notice that when I am trying to turn on the 'Eulerian Wall Film ' and selecting 'Film Material ' water-liquid ' and Surface Tension 0.017633, and clicking 'ok' , I am getting bottom message in the console window even I am not using UDF. Film material was air and surface tension was 0.76.. as default.n
- February 23, 2021 at 7:24 am
  
  Anadi_Mondal
  Subscriber
  
  I can run the simulation if I don't use UDF. So, obviously the problem is in UDF. Do you need any more info ?n
- February 23, 2021 at 12:01 pm
  
  YasserSelima
  Subscriber
  
  yes, lots of questionsnFirst, Are you able to run the case with air film material using the UDF?nSecond, Did you press the initialize at the bottom of wall film material panel?.The error you get SIGSEGV means that Node 0 requested some piece of information from memory and it did not find it. This piece of information could be material property, location in UDM, cell does not exist ... or something else. n
- February 23, 2021 at 6:57 pm
  
  Anadi_Mondal
  Subscriber
  
  Hi YasserSelima,nThank you for constantly supporting me!nnI have tried to run the case with air film but same type of error.nAt first I was not pressing the initialize at the bottom of wall film material panel. Later, I repeat the process by pressing 'initialize' but both cases I got same output in console panel.nBut, what is the correct way, should I press initialize or not at the bottom of wall film material panel?nFor the case of Memory/UDM, I am allocating 12 locations (Number of User Defined Memory Location ) before compiling UDF. I have attached an image of it. My C_UDMI is from C_UDMI (c,t,0) to C_UDMI (c,t,11) in UDF. So I am using 12 locations.nIn all cases errors are as follows:(sometimes I am getting this error with numbers of '999999: mpt_accept: error: accept failed: No error' lines.)nn==============================================================================nnNode 1: Process 25200: Received signal SIGSEGV.nn==============================================================================nMPI Application rank 0 exited before MPI_Finalize() with status 2n The fl process could not be started.nnWhere should I use the command ? if(!Data_Valid_P()) n return;nnShould I use at the bottom of these lines?( This is the inside of DEFINE_DPM_INJECTION_INIT function)nloop(p,I->p_init)n{ni++;n#if RP_HOSTnMessage(
  , i = %d, i);n#endifnc = P_CELL(p);n#if RP_HOSTnMessage(
  passed line 1, i = %d, i);n#endifntc = P_CELL_THREAD(p);n#if RP_HOSTnMessage(
  passed line 2, i = %d, i);n#endifnnn
- February 23, 2021 at 7:19 pm
  
  YasserSelima
  Subscriber
  
  For the film, I am not sure about the answer but we can ask one of the experts here.n Array , Can you please have a look at the Eulerian wall film settings and let Array know the right procedure to initialise the film. There is an error message appeared after changing the film material. The message appears in the photos above. Thank you! nnnFor the UDM, I believe you are selecting the right number. Let's add a value to the UDM before starting. This function should put a value of 1 in all C_UDMI . After compile you should find this in the execute on demand menu. Run this before initialisation.nnDEFINE_ON_DEMAND(init_udm)n{n Domain *domain;n cell_t c;n Thread *t;n n domain=Get_Domain(1);n n /* Fill the UDM with magnitude of gradient. */n thread_loop_c (t,domain)n {n begin_c_loop (c,t)n {n C_UDMI(c,t,0) = 1;n C_UDMI(c,t,1) = 1;n C_UDMI(c,t,2) = 1;n C_UDMI(c,t,3) = 1;n C_UDMI(c,t,4) = 1;n C_UDMI(c,t,5) = 1;n C_UDMI(c,t,6) = 1;n C_UDMI(c,t,7) = 1;n C_UDMI(c,t, = 1;n C_UDMI(c,t,9) = 1;n C_UDMI(c,t,10) = 1;n C_UDMI(c,t,11) = 1;n }n end_c_loop (c,t)n }nnnnThese lines should check that the data is valid before proceeding in calculation. It should be used below the message Message( This is function : film_mass_source);nif(!Data_Valid_P()) n return;nnn
- February 23, 2021 at 7:26 pm
  
