CFD Analysis and Opti­miza­tion of a Novel Left Ven­tric­u­lar Assist Device (LVAD)

by Yinan Bao

Heart disease is the leading cause of mor­tal­ity world­wide, but with today’s advances in tech­nol­ogy and medicine, many indi­vid­u­als suf­fer­ing from heart disease or injury are lit­er­ally getting a new lease on life. Ven­tric­u­lar assist devices or VADs are mechan­i­cal cir­cu­la­tory assist devices that help to either par­tially or fully replace the function of a failing heart by pumping blood from the lower chambers of the heart (ven­tri­cles) to the aorta. There are dif­fer­ent types of VADs such as positive dis­place­ment pumps that would mimic a natural heart­beat, as well as con­tin­u­ous flow pumps where the recip­i­ent has no heart­beat at all! Some other dif­fer­ences are the type of pump – cen­trifu­gal or axial flow, the place­ment of the pump, and which ven­tri­cle is assisted (i.e. right – RVAD, left – LVAD, or both – BiVAD). 

In an earlier blog post and sub­se­quent update, we looked at a con­tin­u­ous cen­trifu­gal pump VAD design for right heart support (RVAD) from Penn State College of Medicine.

The present case study looks at a Left Ven­tric­u­lar Assist Device (LVAD) based on an axial flow pump, from the Masters’ dis­ser­ta­tion of Yinan Bao at McGill Uni­ver­sity (Montreal, Canada), super­vised by Hristo Valtchanov and Prof. Rosaire Mongrain. 

A Novel Hubless LVAD Concept

An axial flow LVAD with an open hub has several advan­tages, firstly that it is more compact for a given flowrate and pressure, which is rather impor­tant con­sid­er­ing that it is placed in the patient’s body. It also can be implanted per­cu­ta­neously through the skin thus avoiding a more invasive surgery. A larger pumping volume, compared to the con­ven­tional axial flow pump, while oper­at­ing at a slower rota­tional speed avoids damage to the blood cells, which is also critical.

Model Def­i­n­i­tion and Para­met­ric Variation

The LVAD device can be sep­a­rated into various sections includ­ing: the inducer, impeller, and coaxial dif­fusers (primary and sec­ondary) as shown in the figures below. All sections have helical blades (rotors or stators). The dimen­sions as shown are based on the initial concept for which physical testing was carried out, whereas the actual device for patients would likely be quite a bit smaller.

In order to explore and optimize the pump design, CFD sim­u­la­tions were under­taken on the flow passages of the LVAD. The model was created in CAESES® with 32 design para­me­ters. Geo­met­ric con­straints were also included in the model def­i­n­i­tion to respect the design objec­tives and to ensure 3D print­abil­ity (constant blade thick­ness, minimum gap sizes, maximum overhang angles).

The fol­low­ing ani­ma­tion illus­trates the vari­abil­ity of the model based on several of the design para­me­ters used in the study.

Automa­tion Workflow and Opti­miza­tion Strategy

CAESES® was used as the central hub in the automa­tion workflow. In addition to CAD gen­er­a­tion, it was also used as the process inte­gra­tion and design opti­miza­tion (PIDO) platform. CAESES was coupled to the Ansys Work­bench in which the 3D meshing and CFD sim­u­la­tions took place, uti­liz­ing the CFX solver. The post-pro­cess­ing was also done with ANSYS.

The oper­at­ing con­di­tions for this design were 2000 RPM and 3~5 L/​min. 

The primary design and opti­miza­tion objec­tives were:

1. Maximize the pressure rise of the pump. Ulti­mately the target dis­charge pressure of the LVAD would be a normal dias­tolic pressure (70 to 80 mmHG), but by max­i­miz­ing the pressure rise across the pump, the final device can be scaled down with increas­ing pressure.
2. Minimize blood damage measured as Hemol­y­sis Index, (HI). HI is computed using a Lagrangian approach based on a Power-law model with coef­fi­cients from Zhang and Taskin HI = C 𝜏α tβ .

Sec­ondary objec­tives were to maximize pump effi­ciency, pressure rise through the diffuser section, and reduce swirl at the outlet.

A two-step process was employed for the opti­miza­tion. In the first step, a series of design of exper­i­ments (DoE) using the Sobol algo­rithm were used to explore the design space and to uncover which of the para­me­ters exhib­ited the highest sen­si­tiv­i­ties. The 32 design para­me­ters were seg­re­gated into 5 groups and studied in stages. From there, a Nelder-Mead Simplex algo­rithm was used to run the formal opti­miza­tion on a subset of design vari­ables exhibit­ing the highest sensitivities.

