中華民國內政部建築研究所中華民國內政部建築研究所

一、研究緣起
太陽能光電板的設置為近年來政府大力推行之能源政策，惟易遭颱風吹落而造成危險，臺灣在夏秋兩季受到颱風侵襲，光電板掉落情形相當驚人，民眾對於強風威脅又更加明顯有感。現在不論任何型式光電板，在環境中所承受風力仍以升力與負風壓為主，然目前國內尚無針對各類型光電板抗風檢測規範。在太陽能光電板的安裝環境中，風力對結構的影響不單單是穩定的升力對支架的靜態拉力，而是隨著環境中風壓的變化有不同的動態響應，更甚至發生結構振動。由於風力對於支架衍生的動態行為相當複雜，須輔以動態分析、更甚至使用結構振動分析，以瞭解受風後力量如何分配至扣件與結構本身，並探討結構及扣點強度。太陽能光電板受風引致振動的行為模式，除了攸關耐風性能，甚至影響後續在管理維護上的作為，但目前國內尚無針對各類型光電板受風振動檢測規範。
二、研究方法及過程
研究將包含以下內容：
1.蒐集包含各國有關太陽能光電板結構風試實驗相關之文獻，探討並建立適用之太陽能綠電風試實驗技術原則。
2.本研究將進行不同風速下之太陽能綠電風致振動試驗，分析太陽能板受風振動行為，並由試驗分析結果提供試驗流程與分析驗證項目，依供未來相關領域使用參考。
三、重要發現
1.本研究以進行太陽能光電板綠電結構之模態驗證來確保有限元素模型之有效性，以掌握後續待測物模型在受風條件下分析之動態特性，利用基於有限元素法之理論模態分析、實驗模態分析與操作模態分析法相互檢驗估測模態頻率與振型向量，進而確認太陽能光電板綠電系統的動態特性。驗證結果於單板模擬之振形與實驗模態結果相近，而於前7模態之頻率誤差皆在5%以內；另於整體太陽能板系統，在板面上之低階模態相近(頻率約為11 Hz附近)，在此範圍之模態誤差約在7%以內。
2.本研究進行太陽能光電板有限元素模型分析時，利用重建模型內部分層結構，並搭配實驗模態分析選定適配之材料參數，以達到最為貼合實際面板之結果。結果顯示依照正確內部分層設定進形模你，可達到較接近實際板面之試驗分析結果。
3.藉由造風機吹試實驗分析，發現太陽能綠電結構架設上角度引起的風致振動，會改變系統的結構頻率，本研究進行了0度與180度兩組試驗，其中發現相較於180度，0度之系統響應較大，恐會對於結構產生過大的外力荷載，使之壽命降低。
4.本研究使用造風機對於太陽能光電板綠電結構進行吹試量測作業，通過各階段風速下的功率頻譜發現，在不同風速下，除去系統響應放大外，10Hz~15Hz處主要振動行為為面板之模態統頻率皆有偏移之現象，對此可能是在吹試過程中風速為自15m/s遞增至30m/s，且各階風速皆維持2分鐘，導致吹試過程中用於固定面板之螺絲、扣件鬆脫引致。
5.在模態振型比對中，第一與第二模態無論在實驗模態分析亦或是造風機吹試，皆無法僅量測、激勵面板獲得，需透過其餘方向才能將其激發識別。
6.對於本案太陽能板面之表面風壓資料進行分析，原則上試驗風速已達最小雷諾數要求，不同風速之無因次風壓係數已趨一致，在後續分析中，可採用相似律進行計算推估更高風速之風壓，模擬可能產生之振動狀況。
7.表面風壓之譜密度資料顯示在低頻區出現一明顯峰值，其為本案太陽能板系統之主要渦散頻率，史特赫數約為0.218，即由風造成之渦散能量主要集中在此區域。而隨風速增加，其頻率將繼續上升，當該頻率與太陽能架構系統之頻率重合時是否造成反應放大，以目前實驗設置要達到重合，則風速預估需在提升一倍至約55 m/s，考量試驗安全性，研究團隊評估後不進行更高風速試驗，而目前取代之方式在數值模型可靠之狀況下，以數值模擬計算提供更高風速之振動結果。
8.本案研究結果僅針對單一架構之太陽能板系統進行討論，本案之試驗量測與分析流程即為初步建立之太陽能板系統試驗分析方式，結案報告將提供分析方式流程資料以供未來延續應用參考。對於建立完整試驗與分析方式，應需進行更多試驗參數之試驗，如板面排列狀況、架構尺寸、不同風向角、板面傾斜角等，方能完整了解太陽能板受風壓之特性，而部分試驗亦須進行風洞縮尺試驗測試與建立資料庫，經分析統整後建立最終之實尺寸太陽能板面吹試試驗應採用之試驗參數。
四、建議事項
由於國內規範目前並無強制執行太陽能板受風試驗之項目，而是以建議或者應採取其他專家協助等方式說明，建議於往後相關規範修訂內容時可將本案完整提出之造風機試驗與分析方式納入，給定較明確之適用狀況，以供相關使用單位參考。而在完整試驗方法與簡化流程提出前，以本案之分析方式可得出較細緻之風壓分佈與局部扣件所承受之應力分佈，具備學理基礎之評估，然若需討論不同尺度之太陽能板案場規劃，由於計算區域太大，以此方式模擬可能無法收斂，然為尚待處理之問題，未來應針對此問題繼續進行相關研究。
【建議一】
建立風雨風洞實驗室之實驗模態分析能量：立即可行建議
主辦機關：內政部建築研究所
協辦機關：屏東科技大學
本案已建立完成太陽能板架構之基礎試驗流程，氣動力量測部分建研所風雨風洞實驗室已備有壓力掃描量測設備，但在實驗模態分析上，過程中需使用之相關技術、軟體與量測儀器於風雨風洞實驗室尚未備齊。後續建議所內添購完整相關儀器與軟體，以利執行相關量測工作，牽涉技術轉移部份可由本案研究團隊協助進行。
