Medical device design using Computational Fluid Dynamics (CFD)

Transcription

1 Medical device design using Computational Fluid Dynamics (CFD) Session: Winter 2016 IMPORTANT NOTE: This project has 8 deliverables, for each one timely work is expected. 1. General Design Specifications Computational Fluid Dynamics (CFD) is used to design blood related medical devices. Some blood medical devices include: ventricular assist devices, pumps, filters, and tubing. When developing procedures for blood medical devices, biophysical factors must be considered. Some of these factors include: physiologic control feasibility, implantability, haemolysis, thrombosis, biocompatibility, and pump performance. CFD is used to predict a variety of blood related data that includes biophysical factors (i.e. blood cell damage), pressure-flow performance, stress levels, hydraulic efficiencies, and the flow profile through the device [1]. The general Design specifications must specifically take into account: The viscosity: Blood is a non-newtonian fluid (as we will see in chapter 6). The assumption of Newtonian viscosity could be acceptable in some cases, but this assumption should be motivated and validated. It could be done, for example by comparing computational results with experimental ones. However, in the scope of this academic exercise, we are going to consider blood as a Newtonian fluid with a viscosity of 3.5cP. The hemolysis: Red blood cells present in blood are susceptible to damage due to stress imposed from many possible factors. Blood damage or trauma depends on the magnitude and duration of stress loading experienced by the cells (Figure 1). The thrombosis: Thrombosis is a process caused by platelet activation (Figure 1). Local blood flow conditions directly affect the shape, size, and structure of the thrombus. Not only does regional stagnant flow contribute to thrombosis, but there also Haemodynamic factors, such as contact between red blood cells and red blood cells, platelets with red blood cells, and platelets with platelets, that also contribute to thrombosis. Collisions between platelets and red blood cells are much higher in low velocity profiles. These collisions increase the probability of aggregation and granule release [1]. Low shear flows do not cause endothelial damage like high shear flows could. Low shear flows may alter the mass transfer properties for the flow but they preserve the

2 structural integrity of the endothelium. In contrast, the endotheliall damage caused by high shear flows trigger platelet clumping onto the vessel wall. These shear stress levels of 10 Pa create platelet aggregates that have low levels of response from platelets or red blood cells in the form of granule release. However, low shear flow could cause thromogenesis by damaging the cells. This is becausee low shear flow is caused by the recirculation flow vortices that trap cells. This increases the local wall shear stress. Furthermore, flow separation and secondary flow conditions can be caused by the abrupt sectional widening of flow regions. These circumstances promote thrombosis [1]. In conclusion, it is imperative to avoid any blood pump design that could allow conditions like haemolysis or thrombosis to occur both upstream and downstream. The flow conditions that need to be avoided or otherwisee minimized while working on the design phase include: (1) recirculation regions, (2) high shear or extraordinarily low shear stress areas, (3) flow separation, (4) surfaces with sharp edges or high roughness, (5) narrow passages or clearances through the pump and (6) flow stasis or stagnation leading to blood pooling [1]. Figure 1: Shear stress exposure time destruction platelets [2]. plot threshold hemolysis of red blood cells and

3 2 Case of study: Nozzle With a Sudden Contraction and Conical Diffuser (Robin1) Link: CFD is a known tool used to analyze and create medical devices such as prosthetic heart valves and stents. Between 2008 and 2013, the U.S. Food and Drug administration hosted a study to assess the CFD s usefulness in assessing medical device safety. 28 laboratories participated in the study known as Robin#1 1. These participants were asked to model the flow in a generic tubular medical device that has similar flow characteristics to a catheter or needle and similar medical device problems such as issues in dialysis tubing. This generic device contained a sudden contraction and conical diffuser. All 28 participants modeled the flow in a generic tubular device with the following specifications: a m diameter cylindrical nozzle with a sudden contraction and 10 conical diffuser, on either side of a 0.04 m long, m diameter throat (Figure 2). Nozzle throat Reynolds numbers (Ret) were specified at 500, 2000, 3500, 5000, and 6500 under steady flow conditions. A Newtonian viscosity of N-s/m2 and a density of 1056 kg/m3 were specified. All other parameters (inlet/outlet lengths, mesh density, cell/element type, axisymmetric vs. 3D mesh, inlet boundary conditions, etc.) were left to the discretion of the individual participants [3]. Furthermore, 5 fluid mechanics laboratories were also asked to perform experimental validations of flow in the nozzle using PIV measurement. Definitions: Nozzle throat Reynolds numbers (Ret) is the Reynolds number based on the nozzle throat diameter (4 mm) Normalized velocity: u u / u Where u is the normalized velocity, u the velocity from PIV or CFD, and u the average velocity at the inlet, calculated as follows. Q is the flow rate at the inlet (calculated from the velocity profile determined from PIV or CFD), and A the area of the m diameter inlet tube 1

