了解模態分析
模態分析 is the process of studying and characterising the inherent dynamic properties of a structure or mechanical system. Those properties — its 自然頻率, its 阻尼 ratios, and its 模態振型 — are collectively the system’s “modal parameters.” Together they describe the unique ways a structure will naturally tend to vibrate when it is disturbed. This knowledge is foundational: it lets engineers design structures that survive dynamic forces, and it lets them diagnose and cure stubborn vibration problems by revealing exactly which natural frequency is being stirred up. Where a 振動頻譜 tells you what frequencies a running machine is producing, modal analysis tells you which frequencies the structure is predisposed to amplify — and that distinction is the key to understanding 共振.
1. The Goal: Identifying Modal Parameters
Every structure has a unique set of modal parameters fixed by its physical make-up — its mass, its stiffness, and its damping. The aim of modal analysis is to pin those parameters down:
- Natural frequencies (resonant frequencies): the specific frequencies at which the structure vibrates with the largest amplitude when excited. Any real structure has many of them, ascending in a series.
- Damping ratios: a measure of how quickly the vibration at each mode decays — in other words, how much energy the structure dissipates. Light damping means a tall, narrow resonance peak; heavy damping means a low, broad one.
- Mode shapes: the distinctive pattern of deformation the structure adopts when it vibrates at one of its natural frequencies. Each natural frequency has its own corresponding mode shape — a first bending mode, a torsional mode, and so on.
With these three quantities in hand, an engineer can predict how the structure will respond to essentially any dynamic load it meets in service, and can foresee trouble before it is built into the hardware.
Why the three parameters work together
No single parameter is enough on its own. A natural frequency tells you 其中 a resonance lies on the frequency axis; the damping ratio tells you how severe it will be if excited; and the mode shape tells you where on the structure the motion is largest — and therefore where a sensor will see it, where a correction will be most effective, and where a 節點 of near-zero motion sits. This is why the parameters are always discussed as a set.
2. Types of Modal Analysis
There are three principal routes to a structure’s modal parameters: two experimental and one purely computational.
1. 實驗模態分析 (EMA)
EMA — closely related to the 敲擊測試 — measures the structure’s response to a known, controlled input force. It is the standard method for testing real hardware. The workflow runs as follows:
- Excite the structure with a measured force, usually from an 儀器化衝擊錘 (its tip carries a force sensor) or from an 電動激振器. This controlled excitation is the essence of impact testing.
- Measure the vibration response at one or more locations with 加速規.
- Compute the 頻率響應函數 (FRF) at each point — the ratio of the output vibration to the input force across frequency.
- Use specialised software to fit the set of FRFs and extract the natural frequencies, damping, and mode shapes. The software can then animate each mode shape so the analyst literally sees how the structure flexes at every natural frequency.
Because both the input force and the output response are measured, EMA yields fully scaled modal parameters — the most complete experimental description available.
2. 運轉模態分析 (OMA)
OMA is used when applying a controlled force is impractical or impossible, or when the behaviour under real operating conditions is what matters. Here only the output response is measured — again with accelerometers — while the structure is excited by its normal operational or ambient forces: wind on a bridge, road inputs into a car body, or the working forces inside a running machine. Advanced algorithms then recover the modal parameters from response-only data. It is a more involved approach and the mode shapes come out unscaled, but for large in-service structures it is often the only feasible one. OMA is conceptually a close cousin of operating deflection shape (ODS) analysis, though ODS describes how a structure actually moves at a given operating condition rather than extracting its underlying modes.
3. 分析模態分析 (FEA)
This is the purely theoretical route, built on a computer model — most commonly 有限元素分析 (FEA). Engineers create a virtual model of the structure and the software predicts its modal parameters before any metal is cut. EMA is frequently performed afterwards to validate and refine the FEA model, closing the loop between prediction and measurement so that future “what-if” studies on the model can be trusted.
3. Applications of Modal Analysis
- Troubleshooting resonance problems: the most common application by far. When a machine vibrates excessively, modal analysis reveals whether a structural natural frequency is being driven by an operating force such as running speed or 葉片通過頻率.
- Design validation: engineers confirm that a new product’s natural frequencies are kept clear of known excitation frequencies — engine RPM, blade pass, gear mesh — so resonance never gets designed in.
- Structural modification: once a resonance is identified, the modal model supports “what-if” studies, answering questions like “where should a stiffener go to push this natural frequency higher?” before any change is made.
- Structural health monitoring: a shift in modal parameters over time can flag developing damage — a growing 軸裂紋, for instance, lowers stiffness and therefore drops a natural frequency.
4. 模態分析與共振問題
所有這些的實際效益在於能夠區分兩件在頻譜上看似相同但需要相反治療方法的事情：激振問題與共振問題。如果高振動來自巨大的激振力——例如殘餘 不平衡 ——解決方法是減小該力。如果它來自於自然頻率恰好與運轉頻率重合的結構，減小該力幾乎無濟於事；治療方法是透過改變質量或剛性來移動自然頻率，或增加阻尼。模態分析是告訴您處於哪種情況的工具。諸如 結構共振 與 frame resonance 等狀況正是以這種方式進行診斷的，而在變速機械上，結果通常會輸入到 坎貝爾圖 以繪製激振階次在整個轉速範圍內與自然頻率相交的位置。
5. 現場測量的適用場景
完整的多點模態測試是一項專門的活動，但可靠性工程師在車間通常會以更緊湊的形式遇到它：在進行動平衡校正工作之前，進行快速衝擊測試以尋找疑似的自然頻率。這一步驟很重要，因為對支撐結構處於共振狀態的轉子進行動平衡校正只是在徒勞無功——響應由結構主導，而非由不平衡主導。像 Balanset-1A 這樣的便攜式雙通道儀器讓工程師能夠在運轉速度下捕捉機器自身軸承中的振動，並確認運轉速度遠離結構自然頻率，因此後續的 現場動平衡 才能真正解決根本原因。一旦排除結構因素，同一儀器即可測量動平衡校正轉子並驗證結果所需的 1× 振幅與相位。透過這種方式，廣泛的模態分析學科與專注的動平衡校正任務相互強化：前者確保您正在解決正確的問題，後者則解決該問題。
