Turbulent Blood Flow in a Ventricular Assist Device

Application Challenge AC7-03 © copyright ERCOFTAC 2021

Evaluation

Influence of the Discretization Error with URANS

Fig. 5.1. Pressure heads via the impeller (top). The line shows the fit

H_{r}=109.3-3.66\cdot 10^{7}h_{i}^{2}

. Pressure heads via the VAD (bottom). The line shows the fit

H_{p}=74.5-7.49\cdot 10^{7}h_{i}^{2}

. The error bars mark the numerical uncertainties (deviations in percent).

Figure 5.1. shows the discretization error uncertainties of the pressure headvia the VADs impeller and the whole pump. The error intervals of total pressure heads were up to 4.8% for the finest grid. From an engineering point of view, these uncertainties are in an acceptable range for the VAD design. In addition, the uncertainty for the pressure head for the whole VAD (4.8%) is higher than for the impeller alone (1.7%). The reason for this is that turbulent phenomena, e.g. detachment in the outlet guide vane, affect the pressure increase via the whole pump. Those effects may have a significant mesh sensitivity and thus affect the uncertainty of the pressure head for the entire VAD. Furthermore, the relatively small uncertainties for the pressure heads suggest that the finest grid resolution is enough to guarantee a grid-independent solution and no further grid refinement seems to be required for these results.

Fig. 5.2. Stress-dependent variables. Upper left: Volume in the pump which exceeds 9 Pa. The line shows the fit

V(\tau \geq 9Pa)=1.98\cdot 10^{-6}-7.78h_{i}^{2}

. Upper right: Volume in the pump which exceeds 50 Pa. The line shows the fit

V(\tau \geq 50Pa)=2.61\cdot 10^{-7}-1.90\cdot 10^{-3}h_{i}^{1}.25

. Bottom left: Volume in the pump which exceeds 150 Pa. The line shows the fit

V(\tau \geq 150Pa)=3.37\cdot 10^{-8}-0.48h_{i}^{2.08}

. Bottom right: Stress-dependent MIH value. The line shows the fit

MIH=54.1-4.01\cdot 10^{5}h_{i}^{1.31}

.

The uncertainties for the stress-dependent MIH (Fig. 5.2, bottom right) indicates a higher but acceptable value for the finest grid with an $\pm$ 8% interval. In contrast, the coarsest grid has a two times higher uncertainty for MIH (15%) as the uncertainty for the pressure head (15%). On the other hand, the uncertainties of the volumes, which exceeds certain stress thresholds, indicate larger error intervals. These uncertainties are up to 4 times larger as the uncertainties for the pump characteristics, as can be seen in Fig. 5.2. In terms of a potential blood damage prediction with this shear stress field, these results indicate that the finest grid is still too coarse. Of course, the uncertainties of the different blood damage indicators will decrease with higher grid resolutions wherein the absolute values will converge to a final state, but even for grid UR-7 a decay of the slope of the fit in Fig. 5.2. is not obvious in the range of data obtained from the flow computations. Unless the grid size for the finest grid is already quite large for design and optimisation studies, it has still recognizable discretisation uncertainties for the shear-dependent variables, which are important for the blood damage evaluation.