【建議二】
針對太陽能板系統進行更完善之試驗量測與建立相關資料庫：中長期建議
主辦機關：內政部建築研究所
協辦機關：社團法人中華民國風工程學會
本案已成功將風壓數據加載至結構系統上，並進行應力與應變計算，惟僅針對特定架構之太陽能板系統，實際應用上之版面數量眾多，一般無法直接將完整案場縮尺進行試驗並量測相應風壓。因此建議後續進行不同尺寸、角度、風向與排列區域之太陽能板風壓量測，並建立數據資料庫，除了可供相似律計算實場風壓外，也可了解發生最大風壓之條件，在進行太陽能板強度吹試試驗時，可挑最極端之組合進行試驗，以得最保守之太陽能板受力狀況。

I.Background
In recent years, the government has been actively promoting the installation of solar panels as an energy policy. However, solar panel systems are vulnerable to typhoons, which can cause danger. In the summer and autumn, Taiwan is often affected by typhoons, and photovoltaic panels are often damaged by strong winds. Regardless of the type of photovoltaic panels, the wind force in the environment is still mainly lift and negative wind pressure. However, there currently needs to be a wind testing standard for all types of photovoltaic panels in Taiwan. The effect of wind force on the solar panel structure is not only the static tension on the bracket by the steady lift force but also different dynamic response with the change of wind pressure, which causes the structure to vibrate. Since the dynamic behavior of wind force on the bracket is quite complex, dynamic analysis is needed to understand how the force is distributed to the fasteners and structure and to investigate the strength of the structural system and fastening points. The behavior of solar panel vibration caused by wind is not only relevant to wind resistance, but also affects the subsequent maintenance.Therefore, developing a process for detecting wind-induced vibration of photovoltaic panels is necessary.
II.Methodology and process
The research will include the following:
1.To collect literature on wind testing of solar panel structures from various countries and to investigate and develop the principles of wind testing techniques for solar panel systems.
2.This study will investigate the wind vibration behavior of solar panels by performing wind vibration tests at different wind speeds. The test results will provide test procedures and analysis test items for reference in related fields.
