1st Annual Workshop on
"Finite Element Modeling in Biology"
June 9-16, 2007 - UMass Amherst

Click here to print

 

As comparative biologists discover the power of finite element modeling and analysis, more and more people want to use it in their own studies. The goal of this workshop was to teach people the fundamental theory behind FEA and introduce them to the process of creating 3D finite element models from CT scans. To do this, three instructors (Betsy Dumont, Ian Grosse and Sean Werle) guided the 10 workshop partipants through the process of building and cleaning 3D surface models from CT scans, meshing those models with solid tetrahedral elements, and analysing them with FE software. The result was a resounding success - everyone went home with an analysis of a complex, 3D finite element model made from their own stack of CT slices!

The workshop got off to a great start with a presentation about FEA and Strand7 FEA software by Ed Martin of Strand7 computing (all the way from Australia!). With a great deal of excitement about what we would do with models we planned to build, we launched into learning to use VGStudio Max. Most of us spent a couple of days with VGStudio Max segmenting our CT scans and extracting the STL surface models that formed the basis of our FE models. By Tuesday, most of us were moving on to the next piece of software, Geomagic Studio 9. Many long hours were spent with Geomagic smoothing out the surface models and editing regions of complex morphology that would prove difficult to mesh. All this work necessitated a mid-week canoe trip down the lovely Connecticut River followed by an excellent meal at a local BBQ joint. The afternoon off was an enjoyable and much-needed break from the computers! The last couple of days of the workshop were spent meshing our models in Strand7, learning more about the theory behind FEA, making decisions about loading the models and assigning material properties, and running analyses.

In the end we all learned a lot, got to know one another, and had a great time. Scroll down the page to read about everyone's project or use these link to go directly to a specific project:

Diapriid Wasp (Joe Cora) Bat Jaw (Tom Eiting) Paranthropus boisei Skull (Ashley Hammond) Savanah Monitor Skull (Casey Holliday) Human Leg (Pranesh Kumar)

Morray Eel Skull (Rita Mehta) Chimpanzee Skull (Caitlin Schrein) Snapping Turtle Skull (Tristan Stayton) Hyena skull (Jack Tseng) Goat Skull (Ed Yoo)

Joe Cora (Ohio State Univeristy). As opposed to the other workshop participants, my objectives for the workshop were not focused on finite element analysis but rather on the tools and techniques used to produce an FE analysis. In particular, I was looking into the feasibility of using microCT scans of small parasitic wasps, family Scelionidae, for taxonomic purposes as well as the potential of FEA on small insect systematic research. A finite element analysis is only as good as the CT scan that produces it, so it is crucial that the CT scan is of high quality and clean. After an initial problem with some strange scans of a couple wet scelionid specimens caused by an unknown filter, I received a diapriid wasp, a sister taxon of Scelionidae, scan that was complete and precise. This scan illustrated many of the internal morphological features that I was interested in seeing which was missing from the previous scans. The early flawed scans and the delay associated with them did not allow me to fully explore the FEA objectives of my trip, but I did perform a preliminary analysis on the diapriid wasp head yielding some interesting results that may be explored in the future. Overall, the insight that I gained from attending the workshop potentially has opened a new portal into small insect research that technology and impetus previously had not allowed.

 

Tom Eiting (UMass Amherst). While I have experience using CT data to analyze structures ranging from turbinals in the nasal passages of mammals to rakers on the gill arches of fish, I have never incorporated Finite Element Analysis (FEA) into my research. Because I am an incoming student to the Dumont lab, where FEA is a critical component of research, attending this workshop seemed a logical route for me to become familiar with the techniques and theory of this powerful analytical tool. Modeling part of the lower jaw of a bat for this workshop served two purposes.  First, because my future work with FEA will be focused principally on bats, these data exposed me to some of the difficulties of working with bats.  Second, the volume I chose was small and relatively simple enough that I was able to tackle all of the steps necessary to perform an FEA project in the span of a week.  This second step was crucial, because early exposure to the whole process will allow me to better understand and predict how to perform future FEA modeling.         Eventually I plan to use FEA to understand airflow and stresses in the nasal chamber of bats.  I am particularly interested in phyllostomids, because they not only use their nose for olfaction and respiration, but they also project echolocation calls through the nose.  I am interested in the evolution and interaction of these three major functions in the noses of phyllostomids.

 

Ashley Hammond (Florida Atlantic University). The robust australopithecine clade, commonly assigned to the genus Paranthropus, have a suite of characters associated with a specialized masticatory apparatus. This includes features such as postcanine megadonty, sagittal crests, large flaring zygomatic arches, and an overall increase in cranial and dental robusticity. Since it has been hypothesized that Paranthropus relied heavily on its posterior dentition, I decided to use finite element modeling to determine how much facial stress would be generated by chewing in different places. I based the model on CTs of OH 5, a relatively complete Paranthropus boisei specimen from Olduvai Gorge. To determine if the locations of stress distribution were different for different bites, I applied equal forces to the first upper incisor and the second premolar in different trials. Based on the preliminary results, it appears that anterior masticatory loading generates more stress than posterior loading. This is consistent with the idea that paranthropines were suited to postcanine dental use and were better able to neutralize forces loaded on posterior teeth than their small anterior teeth.

