Structural damage detection by non-destructive techniques has been investigated widely by researchers in recent years, and several response-based Structural Health monitoring techniques have been developed. In this paper, a vibration-based damage detection algorithm through model updating utilizing incomplete frequency response function and modal data is presented. First, a modified sensitivity equation is presented to relate the changes of mode shapes to the changes of unknown structural parameters. Then, the developed equations are used to correlate changes of frequency response functions to the changes of structural parameters. Parameter estimation is conducted through least square solution of sensitivity equations. In order to prove the capability of the proposed method for structural damage detection purpose, a numerical example using error contaminated data is presented. Results exhibit robustness of method against measurement and mass modeling errors. KEYWORDS damage detection, frequency response function, modified modal method, sensitivity equation | INTRODUCTIONEvery structure may experience changes in structural properties during its design life because of the occurrence of several conditions such as extreme loading, earthquake, strong wind, erosion, and fatigue. Even if a structure is managed to withstand such changes, it is crucial to investigate its integrity. To ensure this objective, many algorithms have recently been developed to detect the location and severity of structural damages and evaluation of its safety. The basic principle of damage detection algorithms is because damage can cause detectable changes in structural response.A general classification of damage identification techniques refers to types of data used to detect damage including static and dynamic data. Most common dynamic features are modal data including mode shapes, [1,2] natural frequencies, [3,4] strain energy, [5] transfer function parameters, [6] measured modal flexibility, [7,8] frequency response function (FRF), [9][10][11][12] time domain structural responses, [13][14][15] or a combination of these features. [16,17] Damage detection by model updating is the process of modification of structural parameters to minimize discrepancy between response of structure and numerical model. The response residual can be minimized by various approaches. [18][19][20] Because model updating is an inverse problem and the relationship between structural parameters and measured response is inherently nonlinear, methods known as sensitivity-based approaches are developed to simplify solving this problem through linearization of equations. [21] Model updating using FRF improves predictive capabilities of updating procedure through providing large amount of information and has been used by some researchers. [22][23][24] One of the main advantages of using FRFs data in sensitivity-based model updating algorithm is avoiding propagation of errors in parameter estimation. [25,26] By sensitivity-based approaches, FRF residual is minimized to eval...