© Emerald Publishing Limited. Purpose - Fused deposition modelling (FDM) is one of the most popular additive manufacturing processes, and is widely used for prototyping and fabricating low-cost customized parts. Current FDM machines have limited techniques to monitor process conditions to minimize process errors, such as nozzle clogging. Nozzle clogging is one of the most significant process errors in current FDM machines, and may cause serious consequences such as print failure. This paper aims to present a physics-based dynamic model suitable for monitoring nozzle clogging in FDM machines. Design/methodology/approach - Liquefier mount of an FDM extruder is analysed as a beam excited by a uniform loading distributed over a partial length. Boundary conditions and applied loads for a direct-type FDM extruder are identified and discussed. Simulation of nozzle clogging was performed by using nozzles of different diameters from 0.5 to 0.2 mm, in step change of 0.1 mm. Sets of experiments were carried out by measuring vibrations of the liquefier block mount during FDM extrusion. Findings - The mount of a liquefier block in an FDM extruder can be used to place a vibration sensor to monitor process errors such as nozzle clogging. Liquefier block mount's transverse vibration amplitudes increase non-linearly when nozzle starts to block. Practical implications - The proposed model can be effectively used for monitoring nozzle clogging in FDM machines, as it is based on the physics relating the FDM process parameters and the nozzle blockage. Originality/value - The novelty of this paper is the unique method of modelling the FDM process dynamics that can be used for monitoring nozzle clogging.