CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM Running, Solving & CFD (
-   -   DieselFoam Spray Evaporation Continuity Error (

spv24 October 31, 2008 06:50

Hello, I try to simulate th

I try to simulate the injection of water into a pipe flow. Therefor I took the dieselFoam tutorials case. I updated the Mesh and boundary files to my geometry with inflow and outflow and I replaced the injector position and the injection direction. This where the only changes I made.
But I get a growing Evaporation Continuity Error from the beginning of the simulation. Is this related to the mesh resolution? What would be a good mesh resolution?

Thanks Sammy

niklas October 31, 2008 07:18

The Evaporation Continuity err
The Evaporation Continuity error assumes closed boundaries and does not account for the inflow/outflow. Neither does it account for liquid mass going out through the open boundaries.

It has survived since the implementation days where I had it to make sure the evaporated mass was accounted for in the PISO.
In fact it could be removed now, since Im sure it works as it should.

spv24 October 31, 2008 07:33

Thanks, but even when I close
Thanks, but even when I close my pipe by putting zeroGradient BC on the outlet and inlet, the Evaporation Continuity Error grows. What could be the reason for this?

Thanks, Sammy

niklas October 31, 2008 07:36

zeroGradient is not closing th
zeroGradient is not closing the pipe.

fixedValue uniform (0 0 0) on velocity is closing the pipe.

spv24 December 1, 2008 08:02

Hello, I still have problem

I still have problems with the mass balance. I simulate a simple pipe flow with inlet and outlet and I inject a constant amount of mass per time unit. when I calculate the mass of the evaporated outflow I don't get the injected mass per time. The pipe is long enough for complete evaporation of the liquid, there are no parcels leaving the geometry. The simulation is runnig long enough for a quasi-steady-state situation. The outflow amount of the evaporated, injected C7H16 doesn't change with time. I calculate the whole mass flow at the outlet. Then I take this number for calculating the mass flow at the outlet of C7H16 by using the concentration in the C7H16 volScalarField at the outlet. I use the filter 'integrated variables' at the outlet. My questions:

Is the dieselFoam solver suitable to solve such problems with injection in a flow (open system)?

What could be my mistake?
What is the unit of the species in the O2 / N2 / C7H16 volScalarField? Is it kg / kg or mol / mol. Both are leading to wrong numbers...

Thanks Sammy

niklas December 1, 2008 08:45

Turn off injection. Do you st
Turn off injection.
Do you still have problem with mass balance?

Try to increase the accuracy of the pressure equation.


spv24 December 4, 2008 09:24

Hi, When I sompute the mass

When I sompute the massflow at the outlet I get the exact result of the added injected massflow plus the mass flow in the pipe. But the integrated massfraction of the injected speciest at the outlet is different. Where can I increase the accuracy of the pressure equation? Shall I use underrelaxation factors?

spv24 December 18, 2008 09:32

Hi Niklas, I have still the
Hi Niklas,

I have still the problem with wrong mass fraction at the outlet. I changed the geomtrys to a simple channel with a cube mesh. I inject C7H16 in a slow flow (1 m/s). When I look at the phi Field at the outlet, there is the added massflow of the injected mass and the mass in the channel flow. But when I look integrate over the C7H16 Field at the outlet, I get a massfraction of about 0.5. The massflow through the channel is about 1.39 e-4 kg /s and the injected mass is constant at 1.0 e-5 kg/s. So the correct concentration should be around 0.067 . What could be the reason for this behaviour of the C7H16 mass-flow-peofile? As air flow I use a pure N2 flow.

My BCs are:

k, epsilon, T, N2, YDefault:

at the inlet: fixedValue
at the wall and the outlet: zeroGradient


at the inlet: fixedValue
at the wall (0 0 0)
at the outlet: zeroGradient




st the inlet and wall: zerogradient
at the outlet: fixedValue

I am wondering about the correct phi-Field at the outlet and the wrong mass fraction in the C7H16 File!

Could you help me?



PS: I put nCorrectors to 5, but the Problem is still there!

niklas December 18, 2008 11:09

Can u mail me the case so I ca
Can u mail me the case so I can look at it myself.

pls dont mail it to my scania mail since I am on vacation.
use my private address niklas dot nordin at nequam dot se

Graham81 January 26, 2010 10:47

Dear sir,

After running a case using th dieselFoam solver, the evaporation continuity error has been steadily rising. From reading the above, can it be that I have misdefined my BC's. I assumed to spray a liquid into a highpressure closed cell. My BC's are defined as zerogradients. Thanks in advance for your time and expertise,


N. A. May 7, 2010 17:40

From where can you get the tutorials for dieselFoam? Is there a weblink where you can download the rest of the tutorials from?

az1362f December 26, 2010 20:21

vapor penetration
hello dears

I have a question about ploting vapor penetration versus time in openfoam, how can I do it?

how did you calculate vapor penetration in your numerical results? how much vapor mass fraction did you select if you used this method to plot vapor penetration? is it possible to explain more about how to plot vapor penetration by means of a numerical code like openFoam?


Graham81 December 27, 2010 08:42

The vapor penetration can be determined through the gaseous mass fraction of the fuel (heptane?) you are injecting. Using paraview these fields can be visualized and exported for dataprocessing per time instance. You will need some kind of treshold criterion though. I used a Virtual Schlieren (matlab routine I could send you) technique to compare simulated vapor penetration results with experiments.


az1362f December 27, 2010 18:55

Hello Dear Pieter;

thank you for your reply, I sent a private massage to you including my email adress and my questions.


az1362f December 30, 2010 11:50

Hello Dear Pieter;

thank you for your reply, I sent a private massage to you including my email adress and my questions.
please send it for me.


All times are GMT -4. The time now is 13:13.