Flow Uniformity in Oxygenators with Different Outlet Port Design : ASAIO Journal

Secondary Logo

Journal Logo

Respiratory Support

Flow Uniformity in Oxygenators with Different Outlet Port Design

Hirano, Ayaka*; Yamamoto, Ken-ichiro; Matsuda, Masato*; Inoue, Masaru; Nagao, Sukemasa; Kuwana, Katsuyuki; Kamiya, Masahiro; Sakai, Kiyotaka*

Author Information
ASAIO Journal 55(3):p 209-212, May 2009. | DOI: 10.1097/MAT.0b013e31819c6f19
  • Free

Abstract

In recent years, there have been more than a million open-heart surgeries around the world requiring extracorporeal circulation.1 The oxygenator containing a hydrophobic gas-permeable membrane performs in extracorporeal circulation functions as an alternative to human lung and allows transfer of oxygen into the blood. The oxygenator currently used in clinical cases is an extracapillary membrane oxygenator. The laminar film on the blood side is the rate-limiting step in gas exchange. For the development of new and improved oxygenators, analyzing the blood flow is important, because it is the main determinant of the size of the laminar film on the blood side.2–6 The blood flow is controlled by jacket structure and by the wrapping pattern7,8 and packing density9 of hollow fibers. The oxygenation performance of oxygenators depends on membrane surface area, axial jacket length, and outside diameter of a hollow fiber.10 Jacket structure is one of the important factors determining the blood flow.

The simulation analysis is an appropriate method of finding out the optimum blood flow in the oxygenators for cost cutting.11,12 Technical visualization is also necessary, because it is difficult to evaluate the adequacy of the simulation analysis.

Magnetic resonance imaging (MRI) and X-ray computed tomography (CT) have been recently used to study dialysis fluid flow distribution in several hollow fiber hemodialyzers.13–16 MRI results may not represent the true dialysis fluid-side flow distribution, because it is impossible to measure the dialysis fluid-side flow distribution by MRI while letting the test solution to flow. X-ray computed tomography needs a contrast medium to visualize the dialysis fluid-side flow. Those results may not represent the true dialysis fluid-side flow distribution, because the test solution having the contrast medium is not the same viscosity and density as the dialysis fluid. In this study, the step-response method17 was used. This method uses electrodes and a conductive fluid to visualize and analyze the blood flow in the oxygenator jackets.18 It has an advantage that it is possible to measure the arrival time reaching each electrode while letting the test solution to flow, which has the same viscosity and density as blood. On the other hand, it has a disadvantage that many electrodes placed in the test membrane oxygenator may cause the disorder of blood flow. Therefore, in the present study, fine electrodes (ϕ = 100 μm) were used leading to suppressed disorder of the blood flow outside the hollow fibers (ϕ = 300 μm).

The objective of the present study is to determine the optimal blood outlet port structure for attaining uniform flow by visualizing the blood-side flow, using the calculated dimensionless fluid arrival time needed for the high-conductive fluid to reach the copper-wire electrodes placed in the membrane oxygenators at flow rates of 1.0 and 5.0 L/min.

Materials and Methods

We evaluated a cylindrical type extracapillary membrane oxygenator, HPO-20 (Senko Medical Instrument Manufacturing Co., Ltd, Tokyo, Japan). HPO-20 has a tangential blood outlet port (referred to herein as “Tangential HPO-20”). We engineered a vertical blood outlet port in Tangential HPO-20 (referred to herein as “Vertical HPO-20”) with a symmetric structure. The structure of the HPO-20 oxygenators is shown in Figure 1. The blood channel at the opposite side of the outlet port in HPO-20 was smaller than that at the outlet port side by approximately 10% for allowing the blood to flow easily to the outlet port. The Vertical HPO-20 is not completely symmetric with respect to the oxygenator structure. The technical data for the HPO-20 are shown in Table 1.

F1-5
Figure 1.:
Structure of HPO-20s.
T1-5
Table 1:
Technical Data on Membrane Oxygenators

As shown in Figure 2, copper-wire electrodes were placed in four sections of the oxygenator jackets. In total, 120 insulated copper-wire electrodes (radial 6 points, and longitudinal 5 points at equal intervals) were placed at planes from section A to D of each HPO-20 at 90° intervals on a cross-section. Fine electrodes (ϕ = 100 μm) were used, so that alteration of the blood flow outside the hollow fibers (ϕ = 300 μm) was suppressed.

F2-5
Figure 2.:
Positions of the electrodes placed in the packed hollow fibers.