The first group of design para­me­ters were related to the blade angles for each section (i.e. inducer, impeller, and primary and sec­ondary dif­fusers). There were 8 para­me­ters in total rep­re­sent­ing the start and end angles; 100 sim­u­la­tions were run in the DoE.

The second group of design para­me­ters were related to the blade geometry vari­a­tions, includ­ing 4 para­me­ters for blade cur­va­ture, and 4 para­me­ters for the number of turns (where 1 turn is equiv­a­lent to 360°) of the respec­tive sections. As with the first group, 100 sim­u­la­tions were conducted.

In the third group, the pump size was varied using 9 design para­me­ters includ­ing outer casing diameter, inlet/​outlet diameter, diameter ratio for inducer/​impeller, as well as gap sizes at the two diffuser interfaces.

The fourth group was related to the lengths of the blades, which were varied with 3 para­me­ters for the respec­tive sections (inducer, impeller, diffuser).

The fifth and final group included 4 para­me­ters rep­re­sent­ing the number of blades for inducer, impeller, and primary/​secondary diffuser.

The fol­low­ing two tables show the nor­mal­ized sen­si­tiv­i­ties of the para­me­ters relative to pressure rise and hemol­y­sis index, respec­tively. The para­me­ters with the highest sen­si­tiv­i­ties (to one or both of pressure and HI) were then selected to form a subset of design vari­ables to run a quicker and more effi­cient opti­miza­tion using the Nelder-Mead Simplex algo­rithm. Thus, the number of para­me­ters was reduced from 32 to the 7 exhibit­ing the highest sen­si­tiv­i­ties for the opti­miza­tion study. (Note that outer diameter and outlet diameter were fixed.)

Results and Conclusions

In total, 500 design variants based on 32 geo­met­ric para­me­ters were sim­u­lated over several stages using Sobol and Nelder-Mead Simplex algo­rithms. The opti­miza­tion study resulted in sig­nif­i­cant improve­ments in several key para­me­ters as shown in the fol­low­ing table, which compares metrics for the baseline design versus the opti­mized design. The design para­me­ters with the greatest influ­ence on delta pressure and HI, were the outer diameter of the pump (which was fixed during the Nelder-Mead Simplex), diffuser diameter ratio, and impeller ratio.

The stream­lines showing the flow patterns through the VAD, as well as the pressure gradient, are compared in the fol­low­ing figure.

In the opti­mized design, there are some inter­est­ing features that facil­i­tate the impres­sive increase in pressure. The blades are diverg­ing in the z‑direction, which further increases the pressure gradient by allowing the blades to act as dif­fusers while de-swirling the flow. This unex­pected feature sub­stan­tially improves pressure recovery and leads to a more compact design. Fur­ther­more, we have a longer path length between blade channels, which allows for a more gradual de-swirling of velocity, further increas­ing the effi­ciency while reducing blood damage.

In con­clu­sion, this case study has illus­trated how CAESES®, coupled to ANSYS/​Workbench was very effec­tive at per­form­ing a design explo­ration and opti­miza­tion of the novel LVAD concept. Fur­ther­more, using a fully auto­mated workflow allowed the explo­ration of unique design features. There were 32 para­me­ters, of which 7 with the highest sen­si­tiv­i­ties were selected for the formal opti­miza­tion, and sig­nif­i­cant improve­ments in pressure rise and effi­ciency were achieved with only modest increase to the Hemol­y­sis index.

About the Author

Yinan Bao grad­u­ated from the Uni­ver­sity of Wis­con­sin-Madison having majored in Mechan­i­cal Engi­neer­ing, after which he com­pleted a research intern­ship at Tsinghua Uni­ver­sity in China. He then went to McGill Uni­ver­sity and obtained his master’s degree in the same field. His research focus is on the opti­miza­tion and CFD analysis for Left Ven­tric­u­lar Assist Devices (LVAD). He is imple­ment­ing multiple opti­miza­tion algo­rithms for the con­fig­u­ra­tion of LVADs to reduce red blood cell damage and to increase their efficiency.

Read More

CAESES® comes with an exten­sive set of capa­bil­i­ties to support the design and opti­miza­tion of medical equip­ment.

Ref­er­ences:

1. Yinan Bao, Hristo Valtchanov, Dr. Renzo Cecere, Prof. Rosaire Mongrain, Com­pu­ta­tional Fluid Dynamics (CFD) Analysis and Opti­miza­tion of Novel Left Ven­tric­u­lar Assist Device (LVAD) Con­fig­u­ra­tion, McGill Uni­ver­sity, 2021
2. Zhang T, Taskin ME, Fang HB, et al., 2011, Study of flow- induced hemol­y­sis using novel Couette-type blood-shearing devices. Artif. Organs;35:1180 – 6.
3. Amy McKean, McGill Blogs, https://blogs.mcgill.ca/spellyourscience/tag/lvad