[pary07]

Hello everyone,

I am running a steady state simulation on single cylinder engine and I want to calculate the tumble and swirl on derived parts. Has anyone done this, as I have tried and I am getting really odd values?

Thanks in advance.

[ashokac7]

Originally Posted by pary07

Hello everyone,

I am running a steady state simulation on single cylinder engine and I want to calculate the tumble and swirl on derived parts. Has anyone done this, as I have tried and I am getting really odd values?

Thanks in advance.

Yes it is proposed under development idea in Steve Portal . But I read somewhere that you can able to get it on cell set. Try that.

[pary07]

Hi Ashok,

Thank you for replying. I did create a cell set and then also made a threshold of it, but I do not know how can I use this to plot tumble or swirl. Also, I never came across this method before, so could you please tell me more about how I can achieve this?
My initial derived part was a constrained plane section which I used as input part on my reports to evaluate expressions but here I cannot use the cell sets as input, but I can indeed use the threshold as input and on using this, it shows no data when I try to run the reports.

[ashokac7]

Originally Posted by pary07

Hi Ashok,

Thank you for replying. I did create a cell set and then also made a threshold of it, but I do not know how can I use this to plot tumble or swirl. Also, I never came across this method before, so could you please tell me more about how I can achieve this?
My initial derived part was a constrained plane section which I used as input part on my reports to evaluate expressions but here I cannot use the cell sets as input, but I can indeed use the threshold as input and on using this, it shows no data when I try to run the reports.

This article is from steve portal for tumble/swirl ratio.

The tumble value is defined as the ratio of angular velocity of the solid body vortex to the axial velocity of the piston.

The tumble flow is a superposition of a circular and an axial movement in the direction of the piston.

In general:
Tu=Vt/Va

where va| can be described by:
Va=mdot/ A*density

Under the assumption that vt and r are in the same plane, vt can be described as :
Vt=w_vector (cross product) r_vector


You need a coordinate system (CS) in the center of the vortex to which the tumble value is related to and the normal vector of the plane mentioned before:

Create a field function for vt.
vt = dot($$Velocity, unit(cross($$Centroid ("CS") , $$normal vector of the plane)))


Next create a mass flow report (MassFlow1Report) to determine the mass flow that enters the domain.

Create a field function for va:
va = $MassFlow1Report / ($cylinder cross section area * $Density)


Last step:
Tu = $vt / $va

[ashokac7]

I am also working on a problem involving swirl. I want to give swirl in the domain. But I know how to specify that on boundary not in region. Do you know anything on this?

[pary07]

Yes, I did come across this link, it is an alternative to the tumble/Swirl report. I tried using the above equation by creating scalar field function vt and va as :
vt=dot($$Velocity, unit(cross($$Centroid ("Laboratory.CartesianCSYS") , $$Normal)))
va=${MassFlow1Report}/(${cylinder_area} *${Density})

Here I am transforming my coordinates, and for the mass flow report I use my section plane as input, but when I run the report, it shows error as cannot evaluate velocity or density. The thing is I am not sure how will this work without using any sum reports.

[ashokac7]

Originally Posted by pary07

Yes, I did come across this link, it is an alternative to the tumble/Swirl report. I tried using the above equation by creating scalar field function vt and va as :
vt=dot($$Velocity, unit(cross($$Centroid ("Laboratory.CartesianCSYS") , $$Normal)))
va=${MassFlow1Report}/(${cylinder_area} *${Density})

Here I am transforming my coordinates, and for the mass flow report I use my section plane as input, but when I run the report, it shows error as cannot evaluate velocity or density. The thing is I am not sure how will this work without using any sum reports.

Did you create the co-ordinate system in the domain. If co-ordinate system lies outside then it may not give you any result. This may be the case.

And I don't understand that why do we need sum report?

[pary07]

My coordinate system is well within the cylinder and indeed at half the stroke. Well, when I create the scalar ff(field function) vt, it comprises of the vector velocity. Now, this is just a ff. So when I execute the expression vt/va, it says cannot evaluate ff Velocity and that I assume is because I cannot assign any part to this ff vt.
How exactly are you thinking about this? Maybe I am wrong somewhere.

[ashokac7]

Originally Posted by pary07

My coordinate system is well within the cylinder and indeed at half the stroke. Well, when I create the scalar ff(field function) vt, it comprises of the vector velocity. Now, this is just a ff. So when I execute the expression vt/va, it says cannot evaluate ff Velocity and that I assume is because I cannot assign any part to this ff vt.
How exactly are you thinking about this? Maybe I am wrong somewhere.

I am also working on this. Same error like you. I will let you know if I find something.

[pary07]

Originally Posted by ashokac7

I am also working on this. Same error like you. I will let you know if I find something.

Hi ashok, did you find anything on this?

[ashokac7]

Originally Posted by pary07

Hi ashok, did you find anything on this?

Sorry Paresh, I found the way around it. I had to create the swirl so what I did is create the cylindrical co-ordinate system at inlet. And apply the flow direction as components. For this only tangential and axial velocity was required. But I was unable to calculate realizable swirl values in region. Instead you can also do the same. Create a plane where you want swirl value. Then crate local cylindrical co-ordinate system at this plane. Calculate X,Y and Z components of velocity (for cylindrical co-ordinate sys) using report. X is radial, Y is Tangential and Z is axial. So swirl will be (Vtangential/Vaxial). I don't know what is accuracy of this. But this should work, may be.