For my thesis I am using the MIT/NREL TLP and NREL 5-MW turbine to design a damper system. I am building a model in Matlab that will model the motions of the structure. For this I am using the WAMIT data that is available in the FAST download. I have some questions regarding the data that is available.
I would like to verify the mass moment of inertia I calculated for pitch, because my natural frequency is not close to the one from the master thesis of Denis Matha (NREL/SR-500-45891). Is the pitch inertia about the CM that is listed in NREL/CP-500-46726 including the inertia of the platform (mass and added mass) and wind turbine?
Is it possible to share the WAMIT output .4 file so I can verify the results of my Matlab model? I don’t have access to WAMIT so I can’t create this file myself.
Thanks in advance,
The pitch inertia about the CM of the MIT/NREL TLP from Table 3 in NREL/CP-500-46726 is for the floating platform only (not including the tower, nacelle, rotor, or tendons) and only considers the body mass, not added mass.
I have a directory made by Denis Matha with many different WAMIT simulations, many of which included the RAO’s from WAMIT (*.4 files); however, it is difficult for me to differentiate which one is which without spending a lot of time examining each one. Regardless, the most important RAO results are plotted in Figure 26 of NREL/SR-500-45891. This plot shows WAMIT-generated and FAST-generated (with tower flexibility enabled) RAOs of the MIT/NREL TLP for two different spoke lengths. Here it is shown that the length of the spokes and the flexibility of the tower have strong effects on the platform-pitch natural frequency. While one can obtain RAOs from WAMIT, WAMIT cannot model the tower flexibility, so in the end, the WAMIT-generated RAOs are not very useful anyway.
Thank you for your reply.
I would like to investigate the differences in stresses in the structure with and without damper. For this I need to create structural model, I have not been able to find a lot of details. Are you aware if there exist more details available of the floater (more specifically the spokes)? Such as cross section, outer dimenions and wall thickness ? There is some information on the main hull (cylinder) but only the length of the spokes is given. Right now I have made an estimate based on this paper by Bachynski (http://www.sciencedirect.com/science/article/pii/S0951833912000627), but he assummed a rigid structure, which is not usefull for my research.
Unfortunately, the structural design (member thickness, compartmentalization, scantling, etc.) was never detailed for the MIT/NREL TLP. Matha et al’s work was more focused on global performance than detailed structural design. You’ll have to make your own assumptions for your own purposes.
As I explained in the first post of this topic I am trying to calculate the eigenfrequencies of the rigid body modes that belong to the MIT/NREL-TLP. For my non-rigid model with tuned mass damper to work and be valid, I must be 100% sure that my inertia’s and stiffness are correct. I am a lot closer now than a while back but I still have some questions. In the attachment the matrices can be found, because I could not put them in this post.
- I determined the stiffness in surge, sway, heave, roll, pitch and yaw w.r.t. the keel.
- I determined the structural mass matrix.
- The added mass matrix is formed by taking the values from the WAMIT results for when the frequency goes to infinity. The added mass is determined w.r.t. the centre of floatation (centre of water plane area).
- The stiffness matrix and the structural mass matrix are transformed to the centre of floatation. And now the structural mass matrix and added mass matrix can be added together.
To find the eigenfrequencies, I used the eigenfunction in Matlab (eig(K/M)). Resulting in:
- Surge: 0.016958 Hz
- Sway: 0.016958 Hz
- Heave: 0.43817 Hz
- Roll: 0.17516 Hz
- Pitch: 0.17523 Hz
- Yaw: 0.07994 Hz
Comparing these values with NREL/SR-500-45891, the eigenfrequencies for surge and sway differ 2.7%, heave 0.15% which I thought is a reasonable difference. The differences between my values for roll, pitch and yaw are 20%, 21% and 18%. The difference is too much and I am trying to understand what causes these differences. Therefore, I have the following questions:
- Is it possible to share the inertias and stiffness that you use at NREL? I know that this was determined with ADAMS, but I am hoping that you can give me this values, that way I can see where the differences are.
- When I do not take the added mass into account the roll and pitch frequencies become 0,204 Hz. This is still an 8.5% difference, but is already much better. Could it be that the added mass in NREL/SR-500-45891 has been forgotten? This could also be a coincidence.
- When I calculate the I_yy of the platform I get 5.082610^8 kgm2, not the 5,71610^8 kgm2 that is given in NREL/CP-500-46726. I determine the inertia of the steel cylinder (radius =9m, walltickness = 0.015m, draft = 47.89m, freeboard = 5m) and concrete mass and add these using Steiner’s formula. Is the value from NREL multiplied by some factor to include for internal stiffeners for example?
Thanks in advance. Best regards,
System Matrices.pdf (80.1 KB)
I don’t have the time to verify all of your numbers, but here are a few general comments:
- The pitch/roll modes of the TLP are heavily coupled to the tower-bending modes, so, the natural frequencies of the pitch/roll modes for a rigid TLP will be quite different than the natural frequencies of the pitch/roll modes for a TLP with flexible tower.
- The linearization functionality of FAST v7 (not yet available in FAST v8) can be used to derive the 6x6 rigid-body equivalent mass and stiffness matrices of a floating wind model. If you wish to generate these from FAST v7, you can find a FAST v7 model of the MIT/NREL TLP here: wind.nrel.gov/public/jjonkman/NR … MW_TLP.zip.
- It would take a bit of effort to dig up the derivation of the platform inertias published in NREL/SR-500-45891, but I’m quite confident that the published values are correct because they themselves were corrections from another paper.
Thanks for your useful comments. I will have to check the values with my model that includes towerbending.
Matha et al. (2009) " Model Development and Loads Analysis of an Offshore Wind Turbine on a Tension Leg Platform, with a Comparison to Other Floating Turbine Concepts", Table 6. gives 0.6311 Hz for the 1st Tower F-A natural frequency of the MIT NREL TLP Final design. Jonkman et al. (2009) “Definition of a 5-MW Reference Wind Turbine for Offshore System Development” Table 9-1 gives 0.3240 Hz for the 1st tower F-A natural frequency of the NREL 5MW Reference wind Turbine.
Onshore Turbine 1st Tower F-A natural frequency is lower than the TLPWT value. Since two systems hold the same turbine, what can be the reason for the TLP to have a stiffer tower behavior than the onshore one?
The tower-base boundary condition has a large influence on the tower natural frequency. The land-based turbine has a fixed / cantilevered boundary condition. The TLP has 6 degree-of-freedom motion with mass / inertia and stiffness.