We applied the step-response method using electrodes and a conductive fluid to visualize and analyze the blood flow in the oxygenator. The experimental system based on the step-response method is shown in Figure 3. The electrical potential at each electrode was continuously measured after a low-conductive fluid was switched to a high-conductive fluid having the same viscosity as blood at flow rates of 1.0 and 5.0 L/min. A 33 wt% solution of glycerol (Wako pure chemical Industries, Ltd., Osaka, Japan) was used as the low-conductive fluid (3.3 mPa s, 10 ns/m at 293 K), and a 33 wt% solution of glycerol containing 3 wt% of NaCl (Wako Pure Chemical Industries, Ltd., Osaka, Japan) was used as the high-conductive fluid (3.3 mPa s, 46 ns/m at 293 K).

F3-5
Figure 3.:
Experimental system based on the step-response method.

The time course of changes of the electrical potential was measured on a potentiometer GR3500, expanded units GR3000 and GR3010 (Keyence Co., Osaka, Japan). Normalized electrical potential Vn(t) (–) was determined using Equation 1 and calculated from the measured data on the electrical potential.

where V, Vmax, and Vmin are the measured, maximum, and minimum electrical potential (V), respectively.

The time course of the changes in the normalized electrical potential is shown in Figure 4. Because the data on the maximum electrical potential at each electrode are different, it is inappropriate to define the time at which the normalized electrical potential becomes 1 as the fluid arrival time. Further unstable electrical potentials were measured at the plateau phase. Hence, we defined the normalized electrical potential when the high-conductive fluid reached the target electrode as 0.5.18 This time was defined as dimensionless fluid arrival time td (–), which was calculated by Equation 2.

F4-5
Figure 4.:
Time-course of changes in the normalized electrical potential.

where tr and ta are the time needed to reach electrode (s), and the mean retention time (s), respectively.

The mean retention time ta was calculated by Equation 3.

dividing volume on the blood side of the oxygenator VB (ml) by blood flow rate QB (ml/min).

The values for ta of both HPO-20s at 1.0 and 5.0 L/min were 7.40 (–) and 1.48 (–), respectively. Distribution of the time needed for the fluid to reach the electrodes was depicted as the dimensionless fluid arrival times in each section. To evaluate the blood flow of test oxygenators in a way easy to understand, the blood-side flow distribution in each oxygenator was visualized using 120 dimensionless fluid arrival time data.

Results and Discussion

Blood Flow of Membrane Oxygenators with Different Outlet Port Positions

To examine the influence of the outlet port position on the blood flow, we visualized the blood-side flow in both HPO-20s by the step-response method at flow rates of 1.0 and 5.0 L/min.

Figure 5 shows the distribution of the dimensionless fluid arrival time to the electrodes placed in the Tangential HPO-20 (1: 1.0 L/min, 2: 5.0 L/min). The step-response method is useful to visualize the blood flow in a membrane oxygenator.18

F5-5
Figure 5.:
Dimensionless fluid arrival time to the electrodes placed in “Tangential HPO-20.”

The blood-side flow of the Tangential HPO-20 was more rapid in section B, whereas that in section C was delayed at the top and bottom regions of the jacket. In particular, the blood-side flow in section C was greatly delayed when compared with that in the other sections because the section C was markedly remote from the outlet port. Thus, the blood-side flow of the Tangential HPO-20 was nonuniformly distributed. This tendency was observed at flow rates of both 1.0 and 5.0 L/min, but the delay in the blood-side flow at 1.0 L/min was more pronounced than that at 5.0 L/min. Delayed distributions of the time needed for the high-conductive fluid to reach the electrodes in all sections were more pronounced in the bottom region of the jacket, because the blood flow was suppressed and complicated by flows rebounding at the bottom in the case of the Tangential HPO-20.

Visualization of the Tangential HPO-20 shows that it is difficult for the test solution to flow out from the side opposite to outlet port. We engineered the Vertical HPO-20 with a vertical blood outlet port by modifying the Tangential HPO-20 to reduce stagnation and channeling. Figure 6 shows the distribution of the dimensionless fluid arrival time to the electrodes placed in the Vertical HPO-20 (1: 1.0 L/min, 2: 5.0 L/min). Comparison of four cross-sections of the Tangential HPO-20 with those of the Vertical HPO-20 was considered to be inappropriate because of the varying distances from the four cross-sections to the outlet port in the two types of test oxygenators. Hence, we compared the overall blood-side distribution rather than the sectional blood-side distribution between Tangential HPO-20 and Vertical HPO-20. With the Vertical HPO-20, the time needed for the high-conductive fluid to reach the outer blood channel was short at the center, whereas there was a delay at the bottom. The type of jacket influences the blood-side flow at the outer bottom region. This brings about a rebound flow, so that it takes much longer for a high-conductive fluid to reach the bottom. However, the blood-side flow in section A of the Vertical HPO-20 was similar to that in section D, whereas the blood-side flows in sections B and C were almost the same. Although sections B and D to the outlet port of the Tangential HPO-20 had the same lengths, the streamline directions were different. This difference causes different distributions of the time needed for the high-conductive fluid to reach the electrodes. On the other hand, sections A and D of the Vertical HPO-20 have the same lengths and angles to the outlet port. Thus, they showed similar distributions of the times needed for the high-conductive fluid to reach the electrodes. Furthermore, the region with markedly delayed flow was decreased in the Vertical HPO-20. The blood-side flow of the Vertical HPO-20 was insensitive to the high-conductive fluid flow rate. Outlet port position is a major factor determining the blood flow distribution. The vertical outlet port produces more uniform blood flow than the tangential outlet port because of the small stagnation and reduced channeling.