III.Conclusions
1.In this study, the validity of the finite element model is verified by performing the experimental modal analysis of the structure of the solar panel, and the dynamic characteristics of the subsequent model to be analyzed under wind conditions are obtained. The validation shows that the mode shape of the single panel simulation is similar to the experimental modal results, and the frequency error in the first seven modes is within 5%; for the whole solar panel system, the low-order mode on the panel surface is similar (frequency around 11 Hz), and the frequency error in this range is within 7%.
2.In this study, the finite element model of a single solar panel was analyzed by reconstructing the inner layer structure of the model and selecting the appropriate material parameters with the experimental mode analysis to achieve the best fit for the actual panel. The results show that the numerical analysis results are closer to the actual panel surface by following the correct inner layer settings for the model.
3.The wind vibration caused by the solar panel structure was found to change the system's natural frequency through the wind-blowing test analysis. It is found that the system response at 0 degree is more significant than that at 180 degree, which may cause excessive external load on the structure and damage.
4.Through the power spectrum at each wind speed, it was found that at different wind speeds, except for the system response amplification, the primary vibration behavior at 10Hz~15Hz was a shift in the modal frequency of the panel. This may be due to the increase of wind speed from 15m/s to 30m/s during the blowing process, and the wind speed was maintained for 2 minutes at each stage, resulting in the loosening of screws and fasteners used to fix the panel during the blowing process.
5.Through the results of mode shapes obtained form finite element analysis and experimental modal analysis, the first and second modes cannot be estimated only from the measurement and excitation panel in either the experimental mode analysis or the wind-blowing test. An application of excitation in other directions is necessary to effectively estimate vibration modes with relatively complete modal information in measurement degree of freedom of the structural system.
6.After checking the surface wind pressure data of the solar panel, it was found that the dimensionless wind pressure coefficients for different wind speeds were consistent, indicating that the test wind speed had reached the minimum Reynolds number requirement. In further analysis, the wind pressure of higher wind speed can be estimated by using the similarity law to simulate the possible vibration condition.
7.The power spectral density of surface wind pressure shows a significant peak in the low-frequency region, which is the main vortex shedding frequency of the solar panel system, with a Strouhal number of 0.218, i.e., the vortex energy caused by the wind is mainly concentrated in this region. The frequency of vortex shedding increases with the increase of wind speed. When the frequency coincides with the natural frequency of the solar panel system, the vibration response may be amplified. If the current experimental setup is to achieve frequency coincidence, the wind speed estimate needs to be doubled to about 55 m/s. Regarding test safety, a higher wind speed test is not conducted. The current replacement method provides the vibration result of higher wind speed by numerical simulation calculation under a reliable numerical model.
8.The results of this study only focus on the single-frame solar panel system, and the test measurement and analysis process of this case is the preliminary test and analysis method of the solar panel system. To develop a complete test and analysis method, more test parameters should be tested, such as panel arrangement, structure size, different wind angles, panel inclination angle, etc. After analysis and discussion, the final test parameters for the real-size solar panel surface wind-blown test were determined.
IV.Suggestions
【Suggestion 1】
Immediately feasible suggestion: development of experimental modal analysis techniques for wind-rain and tine tunnel laboratories.
Organizers: Ministry of the Interior, Architecture and Building Research Institute, National Pingtung University of Science and Technology
The basic wind-blowing test process of solar panel structure has been developed in this project, and pressure scanning measurement equipment has been provided in the wind-rain and wind tunnel laboratory of ABRI. However, in the experimental modal analysis, the relevant technologies, software, and measurement instruments needed in the process have yet to be available in the laboratory. It is proposed to purchase complete related instruments and software to perform relevant measurement work, and the technology transfer can be carried out with the assistance of the research team.
【Suggestion 2】
Medium and long term suggestion: Better testing and measurement of solar panel systems and database development
Organizers: Ministry of the Interior, Architecture and Building Research Institute, Taiwan Association for Wind Engineering
In this case, the wind pressure data has been successfully loaded onto the structural system, and stress and strain calculations have been performed. However, since there are a large number of panels in the actual application, it is generally not possible to directly test the entire site by scale-down and measuring the corresponding wind pressure. It is recommended to measure the wind pressure of solar panels in different sizes, angles, wind directions, and arrangement areas and make a database. When conducting the solar panel strength test, the most extreme combination can be selected to obtain the most conservative solar panel stress conditions.