Numerical and experimental study on the sound generation in the flue organ pipe
European Marie Curie Incoming International Fellowship
European Marie Curie Incoming International Fellowship (Contract Number: 21505) for D. Sc. Seiji Adachi from Japan
The project has been launched at the Fraunhofer IBP for a study on the sounding mechanism of the flue organ pipe. A multidisciplinary approach is taken in this project: in addition to the latest experimental methods such as high resolution acoustic measurements [1-8], flow visualization  and LDA methods [10-11], the state-of-the-art numerical method of computational fluid dynamics (CFD) is employed [12-13]. The project aims to advance the current understanding and to get a sound generation model with which we can simulate the actual flue pipe as accurately as possible. We especially intend to reproduce the sound level, the foot pressures at which the mode transitions occur, as well as the effect of the voicing adjustments on the sound, which is the main concern of the organ builders.
The research on the flue organ pipes has a long history. Since Lord Rayleigh discussed the instability of a jet in his book , the motion of the jet deflected by the acoustic field has been modelled in various ways and applied for the sounding mechanism. Today, using the physical modelling technique, we can simulate the basic behaviour of the flue pipe such as the mode transition and the pitch rise with the increasing foot pressure [15-16]. We are on the right track but our understanding is not perfect. The simulated behaviour is different from that observed in the actual pipe in various points . For example, the sound amplitude is simulated 10 to 20 dB larger than that actually observed. The models can not describe any details of the voicing adjustment.
The main parts of a flue organ pipe are shown in Fig. 1. An air jet comes out from a slit (flue), travels through an opening (mouth) and impinges on an edge (labium) across the mouth. A part of the air entering the pipe excites the sound. On the other hand, the sound deflects the jet and induces the lateral vibration of the jet. The vibration is also excited by the fluid dynamical feedback from the labium.
The sounding mechanism can thus be considered as a self-excited oscillation. The acoustic excitation due to the jet and the jet deflection due to the sound and flow make a feedback loop. If the loop gets a positive gain and it surpasses various losses due to viscosity, thermal conduction and the acoustic radiation, the pipe begins to sound. Whether the sound is generated or not depends on the parameters. For example, by changing the air pressure supplied to the pipe (foot pressure), various oscillatory or non-oscillatory states are obtained.
The difficulty in the research is mainly due to the non-linear characteristics of the fluid dynamics. This was the obstacle preventing Load Rayleigh and others from examining the sounding mechanism in depth. It is, in fact, formidably difficult to predict the motion of the jet analytically. This situation has recently been greatly improved due to the rapid development of computational hardware and software. Today, it is not impossible to simulate the jet motion numerically with an up-to-date software code and sufficient computational resources.
To obtain results that are really useful not only for researchers but also for organ builders, our project emphasizes the agreement between theory and practice. In the research, therefore, the simulation results will be carefully compared with those observed in the acoustic and flow experiments on the actual organ pipes.
The effect of any details of the instrument shape, such as the flue channel dimensions, the nicking at the flue exit and the sharpness of the labium, onto the generated sound can be successfully explained by the physical model. These findings can help the manufacturers of musical instruments to improve the quality of the instruments and to produce new kinds of timbres