[sivakumar]

Hi Foamers,
I want to calculate torque from the simulation, can you guys help me?
I have used MRFSimpleFoam for my simulation. The fan RPM is 750.

please give me some idea.

Thanks and regards,
Sivakumar

[laurentb]

Hi
Have a look to the thread :
Monitoring Torque on rotating body in MRFSimpleFoam
Laurent

[sivakumar]

Hi Guys,
After a long struggle, I came to the conclusion regarding torque.
To calculate torque we need to use this function in the controlDict file.

functions ( forces { type forces; functionObjectLibs ("libforces.so"); outputControl timeStep; outputInterval 1; patches (wallBlade); // pname p; // Uname U; rhoName rhoInf; log true; rhoInf 1.205; CofR (0 0 0); } );
results will be as follows:
forces(pressure, viscous)((2162.2 -313.192 35.8823) (12.5865 3.1166 3.97532)) moment(pressure, viscous)((1162.4 6862.37 100.944) (-12.5067 35.4299 -50.9462))
now the the output pressure force is not exact pressure (unit m2/s2)
so to get exact torque = obtained Torque (moment) / density

Thanks,
Sivakumar

[laurentb]

Hi,

Does somebody can explain in details how to calculate torque with "forces" library data ?

I try to compare experimental and simulation results but there is a big difference.

I calculate the torque as follow : Torque = |moment(pressure) + moment(viscous)| and i obtain twice the experimental value...

Thanks
Laurent

[sivakumar]

hi,
Which solver you are using?
for any simpleFoam you need to divide it with the density what you are using.

Thanks,
Sivakumar

[laurentb]

I'm using MRFSimpleFoam. The density is 1300 kg/m3. If i divide by 1300 the difference is bigger...

[sivakumar]

hi,
What kind of simulation you are doing, I am also using MRFSimpleFoam for axial flow fan.

In my case MRFSimpleFoam torque results keep on changing for each case
Actually, I tried two cases one is with passage (suction side and pressure side of the fan)
secondly two blade and quarter of the cylinder see the fig.

for passage case I am getting 45 N-m (it should be 70-80 N-m)
for 2 blade case I am getting 80 N-m

I am totally confused, is there any problem with MRFSimpleFoam or in my settings.
I am totally gone.

Thanks,
Sivakumar

[sivakumar]

Hi Guys and Laurentb,
I got the right way to calculate torque.

I was doing some wrong calculation, that is why my results were wrong.
now I got the right results, with 2% error.

To Laurentb, I am sorry man, I think I have miss guided you.

Actually we need to multiply the pressure with density to get the exact pressure.

to calculate torque: take the z component of the pressure force and viscous.

torque = F*r

F = ((P *A) *density+ tau *A)

A --> area of the patch which you want to calculate torque.

I hope it is clear now

Thanks,
Sivakumar

[laurentb]

Hi,

Finally the torque computed by the forces library fits my experimental data with 10% error.

My error comes from that i turn off the turbulence in RASProperties file for laminar flow conditions.

If i turn on the turbulence option the force computation works well.

Laurent

[wes1204]

Hi all

I have something to ask.

My case is 2D on x-y plane.

Object look like airfoil and it is in wind tunnel.

I obtain moment using force function added controldic file.

Center of moment is z-axis.

From result, moment has minus value.

In this case, Is object rotated to clockwise or countclockwise??

[sharonyue]

Originally Posted by sivakumar

Hi Guys and Laurentb,
I got the right way to calculate torque.

I was doing some wrong calculation, that is why my results were wrong.
now I got the right results, with 2% error.

To Laurentb, I am sorry man, I think I have miss guided you.

Actually we need to multiply the pressure with density to get the exact pressure.

to calculate torque: take the z component of the pressure force and viscous.

torque = F*r

F = ((P *A) *density+ tau *A)

A --> area of the patch which you want to calculate torque.

I hope it is clear now

Thanks,
Sivakumar

Hi Sivakumar,

Thats cool,Thanks very much.But I still have some problems.

1.How can I get the area of my patch? cuz its not a regular geometry.

2.If my impeller is rotating around y axiz.Should I pick the y component forces?

3.I have many impellers,How can I get the sum torque?

4.As we know,torque is F*radius, is this radius the max radius?

