CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > CONVERGE

recovering .... because dt_chem too large

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   September 15, 2023, 11:27
Exclamation recovering .... because dt_chem too large
  #1
Member
 
Usama
Join Date: Aug 2019
Posts: 32
Rep Power: 7
usamaubk is on a distinguished road
I am constantly getting this printed in Converge status. I am running CI engine simulation with blended fuel. This is being repeated for almost all the main injection events and after.

[ 57.81%] ncyc 40962; Crank 364.341 (deg); dt= 1.250000000e-08 (dt_grow)
recovering .... because dt_chem too large: t_chem= 5.135885e-01 dt= 1.250000e-08 mult_dt_chem= 5.000000e-01 dt_chem= 2.846167e-09
[ 57.81%] ncyc 40963; Crank 364.341 (deg); dt= 1.250000000e-08 (dt_grow)
recovering .... because dt_chem too large: t_chem= 5.125824e-01 dt= 1.250000e-08 mult_dt_chem= 5.000000e-01 dt_chem= 2.890790e-09
[ 57.81%] ncyc 40964; Crank 364.341 (deg); dt= 1.250000000e-08 (dt_grow)
recovering .... because dt_chem too large: t_chem= 5.115406e-01 dt= 1.250000e-08 mult_dt_chem= 5.000000e-01 dt_chem= 2.938424e-09
[ 57.81%] ncyc 40965; Crank 364.341 (deg); dt= 1.250000000e-08 (dt_grow)
recovering .... because dt_chem too large: t_chem= 5.049359e-01 dt= 1.250000e-08 mult_dt_chem= 5.000000e-01 dt_chem= 2.946903e-09
[ 57.81%] ncyc 40966; Crank 364.341 (deg); dt= 1.250000000e-08 (dt_grow)
recovering .... because dt_chem too large: t_chem= 5.039253e-01 dt= 1.250000e-08 mult_dt_chem= 5.000000e-01 dt_chem= 2.999713e-09

This has caused the simulation too slow down a lot. By slowing down i mean in order of weeks. The simulation is stuck (progressing very slow) when the injection is happening. I decreased the mult_dt_chem from 0.5(default) to 0.1. but still no effect on speed (it is still being printed). The time step settings are default with min. time step of 1e-8 ( at some CAD this minimum time step is also forced as dt=dtmin forced by recover_chem). Any idea how to speed up the simulation? and get rid of this?My case setup is fine and I used the same for pure diesel. I am injecting 70000 parcels/nozzle with min cell size of 0.25 at nozzle having total mass of around 40mg. It would be appreciated. Thanks
usamaubk is offline   Reply With Quote

Old   September 21, 2023, 14:02
Default
  #2
New Member
 
Chinmaya Rajath
Join Date: Sep 2020
Posts: 17
Rep Power: 6
crajath is on a distinguished road
Hello

The time step being limited by dt_chem is usually an indication that the temperature in a cell is increasing by a lot within a single time step and CONVERGE is recovering and limit the time step to prevent this from happening. Can you check your heat release rate profile and see if there is a rapid rise there? I would need to evaluate the case setup to help you further.

Please reach out to us at support@convergecfd.com so that we can look at the case setup and help you
__________________
Chinmaya Rajath

Research Engineer, Applications
Convergent Science
crajath is offline   Reply With Quote

Old   September 21, 2023, 16:01
Default
  #3
Member
 
Usama
Join Date: Aug 2019
Posts: 32
Rep Power: 7
usamaubk is on a distinguished road
Quote:
Originally Posted by crajath View Post
Hello

The time step being limited by dt_chem is usually an indication that the temperature in a cell is increasing by a lot within a single time step and CONVERGE is recovering and limit the time step to prevent this from happening. Can you check your heat release rate profile and see if there is a rapid rise there? I would need to evaluate the case setup to help you further.

Please reach out to us at support@convergecfd.com so that we can look at the case setup and help you
Hello,
yes the HRR is rising in the main injection rapidly as seen in the attachment, though the total energy released is similar. can you give suggestions on how this problem can be tackled? I am using a validated spray model with KH-RT. Thanks
Attached Images
File Type: png HRR.png (30.8 KB, 25 views)
File Type: png IHRR.png (50.9 KB, 18 views)
File Type: png Pressure.png (36.7 KB, 17 views)
usamaubk is offline   Reply With Quote

Old   September 21, 2023, 16:10
Default
  #4
New Member
 
Chinmaya Rajath
Join Date: Sep 2020
Posts: 17
Rep Power: 6
crajath is on a distinguished road
Hello,

Can you give me some more details about your spray setup? How many nozzles does you injector have? Can you confirm if you are using 70000 parcels per nozzle for a total injected spray mass of 40 mg for the injector?
__________________
Chinmaya Rajath

Research Engineer, Applications
Convergent Science
crajath is offline   Reply With Quote

Old   September 21, 2023, 16:17
Default
  #5
Member
 
Usama
Join Date: Aug 2019
Posts: 32
Rep Power: 7
usamaubk is on a distinguished road
Hello yes surely.
My setup has an injector with 8 nozzles. The total mass I am injecting is around 42mg and the parcels per nozzle are 70000. injection duration is around 16cads starting from -11aTDC. the injection pressure is adjusted at 1200 bars with a cd of 0.93. Other parameters like nozzle dia etc. are from experiments.
Thanks
usamaubk is offline   Reply With Quote

Old   September 21, 2023, 18:15
Default
  #6
New Member
 
Chinmaya Rajath
Join Date: Sep 2020
Posts: 17
Rep Power: 6
crajath is on a distinguished road
Can you let me know if you have the new_parcel_flag = 1 for the KH breakup model?
__________________
Chinmaya Rajath

Research Engineer, Applications
Convergent Science
crajath is offline   Reply With Quote

Old   September 21, 2023, 18:36
Default
  #7
Member
 
Usama
Join Date: Aug 2019
Posts: 32
Rep Power: 7
usamaubk is on a distinguished road
yes i am creating child parcels. Also i have adjusted the constants from my previous experience with adjusting the same CFD model at different operating conditions. The spray modelling snap is attached here
Attached Images
File Type: png Screenshot 2023-09-21 233148.png (182.0 KB, 19 views)
usamaubk is offline   Reply With Quote

Reply

Tags
converge, dt_chem, error, speed up

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Problems with very large IO dictionaries in parallel kingdomkarl OpenFOAM Programming & Development 4 September 28, 2021 11:28
large domain with large aspect ratio ashkan0065 FLUENT 1 October 24, 2020 18:16
Large problem partitioner tzy CFX 3 February 7, 2018 18:11
Large Epsilon 3D lepsilon Main CFD Forum 0 July 10, 2012 11:43
what happens if y+ is too large or too small Bo Jensen Siemens 12 February 20, 2003 00:31


All times are GMT -4. The time now is 09:53.