[tjwldhks79]

I am investigating the drag coefficient based on the Reynolds number.

In the laminar flow regime, I confirmed a reasonably good agreement for the drag coefficient through two-dimensional analysis.

Now, as we move beyond laminar flow, turbulence occurs, necessitating three-dimensional analysis.

Therefore, I conducted 3D simulations, taking into account Reynolds numbers that vary with the diameter D of the cylinder.

However, while the diameter remains constant, the length of the cylinder does not influence the Reynolds number.(The frontal area varies with length.)

Could you explain why the drag coefficient varies with spanwise direction in this case?

Your insights on this matter would be greatly appreciated.

[shereez234]

Originally Posted by tjwldhks79

I am investigating the drag coefficient based on the Reynolds number.

In the laminar flow regime, I confirmed a reasonably good agreement for the drag coefficient through two-dimensional analysis.

Now, as we move beyond laminar flow, turbulence occurs, necessitating three-dimensional analysis.

Therefore, I conducted 3D simulations, taking into account Reynolds numbers that vary with the diameter D of the cylinder.

However, while the diameter remains constant, the length of the cylinder does not influence the Reynolds number.(The frontal area varies with length.)

Could you explain why the drag coefficient varies with spanwise direction in this case?

Your insights on this matter would be greatly appreciated.

A picture of how you setup the flow past cylinder would be helpful as you can orient cylinder in any direction (vertically straight with Length going from -z to +z or horizontally with length going from -y to +y). This will help to understand what you mean by spanwise.

And very briefly speaking increase in reference surface area (S) would increase Drag force (N)
with Q being dynamic pressure.

[tjwldhks79]

Originally Posted by shereez234

A picture of how you setup the flow past cylinder would be helpful as you can orient cylinder in any direction (vertically straight with Length going from -z to +z or horizontally with length going from -y to +y). This will help to understand what you mean by spanwise.

And very briefly speaking increase in reference surface area (S) would increase Drag force (N)
with Q being dynamic pressure.

The flow direction, as seen in the image, is along the x-axis.

Additionally, the term "spanwise," mentioned in the text, refers to the z-axis.

Naturally, as the spanwise dimension increases, the frontal area expands, leading to an increase in drag force (N), which is expected.

What I am curious about is the drag coefficient.

The drag coefficient varies with the Reynolds number in this scenario.

However, what influences the Reynolds number in this situation is likely the diameter of the cylinder, not the length (spanwise).

Thanks.

[shereez234]

Originally Posted by tjwldhks79

The flow direction, as seen in the image, is along the x-axis.

Additionally, the term "spanwise," mentioned in the text, refers to the z-axis.

Naturally, as the spanwise dimension increases, the frontal area expands, leading to an increase in drag force (N), which is expected.

What I am curious about is the drag coefficient.

The drag coefficient varies with the Reynolds number in this scenario.

However, what influences the Reynolds number in this situation is likely the diameter of the cylinder, not the length (spanwise).

Thanks.

Apologize, I am still not sure what your doubt is. Of course with Reynolds number drag coefficient will vary and not in a linear way as well. The drag crisis will occur when flow changes from laminar to transitional then to turbulent..

cd.png

If Reynolds is same then Cd should not vary even if you increase area, because the increase in Area will be compensated by increase in pure drag force (N) thus balancing Cd..

[tjwldhks79]

Originally Posted by shereez234

Apologize, I am still not sure what your doubt is. Of course with Reynolds number drag coefficient will vary and not in a linear way as well. The drag crisis will occur when flow changes from laminar to transitional then to turbulent..

Attachment 98601

If Reynolds is same then Cd should not vary even if you increase area, because the increase in Area will be compensated by increase in pure drag force (N) thus balancing Cd..

As mentioned in the last sentence, if the Reynolds number remains constant, increasing the area should not change the drag coefficient (Cd).

In the images I provided, the diameter (D) of the cylinder varies with the Reynolds number.

However, when keeping this diameter constant and only altering the height (length along the z-axis) of the cylinder in the simulation, the resulting drag coefficients are different.

All simulation options were kept identical, with the only variation being the height of the cylinder.

Thank you for your attention and consideration.

[shereez234]

Originally Posted by tjwldhks79

As mentioned in the last sentence, if the Reynolds number remains constant, increasing the area should not change the drag coefficient (Cd).

In the images I provided, the diameter (D) of the cylinder varies with the Reynolds number.

However, when keeping this diameter constant and only altering the height (length along the z-axis) of the cylinder in the simulation, the resulting drag coefficients are different.

All simulation options were kept identical, with the only variation being the height of the cylinder.

Thank you for your attention and consideration.

Do you have symmetry plane on two sides?

[LoGaL]

Ye, I am wondring the same thing. If you did not impose a symmetry condition, then the increase of length translates into a larger rectangular pipe with larger characteristic dimension, hence affecting the reynolds number of the flow.

Also, not sure how you are including turbulence? is this still "laminar" model, i.e. bare bone navier stokes? RANS?

[tjwldhks79]

Originally Posted by shereez234

Do you have symmetry plane on two sides?

Are you referring to the two surfaces in the z-axis direction?

Yes, that's correct. I applied symmetry.

Thanks.

[evcelica]

Repeating LoGal's question:
Also, not sure how you are including turbulence? is this still "laminar" model, i.e. bare bone navier stokes? RANS?

What is your inlet turbulence specification? This could possibly be different depending on how you defined inlet turbulence if the length scale used to compute these values is different.