5.I am using MRFInterFoam, and I add this code into controldict.

outputControl timeStep;
        outputInterval 1000;
        patches (intermigOpposite1);
//        pname p;
//        Uname U;
        rhoName rhoInf;
        log true;
        rhoInf 1000;
        CofR (0 0 0);
    }

cuz the liquid is water.isnt it be 1000?

and I get the forces:

forces output:
    forces(pressure,viscous)((-0.111013 9.4727 -0.09283),(0.00142333 -0.00100197 0.0121232))
    moment(pressure,viscous)((-0.0196299 0.941766 0.19644),(0.00167602 0.0451228 0.000146112))

forces output:
    forces(pressure,viscous)((0.447146 12.7018 -0.170442),(-0.00129838 0.00185303 -0.00024443))
    moment(pressure,viscous)((-0.0262434 0.928373 -0.127495),(1.8784e-05 0.0418585 0.000395807))

forces output:
    forces(pressure,viscous)((-0.0942824 4.63998 0.16988),(-0.00217494 -0.0359383 0.00219822))
    moment(pressure,viscous)((0.0744167 1.1975 0.0769798),(0.00104654 0.0454346 0.00164999))

forces output:
    forces(pressure,viscous)((-0.226807 5.13595 -0.159452),(-0.0119485 -0.0779935 -0.00055085))
    moment(pressure,viscous)((0.015837 1.36635 0.137174),(4.43451e-05 0.0472514 0.00681036))

forces output:
    forces(pressure,viscous)((0.300891 1.99927 0.0561256),(-0.00293573 0.00419335 0.000610194))
    moment(pressure,viscous)((0.0233319 9.04241e-06 -0.05736),(0.000348815 0.000324186 0.000962111))

Now just leave the calculation alone. Does the forces looks normal...
If I want to calculate,for example the first one.
Force=(9.4727-0.0010017)*0.033555(my patch area)=0.3178n
torque=0.3178*0.145(radius)=0.046Nm.

Well, 0.046Nm is too small.

There are many problems,Thanks again for your patience.

Regards.

[sivakumar]

Hi,

To calculate torque, just add the components (moment and viscous) which belongs to the axis of rotation. You no need to think of the area.

In your case for the first output

moment about y axis:
y component of moment + viscous = 0.941766+0.0451228

If you get wrong torque, check your settings and blade angle.

if your torque is small, how small it is? what is the experimental torque value?

Thanks,
Sivakumar

[sharonyue]

Originally Posted by sivakumar

Hi Guys and Laurentb,
I got the right way to calculate torque.

I was doing some wrong calculation, that is why my results were wrong.
now I got the right results, with 2% error.

To Laurentb, I am sorry man, I think I have miss guided you.

Actually we need to multiply the pressure with density to get the exact pressure.

to calculate torque: take the z component of the pressure force and viscous.

torque = F*r

F = ((P *A) *density+ tau *A)

A --> area of the patch which you want to calculate torque.

I hope it is clear now

Thanks,
Sivakumar

Hi sivakumar,

If I add them all.and is this the torque?
y component of moment + viscous = 0.941766+0.0451228 +0.928373 +0.0418585 +.....=4.6

Well , my experiment is a little different with my CFD case. but the torque should not differ much. if 4.6's unit is Nm, I think its normal.

But in you last post,
torque = F*r

F = ((P *A) *density+ tau *A)

A --> area of the patch which you want to calculate torque.

Why is this using area?

Anyway,you give me lots of hints,Thanks very much!!

[sivakumar]

Hi,
In the previous post I used force to calculate torque, so OF out put is interns of pressure force and viscous force.

Pressure = F/A, so I used area.

but, in your case just add the Y component of the pressure moment and viscous moment (this is the normal way).

how many % error you are getting in torque?
how good is your results with experimental results?


Thanks,
Sivakumar

[sharonyue]

Originally Posted by sivakumar

Hi,
In the previous post I used force to calculate torque, so OF out put is interns of pressure force and viscous force.

Pressure = F/A, so I used area.

but, in your case just add the Y component of the pressure moment and viscous moment (this is the normal way).

how many % error you are getting in torque?
how good is your results with experimental results?


Thanks,
Sivakumar

Hi Sivakumar,
My experiment is a little different with my case,so I cannot make a comparision between them.But it should not differ much. If my case's torque is about 4 Nm. its normal. If its below 1 or 2Nm its not normal.

And in my experiment, its two phase.Thats a gas-liquid mixer.Now I dont know how to calculate two phase's torque in FOAM.So...

Well,looks like there are two ways calculating torque,1.use the forces to calculate which would need to know the area.2.use the moment just add them all.But in my case I think the first method is not working...