Page 103 - IJB-9-2
P. 103

International Journal of Bioprinting                               Holistic charge-based MEW scaffold model



            A                           B                                         C


















            Figure 4. Characterization of energy surface. (A) Dependence of energy surface on lateral position when z=5. Curve A and B are the intersecting curves
            of the energy surface with the plane x=0 and y=0. Based on the mathematical analysis,   ∂f   is not a function of y, and   ∂f   is not a function of x in Equation
                                                                   ∂x                 ∂y
            VI; the energy surface can be generated by translating Curve A along the path defined by Curve B or vice versa. Curve B is defined as the lateral characteristic
            curve, which is shown in (B). Curve C and D are also shown in (B) when z is toggled at 30 and 1.1, respectively. The dash-dotted line denotes the prescribed
            locations. When z is continuously varied, lateral characteristic curves continuously evolve, and form a characteristic surface as shown in (C).

            sense,  compared  to  the  lateral  characteristic  curve,  the   within the same layer are spatially positioned within the
            characteristic surface contains information on the energy   same horizontal plane. However, the reality is that the
            variation in both lateral and vertical directions, thereby   elevation of fibers at the intersection points is larger than
            yielding a more comprehensive tool. As shown in Figure S3,   elsewhere. This elevation can be equivalently compensated
            the energy variation along the vertical direction can be   for by instantaneously decreasing z value for the incoming
            depicted by the vertical characteristic curve defined as the   jet segment (next section for further explanation). Third,
            energy variation curve along z direction at the prescribed   the model assumes that the jet deposition process is mainly
            locations (Figure S3A). The vertical characteristic curve is   affected by the electrostatic interaction between the jet
            barely affected by most parameters except β. Moreover, the   and the preexisting fibers, which is true only when the
            effects of β on the energy variation in lateral and vertical   incoming jet segment spatially approximates the scaffold.
            directions are synergistic on the lateral deviation of the   Therefore, it is reasonable to assume the existence of an
            incoming jet segment . Therefore, it is reasonable to only   initial z before which the fiber deviation does not occur.
                             [33]
            consider the lateral characteristic curves when predicting   Another model assumption is that the preexisting fibers of
            the effect of different parameters on the lateral deviation of   the same layer share the same β, but this is not exactly true.
            the incoming jet segment.                          Instead, the value of β is determined by the amount of local
                                                               residual charge, which is closely related to the local fiber
            3.2.4. Inaccuracies and limitations in the model   morphologies. In other words, β can vary not only with
            This section focuses on stating assumptions that reveal   the layer number N, but also with the lateral location of the
            the model inaccuracies and limitations, as well as   fibers. This discrepancy helps to understand the effect of
            providing some compensatory ways to reconcile the   lateral location on lateral jet deviation. Finally, the model
            discrepancy between the predicted and actual results.   assumes that the energy surface is formed by the topmost
            First, the calculation of the energy surfaces assumes that   two layers of fibers in the scaffold rather than the entire
            the preexisting fibers are deposited exactly along the   scaffold volume. Although it is mathematically permissible
            prescribed path without any structural disorder. As such,   to extend this model to consider more layers, this extension
            once the structural disorder has been initiated, the actual   is not implemented herein for two reasons. First, since the
            energy surfaces along with characteristic surfaces/curves   rate of the residual charge decay is yet to be established, β
            will diverge from the prediction. Therefore, although this   cannot be reasonably prescribed for the layers deposited
            model helps to understand and predict the initiation of the   before the topmost two layers. The second reason is to
            structural disorder, it is limited in its ability to precisely   consider that more layers will not introduce new modes
            characterize the energy state for a highly disordered scaffold.   of energy surface evolution except for the aforementioned
            However, if the structural disorder is not significant, the   seven  modes.  Rather, as  shown  in  Figure  S2,  increasing
            energy surface can still be approximately predicted by   the layer number N has the equivalent effect of increasing
            the model. Second, this model assumes that all the fibers   β when β > 1 or decreasing β when β ≤ 1. Despite these


            Volume 9 Issue 2 (2022)                         95                      https://doi.org/10.18063/ijb.v9i2.656
   98   99   100   101   102   103   104   105   106   107   108