Search

As recipient of the 2018 Robert M. Nerem Education and Mentorship Medal from the ASME, I was asked to prepare an article to contribute my reflections on these topics, that arise whenever we offer advice and guidance to our younger colleagues. This represents my personal views on Bob Nerem, after whom the award is named, my experiences that have impacted me as a mentor, and some words of advice, offered cautiously and with qualifications, with regard to best practices in mentoring.

This article focuses on different research efforts of Emergent Behaviors of Integrated Cellular Systems (EBICS) for creating biological machines. EBICS’s mission is to create a new scientific discipline for building living, multicellular machines that solve real-world problems in health, security, and the environment. The goal of building biological machines may be achieved through either of two distinct pathways— engineered systems and emergent systems—and the distinctions between them are important and fundamental. While a great deal of progress has been made developing the components for biological machines, one key challenge is the limited understanding of how cells interact with each other and with their environment. In order to create a biological machine, engineers will need to understand the language that cells of different types use to communicate with each other. Biological machines of the future will encompass the complexities of nature, the intricacies of which we are just beginning to comprehend.

The human tongue is a structurally complex and extremely flexible organ. In order to better understand the mechanical basis for lingual deformations, we modeled a primitive movement of the tongue, sagittal tongue bending. We hypothesized that sagittal bending is a synergistic deformation derived from co-contraction of the longitudinalis and transversus muscles. Our model of tongue bending was based on classical bimetal strip theory, in which curvature is produced when one muscle layer contracts more so than another. Contraction was modulated via mismatched thermal expansion coefficients and temperature change (to simulate muscular contraction). Our results demonstrated that synergistic contraction produced curvature and strain results which were in better correspondence to empirical results derived from tagging MRI than were the results of contraction of the longitudinalis muscle alone. This fundamental reliance of tongue bending on the synergistic contraction of its intrinsic fibers supports the muscular hydrostat theory of tongue function.

A two-layer model is used to simulate the mechanical behavior of an airway or other biological vessel under external compressive stress or smooth muscle constriction sufficient to cause longitudinal mucosal buckling. Analytic and finite element numerical methods are used to examine the onset of buckling. Post-buckling solutions are obtained by finite element analysis, then verified with large-scale physical model experiments. The two-layer model provides insight into how the stiffness of a vessel wall changes due to changes in the geometry and intrinsic material stiffnesses of the wall components. Specifically, it predicts that the number of mucosal folds in the buckled state is diminished most by increased thickness of the inner collagen-rich layer, and relatively little by increased thickness of the outer submucosal layer. An increase in the ratio of the inner to outer material stiffnesses causes an intermediate reduction in the number of folds. Results are cast in a simple form that can easily be used to predict buckling in a variety of vessels. The model quantitatively confirms that an increase in the thickness of the inner layer leads to a reduction in the number of mucosal folds, and further, that this can lead to increased vessel collapse at high levels of smooth muscle constriction.

External pneumatic compression of the lower legs is effective as prophylaxis against deep vein thrombosis. In a typical application, inflatable cuffs are wrapped around the patient’s legs and periodically inflated to prevent stasis, accelerate venous blood flow, and enhance fibrinolysis. The purpose of this study was to examine the stress distribution within the tissues, and the corresponding venous blood flow and intravascular shear stress with different external compression modalities. A two-dimensional finite element analysis (FEA) was used to determine venous collapse as a function of internal (venous) pressure and the magnitude and spatial distribution of external (surface) pressure. Using the one-dimensional equations governing flow in a collapsible tube and the relations for venous collapse from the FEA, blood flow resulting from external compression was simulated. Tests were conducted to compare circumferentially symmetric (C) and asymmetric (A) compression and to examine distributions of pressure along the limb. Results show that A compression produces greater vessel collapse and generates larger blood flow velocities and shear stresses than C compression. The differences between axially uniform and graded-sequential compression are less marked than previously found, with uniform compression providing slightly greater peak flow velocities and shear stresses. The major advantage of graded-sequential compression is found at midcalf. Strains at the lumenal border are approximately 20 percent at an external pressure of 50 mmHg (6650 Pa) with all compression modalities.