 

 

Casey Holliday (Marshall University). Many diapsids possess a number of open intracranial joints that have been implicated in hypotheses of feeding function suggesting that: 1, they are mobile joints involved in cranial kinesis; or 2, they are developed to insulate the braincase and cephalic sensory structures from forces transmitted during feeding. Because most animals that possess these features are either too small for in vivo technology (e.g., lizards, birds) or are extinct (e.g., non-avian dinosaurs) Finite Element Analysis offers otherwise unobtainable insight into cranial form and function. To investigate these hypotheses, the microCT data of a Savannah monitor lizard (Varanus exanthematicus) was transformed into a 3D mesh. The basal joint (i.e., basisphenoid-pterygoid joint) was modeled as either being “open”, as found in the live animal, or “closed” where the joint was bridged by virtual materials forming a direct linkage between the braincase and palate. These models were then loaded with a 100N bite force and analyzed in a number of different treatment environments: unilateral and bilateral bite forces, different tooth locations and constraints. It was found that in the “open” model, stress concentrations were found at a number of locations in the palate where other cartilaginous joints are found, and that the pterygoid and face are likely subjected to strong loads. In the “closed” model, stresses were found to divert away from the palate and face but extend into the braincase suggesting that an “open” basal joint does indeed insulate the braincase from feeding-generate forces.

 

Pranesh Kumar (Brighton and Sussex University Hospital, UK). The goal of my project was to study force transmission through the lower limbs of humans while standing and under loads imposed by walking. I built a model of the lower limb from medical CT scans which included cortical bone, trabecular bone, medulary cavities, and the articular cartilages of the knee and hip. In running the finite element analyses, I was particularly interested in three primary of issues. The first was the effect of bone architecture on load transmission. The second was the location of stress concentration points and their correlation with the locations of common fractures. The third was the role of the articular cartilages on force transmission through the joints. 

 

 

 

 

Rita Mehta (UC Davis) is interested in using FEA to study bite force mechanics in moray eels and barracuda. For the workshop she focussed on building and analyzing a model of a morray eel. Ultimately she is interested in how different skull elements affect skull deformation, strain patterns, and loading when apprehending prey by biting.

 

 

Caitlin Schrein (Arizona State University). I am interested in differences between the shapes of skulls among living and extinct apes and how those differences are related to distributing and resisting stress and strain related to mastication. More specifically, I would like to test hypotheses about skull form and adaptations to chewing tough or hard foods, such as mature leaves, seeds or hard fruits.  I plan to use finite element analysis to investigate biological questions pertaining to the structure and optimization of form in the primate skull and how it is affected by forces related to chewing. I built a model using CT scans of a female Pan troglodytes schweinfurthi skull (provided by Dr. Robert McCarthy via the National Museum of Natural History, Dr. Bruno Froehlich, Dr. Bernard Wood, CASHP and the GW NSF-IGERT program).  I plan to use this model as well as other models I have and will build to explore intraspecific and interspecific differences among and between chimpazees, gorillas and orangutans.  I loaded the model using the Bone Load program (Dumont et al., in press), applying muscle forces for the anterior temporalis, superficial and deep masseter muscles and the medial pterygoid.  Muscle force data were adapted from Strait et al. (2005) for a macaque (multiplied by 3). The model was contrained at the TMJ (rotation with no translation) and the left M1 (no rotation or translation) to simulate unilateral biting on the M1.  Isometric bone material properties were assigned (average for a macaque skull; Strait et al. 2005) to the entire cranium.  The mandible was only used to determine the position in space of muscle insertions (for direction of muscle force). 

 

 

Tristan Stayton (Bucknell University). I performed a couple of analyses of my data set, which consisted of models of the skull and lower jaw of the common snapping turtle (Chelydra serpentina). My primary interest was in the stresses produced in the skull during three different biting regimens (biting at the tip of the beak, bilateral biting further posterior, and unilateral biting at the posterior of the beak). Results indicate minimal stresses in the dorsal skull, with moderate stresses in the palate and surprisingly high stresses in the dorsal quadrate and around the tympanum (even given the proximity of these elements to the jaw joint). Additionally, I conducted a convergence analysis designed to determine the degree of resolution in the models needed to obtain consistent results. I started with a "complete" surface model of the turtle skull (200,000 triangles) and successively reduced the number of triangles by 50%, remeshing and each time conducting similar biting analyses using identical parameters. The results appear consistent down to 2500 triangles (nearly a 100x reduction in the complexity of the model). Finally, I modeled similar bites using only the lower jaw, and will compare these to the results expected using beam theory.

 

 

Jack Tseng (Natural History Museum of LA County). I am using finite element modeling to examine the morphological evolutionary history of Hyaenidae (Carnivora, Mammalia) and to test the interpretation of large-bodied taxa of Borophaginae (Canidae, Carnivora) as North American analogs of hyaenids. With emphasis on fossil skulls representing various evolutionary species over the past 18-million-year record of hyaenids in Eurasia, I am exploring the structural evolution of hyaenid crania in distributing and resisting mechanical stress induced from P3 biting (“bone-cracking”). Using the extant spotted hyena Crocuta crocuta as a model species with known bone-cracking behavior, I am currently analyzing and comparing a finite element model of Dinocrocuta gigantea, a large, 10-million-year old relative of Crocuta from Fugu, Shaanxi Province, China (specimen on loan from the Institute of Vertebrate Paleontology and Paleoanthropology [IVPP], Chinese Academy of Sciences). FEA will eventually be applied to other taxa such as Ictitherium, Hyaenictitherium, Adcrocuta, Pachycrocuta, Borophagus, and Epicyon. This study is in collaboration with the IVPP, Natural History Museum of Los Angeles County, Aerospace and Mechanical Engineering at the University of Southern California (USC), and School of Dentistry at USC. Major technical support and discussion of the method have been provided by Dr. Betsy Dumont at the University of Massachusetts, Amherst.  

 

Ed Yoo (Harvard University) was interested in determining the dynamic stress response of the keratin sheath and bone horncore of domestic goat skull under impact loading.