Page 117 - IJB-7-3
P. 117

Li, et al.
                         A                                   B










                         C















           Figure 3. Schematic of nozzle exposed in the air with flowing hydrogel. (A) The temperature distribution. (B) The stress on the microelement.
           (C) Printing process.


           direction of the heat flow will change. The steady-state   coefficient was set to 50 W/m °C. There was a good fit
                                                                                        2
           means that the extrudate temperature, T , will always be   between  the  experimental  and predicted  temperatures,
                                            E
           between the temperatures of the syringe and the AT, that   and it was found that the extrudate temperature rose as
           is, when the AT is lower than the syringe temperature,   the controlled nozzle temperature increased. Because the
           then T ≥ T ≥ T , while when the AT is higher than the   large  ID  nozzle  contained  more  materials  that  carried
                c
                    E
                       A
           syringe temperature, then T ≥ T ≥ T .               the  syringe’s thermal  energy, the  temperature  was less
                                     N
                                         E
                                  A
               Four sets of analyses were conducted using ANSYS   susceptible  to the  AT; consequently, the 23-G nozzle
           software  to  explore  the  temperature  change  during   produced a steeper slope.
           printing using the thermal parameters given in Table 1.   The model was used to explore the influence of the
           Figure 4 presents the simulation results.           AT on the steady temperature of the hydrogel extrudate
               The temperature of the composite hydrogel in the   (Figure 4D), and the simulation results confirmed that the
           nozzle center increased from 21°C to 31°C and decreased   AT had a significant effect on the extrudate’s temperature.
           from 21°C to 19°C gradually (Figure 4B). After more   The influence of the AT on the extrudate’s temperature
           than 40 min of heating/cooling, the temperature reached   was opposite to that of the syringe temperature. It was
           a constant value; this phenomenon could be used as a   confirmed  that  the  smaller  (32-G)  nozzle  containing
           criterion to determine whether a bioink is well heated   fewer materials was more vulnerable to the AT.
           or cooled. Because of the slight heat loss caused by the   During the printing process, the flow of materials
           exposed nozzle, the hydrogel’s steady stage in the syringe   inside the nozzle was affected by the AT. To explore the
           was  not  affected  significantly  by  the  nozzle’s  scaling.   transient change of the material’s temperature, the AT was
           The results of one nozzle were plotted, and several sets   set to 19°C, 27°C, and 32°C, respectively (Figure 4E).
           of  experiments  verified  the  simulation  results.  There   The simulation revealed that the extrudate’s temperature
           was a good agreement between the hydrogel’s measured   responses were relatively fast for both nozzles, reaching
           temperature in the center of the syringe and the heating   a  steady-state  within  10  s. A  smaller  ID  corresponded
           and  cooling  processes’  predicted  temperature.  This   to faster response in the simulation;  furthermore,  it
           confirms  that  the  thermal  simulation  is  reasonably   demonstrated that a smaller nozzle was more vulnerable
           accurate.                                           to the AT.
               The controlled syringe temperature was  increased   The  experimental  results  are  briefly  summarized
           from 19°C to 31°C at a constant AT of 20°C (Figure 4C),   as follows: (a) The central material’s temperature inside
           and the temperature of the extrudate at the tip of the nozzle   the syringe changed slowly and reached  a constant
           was measured. The film coefficient was determined (5 –   temperature after more than 40 min of heating/cooling,
           200 W/m °C in the native flow) using the experimental   (b)  Nozzles  with  larger  IDs  were  less  susceptible  to
                   2
           temperatures  of  the  extrudate.  After  fitting,  the  film   the  AT and  more  vulnerable  to  the  controlled  syringe
                                       International Journal of Bioprinting (2021)–Volume 7, Issue 3       113
   112   113   114   115   116   117   118   119   120   121   122