4 Figure 2. Robin#1 : Conical Diffuser. Nozzle specifications: (a) dimensions of nozzle (inlet and outlet lengths unspecified); (b) cross-sectional cuts defined for data submission for the Conical Diffuser. reproduced from [3]. References : 1. Wood HG and al. The medical physics of ventricular assist devices Rep. Prog. Phys. 68 (2005) biofluid mechanics, the human circulation p Stewart FC and al, Results of FDA s First Interlaboratory Computational Study of a Nozzle with a Sudden Contraction and Conical Diffuser, Cardiovascular Engineering and Technology. Volume 4, Issue 4, pp Hariharan P. and al. Multilaboratory particle image velocimetry analysis of the FDA benchmark nozzle model to support validation of computational fluid dynamics simulations.j Biomech Eng.;133(4):

5 Deliverable 1: Geometry Objective: Get familiar with the software Work due: Report (2 pages) with clear annotated figures. Marking: 3 marks Follow the tutorial and then, respecting the specification of the Robin #1 project, build the 2D geometry in the CFD software. Deliverable 2: Meshs Objective: Investigation of different mesh generation procedures Work due: Report (4 pages) with clear annotated figures of the different meshing Marking: 5 marks Deliverable 3: CFD i) Model 1: Mesh the 2D model with a homogeny low density mesh (about 5 nodes across the diameter) ii) Model 2: Double the mesh densities in your model (medium density) iii) Model 3: quadruple the mesh densities in your model (high density) iv) Model 4: from model 2 refine mesh in the boundary layer and refine mesh in part you expect to have high gradient in velocity and/or shear and lower the mesh density in part where your expect with low gradient. Objective: flow calculation Work due: Report (3 pages) with clear annotated figures Marking: 7 marks Following the specification of the Robin #1 project, compute the flow in this device for the different meshings. Limit your analysis to Ret=500 and Ret=6500 i) Verify the convergence ii) Plot color map of velocity. iii) Plot the maximum velocity along center the pipe from the entrance to the output. iv) Determine the entry length before the flow becomes fully developed. v) Determine the Wall Shear Stress (WSS) where the fluid is fully developed vi) Compare the result of the different models Deliverable 4: PIV data Objective: Experimental data analysis Work due: Report (5 pages) with clear annotated figures Marking: 15 marks Data base link: The data are provided for 2 flow directions. The flow direction presented figure 2 corresponds to the conical diffuser. (The sudden expansion is the reverse problem, using the same geometry but the flow in an opposite direction)

6 Extract the PIV data for Ret=500 and Present: - Color map of velocity - Velocity profiles at key location (1 to 12) - Velocity along the center line - WSS along the wall Deliverable 5: Validation Ret=500 Objective: Mesh optimisation, CFD validation Work due: Report (5 pages) with clear annotated figures Marking: 10 marks Optimise the mesh, to get a solution as accurate as possible with the fewer nodes as possible. Hint: You may want to discretize your domain not homogeneously; increase the mesh density where it s needed only. i) Present the error of WSS relative to experimental data ii) the error in velocity profile relative to experimental data iii) the time of computation function of the number of nodes. Deliverable 6: Validation Ret=6500 Objective: Mesh optimisation, CFD validation Work due: Report (5 pages) with clear annotated figures Marking: 10 marks Reedo 5, but Increasing the Reynolds number to get a turbulent flow (Ret=6500). Test different solvers: laminar, k-epsilon etc.... Discuss the difference obtained. Deliverable 7: Objective: Extracorporeal blood device design Work due: Report (20 pages) with clear annotated figures Marking: 40 marks Design the geometry of an extracorporeal detection cell used during hemodialysis procedure. The cell needs to be clear in orderr to measuree the hematocrit with light transmission. To allow the light go through the flowing blood the diameter of the cell should not be greater than 2mm. The total flow rate could be up to 2l/second. The cell will be mounted on as extracorporeal line with an internal dimeter of 5mm The flow design should ensure that hemolysis and thrombosis are avoided

7 Figure 3 : Example of such existing devices: 1) Pressure measurement in Cardiovascular bypass loop (BLOP4 CPB, medtronics) 2) hematocrit measurement during hemodialisis, light transmission (Critline) 3) blood concentration measurement during hemodialisis, ultrason (BVM Fresenius) Deliverable 8 (individual): Objective: Quality control, recommendation Work due: Report (2 pages max) Marking: 10 marks In the context of quality control in a biomedical company, write a clear recommendation for the use of CFD in medical device design processes.

8