A new model is used to analyze the fully coupled problem of pulsatile blood flow through a compliant, axisymmetric stenotic artery using the finite element method. The model uses large displacement and large strain theory for the solid, and the full Navier-Stokes equations for the fluid. The effect of increasing area reduction on fluid dynamic and structural stresses is presented. Results show that pressure drop, peak wall shear stress, and maximum principal stress in the lesion all increase dramatically as the area reduction in the stenosis is increased from 51 to 89 percent. Further reductions in stenosis cross-sectional area, however, produce relatively little additional change in these parameters due to a concomitant reduction in flow rate caused by the losses in the constriction. Inner wall hoop stretch amplitude just distal to the stenosis also increases with increasing stenosis severity, as downstream pressures are reduced to a physiological minimum. The contraction of the artery distal to the stenosis generates a significant compressive stress on the downstream shoulder of the lesion. Dynamic narrowing of the stenosis is also seen, further augmenting area constriction at times of peak flow. Pressure drop results are found to compare well to an experimentally based theoretical curve, despite the assumption of laminar flow.

Experiments were conducted over a range of Reynolds numbers from 50 to 8000 to study the pressure-flow relationship for a single bifurcation in a multi-generation model during steady expiratory flow. Using the energy equation, the measured static pressure drop was decomposed into separate components due to fluid acceleration and viscous energy dissipation. The frictional pressure drop was found to closely approximate that for an equivalent length of curved tube with the same curvature ratio as in the model bifurcation. The sensitivity of these results to changes in airway cross-sectional shape, non-planar configuration, and flow regime (laminar-turbulent) was investigated. In separate experiments using dye visualization and hot-wire anemometry, a transition to turbulent flow was observed at Reynolds numbers between 1000 and 1500. Transition had very little effect on the pressure-flow relation.

In this companion paper to “Part I: Numerical Simulations,” we report in vitro experimental studies performed on a simple model leg consisting of a “vein” of thin-walled latex tubing surrounded by “tissue” of open-pore foam rubber. Three modes of periodic external compression were investigated: i) uniform compression; (ii) graded compression, decreasing from ankle to knee; and (iii) sequential compression, progressing from ankle to knee. The modes are compared on the basis of three hemodynamic criteria: degree of vessel collapse, level of fluid velocity, and level of shear stress. In uniform compression these measures of merit are distributed very nonuniformly along the length of the leg: they are high near the proximal end of the cuff but low elsewhere, a result due to the formation proximally of a partially occlusive throat. The latter does not form in either graded or sequential compression, with the consequence that favorable values of the three measures of merit occur more uniformly along the length of the pressurized region. It is concluded that either the graded or sequential mode of compression, or perhaps a combination of the two, would be more effective than uniform compression as a prophylaxis against deep vein thrombosis.

This paper presents the results of a numerical study of the technique of periodic external compression for the prevention of deep vein thrombosis. In the model the veins of the lower leg are portrayed as a continuous system rather than as discrete elements as in previous models. Consequently, we are able to explore the detailed effects of different modes of compression including (i) uniform compression, the simultaneous application of uniform pressure over the entire lower leg, (ii) graded compression, the application of nonuniform pressure, maximum at the ankle and minimum at the knee, and (iii) wavelike compression, a wave of compression proceeding from the ankle toward the knee. These numerical results indicate that the effectiveness of uniform compression is severely compromised by the formation of a flow-limiting throat at the proximal end of the compression cuff that reduces both the rate at which blood is discharged from the lower leg and the total blood volume removed. Both of these detrimental effects can be avoided by the use of either wavelike or graded compression. Both alternate methods are shown to produce more uniform augmentation of volume flow rate, flow velocity, and shear stress, throughout the entire lower leg. In the companion paper, Part II [18] (see following article), these same compression modes are tested using a simple hydraulic model consisting of a single latex tube inside a foam cylinder as a highly simplified representation of a human leg.

A simple, continuous, one-dimensional model for the geometry and structure of the bronchial airways is used for the analysis of fluid flow patterns which have been observed in forced expiration maneuvers. Various phenomena within the conducting system associated with flow limitation are investigated: (a) the conditions in which a “choke” (flow limitation) can occur in a compliant system; (b) theoretical flows that are physically impossible; (c) the possibility of having elastic jumps downstream of the choke point; (d) perturbations in the physical parameters of the conducting system.

The mechanics of the lung parenchyma is studied using models comprised of line members interconnected to form 3-D cellular structures. The mechanical properties are represented as elastic constants of a continuum. These are determined by perturbing each individual cell from a reference state by an increment in stress which is superimposed upon the uniform stretching forces initially present in the members due to the transpulmonary pressure. A force balance on the distorted structure, together with a force-deformation law for the members, leads to a calculation of the strain increments of the members. Predictions based on the analysis of the 3-D isotropic dodecahedron are in good agreement with experimental values for the Young’s, shear, and bulk moduli reported in the literature. The model provides an explanation for the dependence of the elastic moduli on transpulmonary pressure, the geometrical details of the structure, and the stress-strain law of the tissue.