F6-5
Figure 6.:
Dimensionless fluid arrival time to the electrodes placed in “Vertical HPO-20.”

Conclusions

The blood outlet port structure of a membrane oxygenator is a major factor determining the pattern of blood flow. The blood flow in an extracapillary membrane oxygenator with a vertical blood outlet port is well distributed so that it produces more uniform blood flow than that with a tangential outlet port because of the small stagnation and reduced channeling.

Acknowledgment

This work was carried out in the Consolidated Research Institute for Advanced Science and Medical Care, the Global COE program “Center for Practical Chemical Wisdom,” and the program entitled “Promotion of Environmental Improvement for Independence of Young Researchers” under the Special Coordination Funds for Promoting Science and Technology provided by the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan.

References

1. Mengoli A: Creation and working of an oxygenator, industrial view. Recherche et Ingénierie Biomédicale 27: 14–15, 2006.
2. Wegner AJ: Oxygenator anatomy and function. J Cardiothorac Vasc Anesth 11: 275–281, 1997.
3. Lund LW, Federspiel WJ, Hattler BG: Gas permeability of hollow fiber membranes in a gas-liquid system. J Memb Sci 117: 207–219, 1996.
4. Dorrington KL, Ralph ME, Bellhouse BJ, et al: Oxygen and CO2 transfer of a polypropylene dimpled membrane lung with variable secondary flows. J Biomed Eng 7: 89–99, 1985.
5. Wickramasinghe SR, Garcia JD, Han B: Mass and momentum transfer in hollow fibre blood oxygenators. J Memb Sci 208: 247–256, 2002.
6. Sakai K, Yanagisawa M, Hosoya N, et al: Comparison of oxygenation and flow characters of inside and outside blood flow membrane oxygenators. Artif Organs Today 3: 57–80, 1993.
7. Nagase K, Kohori F, Sakai K: Oxygen transfer performance of a membrane oxygenator composed of crossed and parallel hollow fibers. Biochem Eng J 24: 105–113, 2005.
8. Krantz WB, Bilodeau RR, Voorhees ME, Elgas RJ: Use of axial membrane vibrations to enhance mass transfer in a hollow tube oxygenator. J Memb Sci 124: 283–299, 1997.
9. Matsuda N, Sakai K, Yamamoto K, Iwasaki H: Effects of hollow fiber packing fraction on blood flow pattern and gas transfer rate of an intravascular oxygenator (IVOX). J Memb Sci 179: 231–241, 2000.
10. Matsuda N, Sakai K: Blood flow and oxygen transfer rate of an outside blood flow membrane oxygenator. J Memb Sci 170: 153–158, 2000.
11. Funakubo A, Taga I, John W: Flow vectorial analysis in an artificial implantable lung. ASAIO J 49: 383–387, 2003.
12. Zhang J, Nolan TDC, Zhang T, et al: Characterization of membrane blood oxygenation devices using computational fluid dynamics. J Memb Sci 288: 268–279, 2007.
13. Poh CK, Hardy PA, Gao D, et al: Effect of flow baffles on the dialysate flow distribution of hollow-fiber hemodialyzers: A nonintrusive experimental study using MRI. J Biomech Eng 125: 481–489, 2003.
14. Osuga T, Obata T, Ikehara H: Detection of small degree of nonuniformity in dialysate flow in hollow-fiber dialyzer using proton magnetic resonance imaging. Magn Reson Imaging 22: 417–420, 2004.
15. Takesawa S, Terasawa M, Sakai K, et al: Nondestructive evaluation by X-ray computed tomography of dialysate flow patterns in capillary dialyzers. ASAIO J 34: 794–799, 1988.
16. Sakai Y, Wada S, Anno I, et al: Nondestructive evaluation of blood flow in a dialyzer using X-ray computed tomography. Artif Organs 3: 197–204, 2003.
17. Rake H: Step response and frequency response methods. Automatica 16: 519–526, 1980.
18. Funakubo A, Fukui Y, Sakuma I, et al: Three dimensional flow measurement and visualization study for the design of membrane oxygenator. Jap J Artif Organs 27: 81–86, 1998 [in Japanese].
Copyright © 2009 by the American Society for Artificial Internal Organs