  YasserSelima
  Subscriber
  
  99999 is the Host . there is no error happening in the hostnthe actual error happens at Node 0: SIGSEGVnnEven when you select 1 processor, Fluent run in parallel. So, the same processor acts as a host and Compute Node 0 at the same time. No serial solver available now in Fluent
- February 24, 2021 at 6:27 am
  
  Anadi_Mondal
  Subscriber
  
  Finally it works. It runs without any error. I am trying to run on my laptop to check it works or not. It has 260k elements, so it will not run on my laptop. I will try this simulation on university's vLabs . Let's see which type of data it gives.nThank you so much for helping constantly. See you again.n
- February 24, 2021 at 9:38 am
  
  Rob
  Forum Moderator
  
  Re the film, open the material properties panel up and check the water values. When you turn on models/interaction options the solver may require additional property data. Never assume that the default value is correct: it most likely is, but not necessarily for your operating conditions. n260k elements will typically need around 0.5-1GB RAM, the extra UDMs will push that but most modern PCs should handle it. As a note, I usually set one extra UDM than I code just to be sure: it's not needed but I took a while to get that C starts counters at zero rather than one. nIf you're not using 2021R1 I'd also suggest getting the new version. We've been doing a lot of work on the EWF and other interaction models, and are continuing to do so. Read the update and beta sections of the manual. n
- February 24, 2021 at 11:49 am
  
  YasserSelima
  Subscriber
  
  Finally it works. It runs without any error. I am trying to run on my laptop to check it works or not. It has 260k elements, so it will not run on my laptop. I will try this simulation on university's vLabs . Let's see which type of data it gives.Thank you so much for helping constantly. See you again./forum/discussion/comment/107925#Comment_107925
  Congrats! Happy for you!nnThank you @Rob, !n
- March 10, 2021 at 7:11 am
  
  Anadi_Mondal
  Subscriber
  
  and ,nHope you are doing well.nI am experiencing a new problem. I have no error to run the simulation but it's only solving for either liquid or vapor phase. Not getting residual for liquid and vapor phase together.nMy inlet has only vapor (I selected volume fraction 1 in 'multiphase' option) and I have added water mass flux at the injection wall near inlet through EWF. But the simulation is not showing any value for water.(please see the attached image)nIf I selected '0' volume fraction in the inlet , the I am getting residuals for water only. No data for vapor phase.nI selected injections from interior_surface after 1s but not getting any injected droplets value.nWhat might be the problem?nI have added my mesh and residuals for your better understanding.nnThank YounAnadin n
- March 10, 2021 at 9:14 am
  
  YasserSelima
  Subscriber
  
  Welcome back!nHopefully you do not need to debug the UDF again! Let's see what Rob says. I never used Eulerian film. If he says debug, tag me againn
- March 12, 2021 at 7:17 am
  
  Anadi_Mondal
  Subscriber
  
  nWill you kindly go over my problem?nHere is the description of the modeln1) The simulation is to visualize Annular flow in tube. I choose Eulerian -Eulerian model with water as primary phase and vapor as secondary phase. Vapor will flow inside the core of tube and liquid as film at wall. Moreover, I am using DPM to inject liquid droplet from the interior_surface(please see above mesh). Will this interior surface create blockage to flow liquid/vapor in tube?n2) Only vapor is flowing from the inlet (keep the volume fraction 0 under multiphase tab) and I am using a injection wall (EWF) near inlet for liquid film. Here I have added a mass flux(water)n3) I enable Eulerian Wall Film at injection wall, stabilization and annular wall under boundary conditions and finally initialize EWF from Model(Eulerian Wall Film)nThe simulation is running but giving residuals only for water.nWould you tell me where is the mistake?.Thank YounAnadin
- March 12, 2021 at 4:41 pm
  
  Rob
  Forum Moderator
  
  As you've opened another thread I'll lock this one. Asking new questions in new threads is recommended. n