A whole head finite element model for simulation of bone conducted (BC) sound transmission was developed. The geometry and structures were identified from cryosectional images of a female human head and eight different components were included in the model: cerebrospinal fluid, brain, three layers of bone, soft tissue, eye, and cartilage. The skull bone was modeled as a sandwich structure with an inner and outer layer of cortical bone and soft spongy bone (diploë) in between. The behavior of the finite element model was validated against experimental data of mechanical point impedance, vibration of the cochlear promontories, and transcranial BC sound transmission. The experimental data were obtained in both cadaver heads and live humans. The simulations showed multiple low-frequency resonances where the first was caused by rotation of the head and the second was close in frequency to average resonances obtained in cadaver heads. At higher frequencies, the simulation results of the impedance were within one standard deviation of the average experimental data. The acceleration response at the cochlear promontory was overall lower for the simulations compared with experiments but the overall tendencies were similar. Even if the current model cannot predict results in a specific individual, it can be used for understanding the characteristic of BC sound transmission in general.

In this study, novel deep learning models based on high-speed flame images are proposed to diagnose the combustion instability of a gas turbine. Two different network layers that can be combined with any existing backbone network are established—(1) An early-fusion layer that can learn to extract the power spectral density of subsequent image frames, which is time-invariant under certain conditions. (2) A late-fusion layer which combines the outputs of a backbone network at different time steps to predict the current combustion state. The performance of the proposed models is validated by the dataset of high speed flame images, which have been obtained in a gas turbine combustor during the transient process from stable condition to unstable condition and vice versa. Excellent performance is achieved for all test cases with high accuracy of 95.1–98.6% and a short processing time of 5.2–12.2 ms. Interestingly, simply increasing the number of input images is as competitive as combining the proposed early-fusion layer to a backbone network. In addition, using handcrafted weights for the late-fusion layer is shown to be more effective than using learned weights. From the results, the best combination is selected as the ResNet-18 model combined with our proposed fusion layers over 16 time-steps. The proposed deep learning method is proven as a potential tool for combustion instability identification and expected to be a promising tool for combustion instability prediction as well.

A three-dimensional finite-element (FE) model of a human dry skull was devised for simulation of human bone-conduction (BC) hearing. Although a dry skull is a simplification of the real complex human skull, such model is valuable for understanding basic BC hearing processes. For validation of the model, the mechanical point impedance of the skull as well as the acceleration of the ipsilateral and contralateral cochlear bone was computed and compared to experimental results. Simulation results showed reasonable consistency between the mechanical point impedance and the experimental measurements when Young’s modulus for skull and polyurethane was set to be 7.3 GPa and 1 MPa with 0.01 and 0.1 loss factors at 1 kHz, respectively. Moreover, the acceleration in the medial-lateral direction showed the best correspondence with the published experimental data, whereas the acceleration in the inferior-superior direction showed the largest discrepancy. However, the results were reasonable considering that different geometries were used for the 3D FE skull and the skull used in the published experimental study. The dry skull model is a first step for understanding BC hearing mechanism in a human head and simulation results can be used to predict vibration pattern of the bone surrounding the middle and inner ear during BC stimulation.

Bone conduction (BC) sound is the perception of sound transmitted in the skull bones and surrounding tissues. To better understand BC sound perception and the interaction with surrounding tissues, the power transmission of BC sound is investigated in a three-dimensional finite-element model of a whole human head. BC sound transmission was simulated in the FE model and the power dissipation as well as the power flow following a mechanical vibration at the mastoid process behind the ear was analyzed. The results of the simulations show that the skull bone (comprises the cortical bone and diploë) has the highest BC power flow and thereby provide most power transmission for BC sound. The soft tissues was the second most important media for BC sound power transmission, while the least BC power transmission is through the brain and the surrounding cerebrospinal fluid (CSF) inside the cranial vault. The vibrations transmitted in the skull are mainly concentrated at the skull base when the stimulation is at the mastoid. Other vibration transmission pathways of importance are located at the occipital bone at the posterior side of the head while the transmission of sound power through the face, forehead and vertex is minor. The power flow between the skull bone and skull interior indicate that some BC power is transmitted to and from the skull interior but the transmission of sound power through the brain seem to be minimal and only local to the brain–bone interface.