Weird Cd and Cl values and convergence issue

JX- · February 15, 2017, 03:10

dear all,

i have tried to solve my model for the first 100 iterations using the first order upwind turbulence model. (k-epsilon with reliazable and non-eq wall function)

the solver used was a coupled solver.

throughout the first 50 iterations, the graph seems fine. however the Cd values were unrealistic. it started at like 1.xx till like 5.xx and then it went hairwire.

i wonder what went wrong? could it be my mesh?

any ideas or opinions are highly appreciated.

thanks,
Jx

JX- · February 15, 2017, 03:15

found out why i have weird Cd values. will update if i encounter any problems.

diamondx · February 15, 2017, 03:15

check your reference values, the area !

TimTom22 · February 15, 2017, 04:03

Hey, JX... what did you find out that was giving you those unrealistic values? I was running a NACA 0012 a while back trying to validate the simulation with its data from a book and my Cd's were also unrealistic.

JX- · February 15, 2017, 04:12

Quote:

Originally Posted by TimTom22

Hey, JX... what did you find out that was giving you those unrealistic values? I was running a NACA 0012 a while back trying to validate the simulation with its data from a book and my Cd's were also unrealistic.

hey TimTom22,

i have not tested it out yet, but im pretty sure it was because when i was setting the areas to be monitored for drag, lift and moment values, i accidentally highlighted the ground/road face as well. i guess thats why i got the values to be so high and then halfway thru i experienced a warning message saying reverse flow is happening at the pressure outlet etc etc.

i will try it out again and will post a full thread on what im doing.

diamondx · February 15, 2017, 04:14

Also make sure the scale of your mesh is correct. It might be look stupid, but I've seen so many people forgetting to set the scale of their mesh.

Sent from my Nexus 6P using CFD Online Forum mobile app

JX- · February 15, 2017, 04:16

Quote:

Originally Posted by diamondx

Also make sure the scale of your mesh is correct. It might be look stupid, but I've seen so many people forgetting to set the scale of their mesh.

Sent from my Nexus 6P using CFD Online Forum mobile app

i have double checked on my mesh's scale and im pretty sure its correct. my model is about 300mm in length and my entire enclosure as opened in fluent shows its about 7m in length.

JX- · February 15, 2017, 04:29

https://www.dropbox.com/s/cy9rqksky6...72529.jpg?dl=0

As expected TimTom22. Can see from the picture i uploaded.

JX- · February 15, 2017, 04:37

Can somebody explain to me whats going on here? I have gone clueless... Been facing this issue for the past 48 hours!!

https://www.dropbox.com/s/j8ru3drq0b...73312.jpg?dl=0

diamondx · February 15, 2017, 04:49

Have you checked the area in the reference values, that what fluent uses for calculating cl. What about the lift force, are they correct ??

Sent from my Nexus 6P using CFD Online Forum mobile app

JX- · February 15, 2017, 04:55

Quote:

Originally Posted by diamondx

Have you checked the area in the reference values, that what fluent uses for calculating cl. What about the lift force, are they correct ??

Sent from my Nexus 6P using CFD Online Forum mobile app

yea the area i have computed from the reports under projected areas in the axis of air flow.

didnt bother checking the values as the Cd and Cl values went all wrong after 20 plus iterations. i wonder what went wrong... you seen the image i posted in the dropbox link yet?

JX- · February 15, 2017, 05:13

a couple of pictures about my mesh around the 3d model

JX- · February 15, 2017, 06:14

this is what i got after i have excluded the 'wheels' in my solution monitors. just the first 100 iterations to initialize the problem (k-epsilon, realizable, non-equilibrium wall function 2 eqn turbulence model, coupled solver, first order everything, flow courant number of 25, ERF of 0.25 for both momentum and pressure)

turned out that was causing the sudden drop and increase in crazy Cd and Cl values. however, is what im doing the right way to analyse the drag and lift around my 3d model?

some pictures for references...

kuzma169 · February 16, 2017, 02:16

Quote:

