CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Wind Turbine Simulation (http://www.cfd-online.com/Forums/main/69229-wind-turbine-simulation.html)

akrsjak October 16, 2009 03:05

Wind Turbine Simulation
 
Hello,

Can somebody help me? I am searching for the right software to simulate a wind tubine :confused::confused::confused:
I want from this software:
1. Full visualization of the results, also animation of solids and fluids
2. Output parameters like speed, pressure, tourge, ...
3. My boundary conditions are only the wind speed and wind direction

Thanks a lot,
Andrej

andy_ October 16, 2009 11:45

What is your budget and how large are your computational facilities?

It is not clear what information you want from the simulation. What do you want to do with it?

Wind direction and speed is not enough input information. I assume you know the geometry but what about varying blade angles, load, boundary layer condition, etc...?

Do you need all the blades, the tower and ground or just the flow main part of the blade? Details of tip flow?

Many questions about wind turbine design are better served with BEM-type approaches. If you need CFD, are steady-state RANS approaches appropriate or does the desired information dictate unsteady LES approaches?

akrsjak October 16, 2009 12:00

1. The budget is very high, more than $100.000, so we can buy any software and hardware on market, there is no problem.
2. We want to simulate an easy three blade horizontal turbine and compare it to vertical designs and also optimize the blades design.
3. We want to run more types of simulations:
a) inputs are: wind speed and design, output is: torque, rotor speed, behaivior of the medium around blades and all solid parts in time (velocity, pressure, temperature,...)
b) we want also to run stress-strain analysis, but we can do this, but need outputs from a)
4. We want to simulate the whole construction (body, stator, rotor, blades) -everything.

andy_ October 16, 2009 13:32

Not sure this isn't handled better via email (I have a signficant interest in this area) but it is also a fairly popular topic on the forum. Let me know.

> 1. The budget is very high, more than $100.000, so we can buy any software and
> hardware on market, there is no problem.

Your budget is not high if you wish to pursue a well resolved unsteady approach for the full geometry. This would give you a significantly enhanced capability when it came to predicting stall, turbulence effects and a fully coupled fluid/structure.

If you back off and look to simulating the flow using steady-state turbulence models the hardware costs come down to a modest parallel computer, the option to use established commercial codes but the uncertainty about accuracy grows dictating more resources to get a handle on it.

The former is where things are going in the near term or now if you are ambitious and the latter is where the industrial wind turbine companies that have adopted CFD are now. The (high tech) aerospace industry is a bit further on although their needs in terms of simulating attached/detached flow, on/off design, thick/thin blades, stable inlet condition and the like is reflected to do some degree in the modelling approaches adopted.

> 2. We want to simulate an easy three blade horizontal turbine and compare it to
> vertical designs and also optimize the blades design.

On design - OK. Optimise it for what? What much do you have to determine if the results are good or bad?

> 3. We want to run more types of simulations:
> a) inputs are: wind speed and design, output is: torque, rotor speed, behaivior of
> the medium around blades and all solid parts in time (velocity, pressure,
> temperature,...)

If you want rotor speed as an output you need to a model for the work extracted.

The simulation of the medium around the blade involves turbulence, stall and possibly sound generation if this of interest. These are nonlinear effects where the small scale motion can have significant influence on the large scale motion in a way they don't for the stress in the blade. (Although simulating things like cracking fibres, delamination, etc... can make the structural side a bit more fun but no where near as much as the fluid side!).

> b) we want also to run stress-strain analysis, but we can do this, but need outputs
> from a)

The dynamic side is not of interest within this project?

> 4. We want to simulate the whole construction (body, stator, rotor, blades)
> -everything.

One big simulation following all the molecules in the solid and fluid would put many of us engineers out of a job. Thankfully that is far in the future and we are still need to sort out how to extract the desired information from applying the range of modelling assumptions currently available.

akrsjak October 17, 2009 03:30

check this out:
http://www.youtube.com/watch?v=1j2wjbtiEuM
we want to simulate something complex like this, but the rotor will be driven by wind and not heat...
there are softwares that can do complex simulation and I wanted only to know the best choise...

andy_ October 17, 2009 06:58

> we want to simulate something complex like this, but the rotor will be driven
> by wind and not heat...

At the risk of causing offense, the rotor is driven by wind in your example. Heat generated the wind but it is primarily the mechanical energy in the wind that is being extracted to turn the rotor.

> there are softwares that can do complex simulation and I wanted only to know
> the best choise...

The best choice will depend on what type of information you require from the simulation. This is not clear to me. Are you going to manufacture wind turbines or provide a consulting service?

I had assumed in my previous reply that you were familiar with R&D in the wind turbine industry and were looking to bring in CFD as part of the process to address some of the issues that are not handled by less computationally intensive approaches. There is a fair amount of activity along these lines at the moment and 100k is a typical amount for an industrial company to allocate to see if CFD is actually going to provide something useful.

Almost every supplier of CFD software will claim it is suitable for your purposes which is very unlikely to be the case. In addition, the appropriate modelling approach is in a state of transition as the cost of computing continues to fall. Investing in the established methodology as expressed in current commercial codes may or may not be appropriate depending on your situation.

I will predict that if you simply purchase a computer and some CFD software and they are used by people inexperienced in wind turbines and the modelling assumptions in the CFD software you will end up with a set of nice looking movies that provide very little useful information for making engineering decisions. This will be because you will not have a rational basis on which to determine what is reliably simulated and what is not.

Of course, you can train up internal people and/or hire expertise but it may be wise to put this is in the initial programme because these are substantial costs in terms of time and money. Again, almost all CFD software suppliers will suggest training up internal people, almost all engineering consultants will suggest hiring external expertise and universities may well suggest both. The most appropriate solution will depend on where you are now and where you want to go.

Lysistrata October 18, 2009 08:53

Quote:

Originally Posted by akrsjak (Post 232939)
1. The budget is very high, more than $100.000, so we can buy any software and hardware on market, there is no problem.
2. We want to simulate an easy three blade horizontal turbine and compare it to vertical designs and also optimize the blades design.

There are several different vertical axis wind turbines. The most efficient ones use blades that can pitch. You did know that, didn't you? Or were you going to compare horizontal axis with fixed pitch vertical axis turbines?

akrsjak October 18, 2009 09:15

What kind of VAWT or HAWT are most efficient? We live in Slovakia and the wind condintions are not good. Avarage wind speed is only 2-3m/s... therefore we need the most efficiect available solution. Do you have some links where we can compare different designs?

What do you mean with pitch? You mean that during function the pitch can be modified?

Lysistrata October 20, 2009 12:05

Quote:

Originally Posted by akrsjak (Post 233093)
What kind of VAWT or HAWT are most efficient? We live in Slovakia and the wind condintions are not good. Avarage wind speed is only 2-3m/s... therefore we need the most efficiect available solution. Do you have some links where we can compare different designs?

What do you mean with pitch? You mean that during function the pitch can be modified?

I would not bother with wind if you only get averages of 2 or 3m/sec. Wind turbines are not efficient at low Reynolds numbers. To get high Rn you will need to use very large blades and that means a lot of material (and hence high cost) to get a reasonable amount of power.

By pitch, I mean that the blades of a VAWT can swivel. This means that they can be made to pitch so as to delay stall. The machine is more compicated, but far more efficient. VAWT are not as effiicient as HAWT for generating electricity, but there are some unusual applications where they are better suited.


All times are GMT -4. The time now is 07:53.