I am a structural engineer mainly working on telecom towers. As usual I use codes to predict the wind loads on structures but recently I am curious to learn how CFD can be effectively used to get the more realistic drag coefficients.

I tried Simscale to compare between drag coefficients from codes so I created a simple circular pipe section(114mm OD x 4.5mm thk.). Most of the **code suggest Cd as 1.2** for smooth circular section.

below results shows the variation of drag coefficient with time. **Initially for up to150s the variation is too much (25.19 to -6.9). from 150s to 250s the variation is small (-1.17 to -1.06) after250s it is average of -0.93.**

My question is how can we **determine correct drag coefficient** and why there is too much of **fluctuation at beginning?**

How can we determine a **over all drag coefficient for cluster of steel work** for example a portion of lattice tower?