How to implement the 8° upflow angle required by the standard IEC 61400-1?

[Sarah.Lott]

Hello,

I want to run simulations in compliance with IEC 61400-1 Edition 4.0 (2019). The following is written there:
“In all cases, the influence of an inclination of the mean flow with respect to a horizontal plane of 8° shall be considered. This flow inclination angle shall be assumed to be invariant with height.”
So, I was wondering how to implement this 8° upflow angle.

We have to distinguish between two cases:

• Turbulent inflow → WindType=3 (in InflowWind) → use of TurbSim
• Uniform inflow → WindType=2 (in InflowWind) → create time series data with IECWind or individually

First question – turbulent inflow:

I assume that the upflow angle of 8° can be defined using the parameter VFlowAng. This parameter appears twice: in InflowWind and in TurbSim.

Where do I set it, in InflowWind or in TurbSim? Does it even make a difference?

Second question – uniform inflow:

I can think of three different options to define the upflow angle:

• A: use of parameter VFlowAng in InflowWind, InflowWind|VFlowAng = 8
• B: in the definition of the uniform wind field the resulting wind speed can be split into a horizontal and a vertical component, InflowWind|VFlowAng = 0
• C: use of the 9^th column (upflow angle) in the definition of the uniform wind field, InflowWind|VFlowAng = 0

Let me use an example to explain the differences.
Suppose I would like to simulate a wind speed of 9m/s, then the input data would look like this

Option A:
InflowWind|VFlowAng = 8
Time series data for uniform wind field:
! Time (s) Wind speed (m/s) Wind direction (deg) Vertical speed (m/s) Horizontal shear (-)
Power law exponent (-) Vertical shear (-) Gust speed (m/s) Upflow angle (deg)
0.0000 9.0000 0.0000 0.0000 0.0000 0.2000 0.0000 0.0000 0.0000

Option B:
InflowWind|VFlowAng = 0
Time series data for uniform wind field:
! Time (s) Wind speed (m/s) Wind direction (deg) Vertical speed (m/s) Horizontal shear (-)
Power law exponent (-) Vertical shear (-) Gust speed (m/s) Upflow angle (deg)
0.0000 8.9124 0.0000 1.2526 0.0000 0.2000 0.0000 0.0000 0.0000

Option C:
InflowWind|VFlowAng = 0
Time series data for uniform wind field:
! Time (s) Wind speed (m/s) Wind direction (deg) Vertical speed (m/s) Horizontal shear (-)
Power law exponent (-) Vertical shear (-) Gust speed (m/s) Upflow angle (deg)
0.0000 9.0000 0.0000 0.0000 0.0000 0.2000 0.0000 0.0000 8.0000

For the evaluation at hub height, I would expect the horizontal wind speed to be
v_hor = cos(8°) * 9 m/s = 8.912 m/s
and the vertical wind speed to be
v_vert = sin(8°) * 9 m/s = 1.253 m/s

These are the results from an OpenFAST simulation using the three wind fields as defined above:

=> option B and option C meet the expectations

I defined another point (~22m below hub height) to output the velocities in InflowWind. Here, too, there are significant differences between the various options, as you can see below.

The upflow angle is calculated as alpha = arctan(VelZ / VelX).
Option A: alpha = 8.00 deg
Option B: alpha = 8.62 deg
Option C: alpha = 8.00 deg

=> option A and option C meet the expectations

Can you explain to me how the different results come about? What is the correct way to define the upflow angle?

Why I came up with this question?
In my simple example, it wouldn’t matter at all which option is chosen for the uniform wind field, because the turbine loads hardly differ at all. But I’ve run simulations for DLC 1.4 using ECD, and there are significant differences.

Best regards,
Sarah

[Bonnie.Jonkman]

Hi Sarah,

There are a few subtle differences that may help in your analysis:

• The TurbSim VFlowAng uses the sine and cosine at each individual point in the 3-D grid. So, the wind speed at individual points is rotated, but the grids are not. Figure 6 in the TurbSim User’s Guide shows this rotation:

  

• The InflowWind VFlowAng parameter rotates the grids relative to the point [0, 0, HubHeight]. At each timestep, InflowWind essentially has a box of wind speeds around the turbine, and this rotates the full box. This is the option I’d use for turbulence cases; it seems to make the most sense with Taylor’s Frozen Turbulence Hypothesis.

• The UniformWind files (formerly called Hub-Height Wind files) can also be rotated using the the upflow angle in column 9. If you just manually set column 4 with the vertical component calculated to be 8-degrees, each point in space at a given time will have the same vertical wind speed as you defined. However, they could have different horizontal speeds, so not all points would have 8-deg upflow. If you use column 9, it calculates the wind speed at each point in space and then rotates it using sine and cosine (somewhat like the TurbSim VFlowAng parameter with the turbulence planes). You will notice a difference in the two options if your UniformWind file has shear. If it doesn’t have shear defined (i.e., columns 5-7 are 0), your options B and C should be the same.

• The other thing to note is that the InflowWind WindHubVel* output channels are not fixed to the inertial frame. They report the wind measurements at the position in space where the structural code says the hub is located. I would expect that if you requested outputs at the position [0, 0, HubHeight] using Wind#Vel*, that all 3 methods would give you the same results.

For turbulence cases (full-field wind files), I would use the InflowWind VFlowAng parameter for the upflow angle.

It’s difficult for me to say what the “correct” way to implement the upflow angle for the uniform wind files is. I think your options A and C are both reasonable, given that these wind files are already pretty idealized. I’d probably choose option A, but I know another major code uses option C.

Here’s another figure from the TurbSim User’s Guide to show the differences between the turbulence and uniform wind files as implemented in InflowWind:

[Sarah.Lott]

Hi Bonnie,

thank you very much for the quick response and the detailed explanations. That’s really helpful.

Best regards,
Sarah