Originally Posted by JX-

dear all,

i have tried to solve my model for the first 100 iterations using the first order upwind turbulence model. (k-epsilon with reliazable and non-eq wall function)

the solver used was a coupled solver.

throughout the first 50 iterations, the graph seems fine. however the Cd values were unrealistic. it started at like 1.xx till like 5.xx and then it went hairwire.

i wonder what went wrong? could it be my mesh?

any ideas or opinions are highly appreciated.

thanks,
Jx

Sent from my iPad using CFD Online Forum mobile app

February 15, 2017, 03:10	Weird Cd and Cl values and convergence issue	#1
JX- New Member Jun Join Date: Oct 2016 Posts: 19 Rep Power: 9	dear all, i have tried to solve my model for the first 100 iterations using the first order upwind turbulence model. (k-epsilon with reliazable and non-eq wall function) the solver used was a coupled solver. throughout the first 50 iterations, the graph seems fine. however the Cd values were unrealistic. it started at like 1.xx till like 5.xx and then it went hairwire. i wonder what went wrong? could it be my mesh? any ideas or opinions are highly appreciated. thanks, Jx

February 15, 2017, 03:15		#3
diamondx Super Moderator Ghazlani M. Ali Join Date: May 2011 Location: Tokyo, Japan Posts: 1,385 Blog Entries: 23 Rep Power: 28	check your reference values, the area ! __________________ Regards, New to ICEM CFD, try this document --> https://goo.gl/KAOIwm Ali

February 15, 2017, 04:14		#6
diamondx Super Moderator Ghazlani M. Ali Join Date: May 2011 Location: Tokyo, Japan Posts: 1,385 Blog Entries: 23 Rep Power: 28	Also make sure the scale of your mesh is correct. It might be look stupid, but I've seen so many people forgetting to set the scale of their mesh. Sent from my Nexus 6P using CFD Online Forum mobile app __________________ Regards, New to ICEM CFD, try this document --> https://goo.gl/KAOIwm Ali

February 15, 2017, 04:49		#10
diamondx Super Moderator Ghazlani M. Ali Join Date: May 2011 Location: Tokyo, Japan Posts: 1,385 Blog Entries: 23 Rep Power: 28	Have you checked the area in the reference values, that what fluent uses for calculating cl. What about the lift force, are they correct ?? Sent from my Nexus 6P using CFD Online Forum mobile app __________________ Regards, New to ICEM CFD, try this document --> https://goo.gl/KAOIwm Ali

February 15, 2017, 04:03		#4
TimTom22 New Member timtom Join Date: Feb 2017 Posts: 1 Rep Power: 0	Hey, JX... what did you find out that was giving you those unrealistic values? I was running a NACA 0012 a while back trying to validate the simulation with its data from a book and my Cd's were also unrealistic.

February 15, 2017, 04:29		#8
JX- New Member Jun Join Date: Oct 2016 Posts: 19 Rep Power: 9	https://www.dropbox.com/s/cy9rqksky6...72529.jpg?dl=0 As expected TimTom22. Can see from the picture i uploaded.

February 15, 2017, 04:37		#9
JX- New Member Jun Join Date: Oct 2016 Posts: 19 Rep Power: 9	Can somebody explain to me whats going on here? I have gone clueless... Been facing this issue for the past 48 hours!! https://www.dropbox.com/s/j8ru3drq0b...73312.jpg?dl=0

February 15, 2017, 05:13		#12
JX- New Member Jun Join Date: Oct 2016 Posts: 19 Rep Power: 9	a couple of pictures about my mesh around the 3d model

February 15, 2017, 06:14		#13
JX- New Member Jun Join Date: Oct 2016 Posts: 19 Rep Power: 9	this is what i got after i have excluded the 'wheels' in my solution monitors. just the first 100 iterations to initialize the problem (k-epsilon, realizable, non-equilibrium wall function 2 eqn turbulence model, coupled solver, first order everything, flow courant number of 25, ERF of 0.25 for both momentum and pressure) turned out that was causing the sudden drop and increase in crazy Cd and Cl values. however, is what im doing the right way to analyse the drag and lift around my 3d model? some pictures for references...