Date of Award


Level of Access Assigned by Author

Campus-Only Thesis



Degree Name

Master of Science (MS)


Mechanical Engineering


Vincent Caccese

Second Committee Member

Wilhelm Friess

Third Committee Member

Andrew Goupee


Traumatic brain injuries (TBI) are a leading cause of deaths and disabilities in the US and account for 30 % of injury related deaths. Each day in the US, 153 people die from TBIs and those who survive may face severe disabilities that could last a life time. To predict the response of different head impact scenarios, numerous standards have been released that act as guidelines or criteria’s that companies and researchers must adhere by. Such standards for example are the ASTM F1446-15b or the NOCSAE ND 001, both of which utilize a linear impact system with a uniaxial acceleration response. These acceleration responses can then be processed through head injury predictors such as the Head Injury Criterion (HIC) that rely on the translational acceleration of the center of gravity. The HIC calculates a value of injury based on the linear motion of an impact response of a standardized headform, while a criterion such as the Brain Injury Criteria (BrIC) utilizes the angular motion which is not measured by many standard helmet impact tests. These predictors calculate an injury response where it is compared to an injury threshold formulated by the predictor or against the Abbreviated Injury Scale (AIS) that indicates the predicted extent of the injury. While these techniques do have an immense amount of background research and do help to predict injuries, they do not have an internal mechanical simulated brain that would account for internal dynamic factors that occur within the skull. The brain is a very soft and theoretically an incompressible structure and has a different relative motion from the skull as it keeps traveling forward while the head bounces back (or vice versa) during an impact or high vibrations. This dynamic characteristic intuitively will cause additional dynamic responses to the brain that must be captured to help refine the prediction of brain injuries. This thesis represents a first step toward developing a head impact testing surrogate with a simulated brain that includes embedded acceleration sensors.

The focus of this research is to develop the hardware and to study the functionality of an accelerometer instrumentation array when embedded into a gel matrix used to simulate brain response. The response of the brain simulating gel is compared to that of the relatively rigid test apparatus with no gel to show the difference in the acceleration signals. A hemispherical 3D-printed shell is mounted with four tri-axial MEMS accelerometers and filled with Dow Corning Sylgard® 527 Dielectric gel. This gel was found to simulate the mechanical properties of the brain quite well and is the primary material used in this type of brain research. Four accelerometer packages are positioned as a nine-accelerometer when embedded into the gel material. The test apparatus was than subjected to a variable frequency 3 G harmonic motion wave formulation to analyze the acceleration response due to low to moderate frequency vibrations. The MEMS (Micro-Electro Mechanical System) accelerometers, when embedded into the Sylgard® 527 Dielectric gel, were able to function properly in the gel material and produce comparable results. The Sylgard® gel, when subject to a vibration, provided evidence as to the relative motion and amplitude of acceleration that differs from that of the hemispherical shell, as well as the differences in response between the boundary of the shell and gel matrix as the radial direction increases from the centroid of the gel. It can also be concluded that accelerometers must be embedded into the brain gel model to capture these internal effects.