<p>The natural vibration period of a structure is an essential parameter for earthquake-resistant design, performance assessment, and structural health monitoring. However, it is challenging to quantify for existing structures and ancient construction typologies that are not properly covered by design standards. This paper investigates the natural vibration periods of historical masonry minarets through a database of 111 structures from Egypt, Turkey, Jordan, and Bosnia and Herzegovina, compiled from literature publications and local archives. The database includes minarets with different geometries, architectural styles, conservation states, ages, and dynamic information determined either from ambient vibration tests (AVTs) or finite element analyses (FEM), depending on data availability in the original sources. Empirical models are proposed to estimate the natural vibration period using simple geometric parameters: the total height <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(H\)</EquationSource> </InlineEquation> and a representative external plan dimension <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(D_{rep}\)</EquationSource> </InlineEquation>, defined as the main body external diameter when available, or as the reported base dimension otherwise. The equations are developed through log-log regression and explicitly include the residual dispersion in logarithmic space, allowing the uncertainty of the estimates to be quantified. To avoid calibrating the empirical formulations on potentially uncertain numerical results, only AVT-derived records are used for the proposed regression models, while FEM-only records are analyzed separately as a comparison subset. Results show that height is a relevant predictor, while combining <InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(H\)</EquationSource> </InlineEquation> with <InlineEquation ID="IEq4"> <EquationSource Format="TEX">\(D_{rep}\)</EquationSource> </InlineEquation> significantly improves prediction accuracy. The resulting bivariate model, <InlineEquation ID="IEq5"> <EquationSource Format="TEX">\(LL{\rm{ - }}BI\left( {H,{D_{rep}}} \right)\)</EquationSource> </InlineEquation>, provides lower dispersion and stronger statistical performance than the height-only formulation. A focused analysis of Cairo’s minarets further indicates that both slenderness and the bivariate formulation are informative within this more homogeneous subset. The proposed model provides a practical tool for preliminary period estimation and large-scale seismic risk screening of masonry minarets.</p>

Dynamic characterization of masonry minarets using regression models

期刊论文
出版模式：

开放获取
发表日期：

2026年6月24日

查看全文

Mariam A. Sallam,
Marco Fasan,
Hany M. Hassan,
Chiara Bedon,
Giovanni Smiroldo,
Hesham E. Abdel Hafiez

摘要

The natural vibration period of a structure is an essential parameter for earthquake-resistant design, performance assessment, and structural health monitoring. However, it is challenging to quantify for existing structures and ancient construction typologies that are not properly covered by design standards. This paper investigates the natural vibration periods of historical masonry minarets through a database of 111 structures from Egypt, Turkey, Jordan, and Bosnia and Herzegovina, compiled from literature publications and local archives. The database includes minarets with different geometries, architectural styles, conservation states, ages, and dynamic information determined either from ambient vibration tests (AVTs) or finite element analyses (FEM), depending on data availability in the original sources. Empirical models are proposed to estimate the natural vibration period using simple geometric parameters: the total height \(H\) and a representative external plan dimension \(D_{rep}\) , defined as the main body external diameter when available, or as the reported base dimension otherwise. The equations are developed through log-log regression and explicitly include the residual dispersion in logarithmic space, allowing the uncertainty of the estimates to be quantified. To avoid calibrating the empirical formulations on potentially uncertain numerical results, only AVT-derived records are used for the proposed regression models, while FEM-only records are analyzed separately as a comparison subset. Results show that height is a relevant predictor, while combining \(H\) with \(D_{rep}\) significantly improves prediction accuracy. The resulting bivariate model, \(LL{\rm{ - }}BI\left( {H,{D_{rep}}} \right)\) , provides lower dispersion and stronger statistical performance than the height-only formulation. A focused analysis of Cairo’s minarets further indicates that both slenderness and the bivariate formulation are informative within this more homogeneous subset. The proposed model provides a practical tool for preliminary period estimation and large-scale seismic risk screening of masonry minarets.