Prevent Rigid Body from rotating

jaxk · January 24, 2023, 13:31

Aim: prevent Rigid Body from rotating around z-axis while ‘Z axis Rotational Degrees of Freedom’ is on.

Hi, I have a transient simulation of a RB in a flow. The RB is free to translate in x and rotate in z. The flow results in a moment on the RB, and as a result the RB will rotate, which is at some point in time unwanted. At that moment I would like to add an External Torque equal but opposite to the torque caused by the flow, such that the total torque on the RB is zero. Does anyone know which value I should use for the torque?
The RB is located at object1 and has its centre of mass initially at its own coordinate frame ‘coord1’. The ZZ component of its mass moment of inertia is ‘Iz’.

I tried T=torque_z()@object1 and also T=Iz*rbstate(Angular Acceleration Z)@Rigid Body 1.
And I tried T=torque_z_coord1()@object. In POST there is ‘Normal Moment on object1 (Z)’ but I can not try this value because it is not available in PRE.

None of the External Torques result in a value which lead to zero total torque on (and thus zero rotation of) the RB. Is there anyone with a suggestion? Thanks!

ghorrocks · January 24, 2023, 16:54

Why don't you just do an initial simulation where you lock that rotation axis, and use it as an initial condition for a simulation which then releases that axis?

jaxk · January 25, 2023, 04:39

That works, but would result in a lot of separate simulations since the RB should move at certain times and not at other times, alternately.

To remove the force on the RB caused by the flow I can simply add an opposing External Force of force_x()@object1. I was hoping for something similar to remove the torque.

Since at the External Torque section I can add a torque on the centre of mass of the RB, I need to know the torque caused by the flow with respect to this centre of mass.

evcelica · January 25, 2023, 12:39

So when you use the same method you do for translation and apply torque_z()@object1, what happens?

If it isn't resulting in 0 net torque, perhaps it is due to the force and coordinate frame offset being included in the torque values?
Export your applied torque from your function as well as the resulting torque, force, and translation and you should be able to post process these numbers to see what if going on.

But If it is already rotating and you want to stop it, you don't want to just equal the flow torque, that will keep it moving at the same rotational velocity.
Your statement: "zero total torque on (and thus zero rotation of)" is incorrect unless the object is not already rotating. You mean "zero total torque on (and thus zero angular acceleration of)"
You have to apply a torque related to the angular velocity if you want to stop it when it is moving.

If you want to keep it rotating at the same velocity, then post process and fix what your function is doing like I said earlier.
Another thought would be, If you can, make the moment of area a function of time, where you increase it to a very high value when you want constant angular velocity. But it sounds like you want to actually be able to stop it when it is already rotating?

jaxk · January 26, 2023, 04:31

Indeed, the result is no 0 net torque, which is what I want. I also think it is due to the coordinate frame offset. But I thought the value of the torque_z would be the same if the coordinate frame has an offset in x.
I had a look in post, where I added various coordinate frames along the x-axis. When I evaluate the torque_z, I can select on of these coordinate frames, which each give a different torque_z value. So it looks like the value does depend on the location of the coordinate frame, but I haven't figured out the math behind it for the coordinate transformation.

I agree with your comment on the torque/angular acceleration

Thanks for the 'moment of area'-suggestion! I'll give it a try

jaxk · January 26, 2023, 05:56

Parameter 'zzValue' in object '/FLOW:Flow Analysis 1/RIGID BODY:Rigid Body 1/MASS MOMENT OF INERTIA' is not allowed to be assigned an expression value that depends on variables. It must be assigned a numeric value, or an expression that resolves to a constant value.

It is not allowed

evcelica · January 30, 2023, 17:36

So varying the Moment of inertia is not an option.
You will have to figure out the math behind the output torque values then. I'd start looking into subtracting out the vertical force * deformed X coordinate from global origin. I believe that is included in your torque variable.
Forces on the surface can be normal (Pressure) forces as well as tangential (shear) forces, so I can see how this could become a bit tricky.

January 24, 2023, 13:31	Prevent Rigid Body from rotating	#1
jaxk New Member Join Date: Dec 2022 Posts: 29 Rep Power: 3	Aim: prevent Rigid Body from rotating around z-axis while ‘Z axis Rotational Degrees of Freedom’ is on. Hi, I have a transient simulation of a RB in a flow. The RB is free to translate in x and rotate in z. The flow results in a moment on the RB, and as a result the RB will rotate, which is at some point in time unwanted. At that moment I would like to add an External Torque equal but opposite to the torque caused by the flow, such that the total torque on the RB is zero. Does anyone know which value I should use for the torque? The RB is located at object1 and has its centre of mass initially at its own coordinate frame ‘coord1’. The ZZ component of its mass moment of inertia is ‘Iz’. I tried T=torque_z()@object1 and also T=Iz*rbstate(Angular Acceleration Z)@Rigid Body 1. And I tried T=torque_z_coord1()@object. In POST there is ‘Normal Moment on object1 (Z)’ but I can not try this value because it is not available in PRE. None of the External Torques result in a value which lead to zero total torque on (and thus zero rotation of) the RB. Is there anyone with a suggestion? Thanks!

