[niyas]

Hi,
In this post, I will be discussing the method for obtaining angular velocity in a cylinder in a 3D numerical simulation. I am utilizing Fortran-based in-house coding to simulate a bar and a cylinder. The simulation for the bar (referred to as Figure-1) was successful and I had no issues. For the bar, I used the following initial velocity conditions: a positive 'u' velocity along the x-axis, a 'v' velocity of zero along the y-axis, and a 'w' velocity of zero along the z-axis. This resulted in a streamlined pattern, calculated using √(u^2+v^2+w^2 ), as shown in Figure-2.

Regarding the 3D cylinder simulation (Figure-3), using the same procedure for initializing the velocities did not produce the desired result, as seen in Figure-4. I want to achieve a positive angular velocity within the cylinder, but I am unsure how to specify the 'u' and 'v' velocities to achieve this. I know the 'w' velocity must be zero, but I need a solution for the 'u' and 'v' velocities that will result in the desired angular velocity, as shown in Figure-5.

[sbaffini]

Originally Posted by niyas

Hi,
In this post, I will be discussing the method for obtaining angular velocity in a cylinder in a 3D numerical simulation. I am utilizing Fortran-based in-house coding to simulate a bar and a cylinder. The simulation for the bar (referred to as Figure-1) was successful and I had no issues. For the bar, I used the following initial velocity conditions: a positive 'u' velocity along the x-axis, a 'v' velocity of zero along the y-axis, and a 'w' velocity of zero along the z-axis. This resulted in a streamlined pattern, calculated using √(u^2+v^2+w^2 ), as shown in Figure-2.

Regarding the 3D cylinder simulation (Figure-3), using the same procedure for initializing the velocities did not produce the desired result, as seen in Figure-4. I want to achieve a positive angular velocity within the cylinder, but I am unsure how to specify the 'u' and 'v' velocities to achieve this. I know the 'w' velocity must be zero, but I need a solution for the 'u' and 'v' velocities that will result in the desired angular velocity, as shown in Figure-5.

You may want to look at this page and consider the function ATAN2 in Fortran.

In the end, your cartesian velocity components and at given and where you know the tangential velocity are given by:

[niyas]

@sbaffini, the command you suggested was very effective (Figure-1). Thank you!!

Is it possible to set the 'u' and 'v' velocities as constant within the cylinder?

At present, the 'u-velocity' (Figure-2) and 'v-velocity' (Figure-3) have varying values within the cylinder.

Is there a method to make them constant?

I aim to maintain a steady cartesian flow-field velocities within the cylinder.

[sbaffini]

I probably don't understand. How can the two velocities be constant AND parallel to the cylinder surface at the same time?

You kind of made a clear picture of the streamlines you wanted, and those can't be achieved by constant u and v.

[niyas]

"In a 3D bar, there is an initial condition of a positive uniform velocity of 5.5 m/s (u = 5.5 m/s and v = 0) flowing from one end to the other.

Similarly, I need to create a uniform velocity field within a cylinder as an initial condition. Instead of using linear velocity (u), should I use angular velocity (omega)?

Omega is calculated as the tangential velocity divided by the radius.

If I opt for omega, then I won't have 'u', 'v', and 'w'. Instead, I'll only have 'omega' and 'w'. Will this necessitate a modification of the entire code structure, including the governing equations, etc.?

My aim is to replicate the results obtained in a 3D bar in a 3D cylinder. Initially, I thought I could simply change the geometry from a bar to a cylinder to obtain the same results. Will this approach work?

[sbaffini]

Nothing of this really helps.

Flow in your geometry can either be fully tangential with no axial component, like in my first answer, or only axial with no tangential component (in this case your velocity will have just the z component of your choice) or both (you combine my first answer with your preferred z component).

Any other initialization will give you a velocity going into or out of walls.

If you have a reference work you want to replicate, it is better if you just put a link here of the whole work

[niyas]

I appreciate your valuable insights and thank you for taking the time to share them.

If it's not too much trouble, would you kindly provide me with your email address? I'd like to share more details with you about the project. Alternatively, if you prefer, you could simply send a quick "hi" to my email address at niyasdeen.aero@gmail.com.

[sbaffini]

Originally Posted by niyas

I appreciate your valuable insights and thank you for taking the time to share them.

If it's not too much trouble, would you kindly provide me with your email address? I'd like to share more details with you about the project. Alternatively, if you prefer, you could simply send a quick "hi" to my email address at niyasdeen.aero@gmail.com.

Don't take me wrong, but I can't. I mean, my email is kind of everywhere and you can pick it up and just send an email, but don't expect an answer anytime soon or an answer at all. I usually make exceptions only when the topic is very specific and I am kind of into the argument myself, but this is not the case.

My suggestion is to clearly state your problem here, what you want to do, how, where you have problems and which problems. Then anyone, including me, is here to help.

But, at least personally, I'm willing to help when my contribution is around 10% at most, possibly less. I want the freedom to just drop any correspondence when I feel that there aren't the conditions anymore to help. If you write to me, I still have that freedom of course, but then I look like an a..hole, which is not nice in any case.

In your specific case, you should better describe your problem, in my opinion. But maybe someone else will understand, which is your main advantage from keeping this thing here instead of private email.

[niyas]

Thank you for sharing your thoughts. I will take your suggestions into consideration and provide more details about my problem in the future.

Additionally, I wanted to let you know that I'm planning to move forward with no velocity as an initial condition, as you mentioned. However, I've bumped into another issue that I will create in a separate thread. Thank you again for your help and advice.