[loyhj94]

Dear readers,

I am currently trying to get my vehicle to travel in the centre of my curvature domain. It is a 2D simulation with dynamic meshing. However, I am not able to get the angle of motion accurately. My domain is a constant radius of 300m. I have attached my current UDF file below. Please advice. Thank you.

#include"udf.h"

DEFINE_CG_MOTION(velocity, dt, cg_vel, cg_omega, time, dtime)
{
real omega;

omega = 0.4*time;

cg_vel[0] = 0.0; /* x-velocity*/
cg_vel[1] = 70.0*time;
cg_vel[2] = 0.0;

cg_omega[0] = 0.0; /* angular motion*/
cg_omega[1] = 0.0;
cg_omega[2] = omega;

}

[pakk]

However, I am not able to get the angle of motion accurately

Can you be more specific?
What did you expect, and what did you get instead?

[loyhj94]

Originally Posted by pakk

Can you be more specific?
What did you expect, and what did you get instead?

I expect my vehicle to travel in the centre of my domain consistently. However, the vehicle tends to get too close to the left or the right side or touch the sides of the domain, causing the simulation to fail.

[pakk]

Let me rephrase my question. Which trajectory do you want your vehicle to follow (mathematically)?

What you wrote now are equations for a vehicle that is speeding up in y-direction, and rotating faster and faster around the z-axis, I think.

Your problem is not in the coding of the UDF, but in writing down the correct equations for you problem.

[loyhj94]

Originally Posted by pakk

Let me rephrase my question. Which trajectory do you want your vehicle to follow (mathematically)?

What you wrote now are equations for a vehicle that is speeding up in y-direction, and rotating faster and faster around the z-axis, I think.

Your problem is not in the coding of the UDF, but in writing down the correct equations for you problem.

I have attached a link to my picture. My vehicle will be moving from the starting point to the end of the domain which curves to the left.

https://ibb.co/e8pAOH

[pakk]

Great. Now write an equation for this trajectory!

[loyhj94]

Originally Posted by pakk

Great. Now write an equation for this trajectory!

I have tried some equations I found online like:
omega = (2.0*asin(Chord/(2.0*Radius)))*time and
omega = atan((pow(cg_vel[1],2))/(g*r))
However, none of these equations as well as the current one works. Perhaps you will be aware of an equation that I may use?

[pakk]

You don't need to go searching online for equations...

Think back to your lessons on school. Did you never learn equations for circles?

[loyhj94]

Originally Posted by pakk

You don't need to go searching online for equations...

Think back to your lessons on school. Did you never learn equations for circles?

Is it possible to set the centre of the circle for me to use the equation of circle?

[pakk]

Let me tell you what is possible.

- It is possible that you think a bit about which equation to put in. This is movement on a circle, you surely had something about it in school. Think back to those lessons about sine and cosine...
- It is possible that you then put your equation here, and we will help you to fix small mistakes that you might have made.
- It is possible that you then try to translate this mathematical equation into UDF code. Read what I write: "try". You might fail, but at least make an attempt.
- It is possible that we then help you to fix small mistakes in that attempt.
- It is possible that you can then use that code to run your simulation.

But all of that is only possible if you cooperate.

If you want to learn how to do it, I can be your teacher for free.
If you don't care about learning, and only want a solution for your problem, then you can hire me. US$2200,-- please.

[loyhj94]

I have tried using equation of circle to run my vehicle and it seems to work. However, towards the middle of the domain, my vehicle started to drift off from the centre of the domain. I believe it is the value of my radius that is wrong. Do you know how to find the correct value of my radius? Please advise.

#include"udf.h"

DEFINE_CG_MOTION(velocity, dt, cg_vel, cg_omega, time, dtime)
{
real omega;
real x,y;

y = time;
x = -sqrt(20.16-(pow(2.0,y)));
omega = -0.0607;

cg_vel[0] = x; /* x-velocity*/;
cg_vel[1] = 19.444;
cg_vel[2] = 0.0;

cg_omega[0] = 0.0; /* angular motion*/
cg_omega[1] = 0.0;
cg_omega[2] = omega;

}

[pakk]

You have used an equation for the circle that fails if y becomes larger than the radius.

I would have used equations with sine and cosine. If your car drives with constant velocity magnitude, then the position is:

x_pos = R * cos(omega * time)
y_pos = R * sin(omega * time)

So the velocity becomes the derivative:

x_vel = -omega * R * sin(omega * time)
y_vel = omega * R * cos(omega * time)

#include"udf.h"

DEFINE_CG_MOTION(velocity, dt, cg_vel, cg_omega, time, dtime)
{
    real omega = -0.0607;
real R = 300;

cg_vel[0] = -omega * R * sin(omega * time);
cg_vel[1] = omega * R * cos(omega * time);
cg_vel[2] = 0.0;

cg_omega[0] = 0.0; /* angular motion*/
cg_omega[1] = 0.0;
cg_omega[2] = omega;

}

[loyhj94]

Originally Posted by pakk

You have used an equation for the circle that fails if y becomes larger than the radius.

I would have used equations with sine and cosine. If your car drives with constant velocity magnitude, then the position is:

x_pos = R * cos(omega * time)
y_pos = R * sin(omega * time)

So the velocity becomes the derivative:

x_vel = -omega * R * sin(omega * time)
y_vel = omega * R * cos(omega * time)

#include"udf.h"

DEFINE_CG_MOTION(velocity, dt, cg_vel, cg_omega, time, dtime)
{
    real omega = -0.0607;
real R = 300;

cg_vel[0] = -omega * R * sin(omega * time);
cg_vel[1] = omega * R * cos(omega * time);
cg_vel[2] = 0.0;

cg_omega[0] = 0.0; /* angular motion*/
cg_omega[1] = 0.0;
cg_omega[2] = omega;

}

Thank you for the help. But I don't understand the code fully. Do you mind explaining the code to me in detail? I really appreciate the help you offer.

[pakk]

Originally Posted by loyhj94

Thank you for the help. But I don't understand the code fully. Do you mind explaining the code to me in detail? I really appreciate the help you offer.

What is the part that is giving you problems to understand?

[loyhj94]

It's fine. I managed to understand the code. Thank you for the help!