January 24, 2023, 16:54		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,712 Rep Power: 143	Why don't you just do an initial simulation where you lock that rotation axis, and use it as an initial condition for a simulation which then releases that axis? __________________ Note: I do not answer CFD questions by PM. CFD questions should be posted on the forum.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
How to define movement/deformation relative to a rigid body motion in Fluent by udf?	eagle_001	Fluent UDF and Scheme Programming	6	May 17, 2022 03:41
Rigid Body airfoil CFX	androm	CFX	15	March 18, 2020 12:12
Fluid Domain moving with Rigid body	Lloyd Sullivan	CFX	3	August 17, 2018 09:58
Rigid body in creeping flow - COMSOL	MultiPhy	CFD Freelancers	1	June 8, 2017 01:38
Rigid Body State Variables in Solid Immersed Simulation	Hamidreza	CFX	1	October 19, 2009 06:14

January 25, 2023, 04:39		#3
jaxk New Member Join Date: Dec 2022 Posts: 29 Rep Power: 3	That works, but would result in a lot of separate simulations since the RB should move at certain times and not at other times, alternately. To remove the force on the RB caused by the flow I can simply add an opposing External Force of force_x()@object1. I was hoping for something similar to remove the torque. Since at the External Torque section I can add a torque on the centre of mass of the RB, I need to know the torque caused by the flow with respect to this centre of mass.

January 25, 2023, 12:39		#4
evcelica Senior Member Erik Join Date: Feb 2011 Location: Earth (Land portion) Posts: 1,167 Rep Power: 23	So when you use the same method you do for translation and apply torque_z()@object1, what happens? If it isn't resulting in 0 net torque, perhaps it is due to the force and coordinate frame offset being included in the torque values? Export your applied torque from your function as well as the resulting torque, force, and translation and you should be able to post process these numbers to see what if going on. But If it is already rotating and you want to stop it, you don't want to just equal the flow torque, that will keep it moving at the same rotational velocity. Your statement: "zero total torque on (and thus zero rotation of)" is incorrect unless the object is not already rotating. You mean "zero total torque on (and thus zero angular acceleration of)" You have to apply a torque related to the angular velocity if you want to stop it when it is moving. If you want to keep it rotating at the same velocity, then post process and fix what your function is doing like I said earlier. Another thought would be, If you can, make the moment of area a function of time, where you increase it to a very high value when you want constant angular velocity. But it sounds like you want to actually be able to stop it when it is already rotating?

January 26, 2023, 04:31		#5
jaxk New Member Join Date: Dec 2022 Posts: 29 Rep Power: 3	Indeed, the result is no 0 net torque, which is what I want. I also think it is due to the coordinate frame offset. But I thought the value of the torque_z would be the same if the coordinate frame has an offset in x. I had a look in post, where I added various coordinate frames along the x-axis. When I evaluate the torque_z, I can select on of these coordinate frames, which each give a different torque_z value. So it looks like the value does depend on the location of the coordinate frame, but I haven't figured out the math behind it for the coordinate transformation. I agree with your comment on the torque/angular acceleration Thanks for the 'moment of area'-suggestion! I'll give it a try

January 26, 2023, 05:56		#6
jaxk New Member Join Date: Dec 2022 Posts: 29 Rep Power: 3	Parameter 'zzValue' in object '/FLOW:Flow Analysis 1/RIGID BODY:Rigid Body 1/MASS MOMENT OF INERTIA' is not allowed to be assigned an expression value that depends on variables. It must be assigned a numeric value, or an expression that resolves to a constant value. It is not allowed

January 30, 2023, 17:36		#7
evcelica Senior Member Erik Join Date: Feb 2011 Location: Earth (Land portion) Posts: 1,167 Rep Power: 23	So varying the Moment of inertia is not an option. You will have to figure out the math behind the output torque values then. I'd start looking into subtracting out the vertical force * deformed X coordinate from global origin. I believe that is included in your torque variable. Forces on the surface can be normal (Pressure) forces as well as tangential (shear) forces, so I can see how this could become a bit tricky.