Data related to the experimental design for powder bed binder jetting additive manufacturing of silicone

The data included in this article provides additional supporting information on our recent publication (Liravi et al., 2018 [1]) on a novel hybrid additive manufacturing (AM) method for fabrication of three-dimensional (3D) structures from silicone powder. A design of experiments (DoE) study has been carried out to optimize the geometrical fidelity of AM-made parts. This manuscript includes the details of a multi-level factorial DOE and the response optimization results. The variation in the temperature of powder-bed when exposed to heat is plotted as well. Furthermore, the effect of blending ratio of two parts of silicone binder on its curing speed was investigated by conducting DSC tests on a silicone binder with 100:2 precursor to curing agent ratio. The hardness of parts fabricated with non-optimum printing conditions are included and compared.


Subject area
Engineering, Materials Science More specific subject area Additive Manufacturing Type of data Table, figure How data was acquired Design of Experiments, Thermocouple Data format Raw, Analyzed Experimental factors The samples were 3D printed based on the experimental design factor treatments in a completely randomized fashion.

Experimental features
For geometrical fidelity optimization, the effects of different values of two factors (layer thickness (LT) and binder dispensing frequency (Fr)) on height and diameter of 3D printed cylinders were studied. The effects of factors on all three responses were simultaneously investigated using desirability function method. For measurement of powder-bed's temperature a thermocouple was used.

Value of the Data
The raw data of dimensional features provided in Table 1 provides the readers with the chance of fact checking the results by following the analysis steps.
The desirability function response optimization (Table 5) shows the values of LT and Fr (in the investigated region) resulting in dimensional features closest to their target values.
The temperature vs. time data provided in Fig. 1 supports our interpretation of thermal analysis of silicone binder using differential calorimetry scanning (DSC).
The thermal behavior of 100:2 silicone binder provided in Fig. 2 shows that increasing the amount of curing agent does not speed up the full crosslinking process, however, it reduces the crosslinking initiation temperature.
The comparison of hardness values shown in Fig. 3 and Tables 6-10 is indicative of the insignificant effect of process parameters on the hardness of fabricated parts for the selected silicone binder and powder.

Data
In order to optimize the 3D printing parameters, a multi-level experimental design was formed with layer thickness (LT) and dispensing frequency (Fr) of the silicone binder deposition as the control factors. The height (H), inner diameter (ID), and the diameter difference (DD) between the inner and outer circles fitted to the cross section of parts are the responses. The outer diameter (OD) is the diameter of the largest circle fitted to the cross-section of the cylindrical parts so that it covers the entire cross-section including the irregular edges. The diameter of the circle that only covers the central parts of the cross-section and not the irregularity caused by the lateral infiltration of silicone binder is ID. The structure of DoE and the measurement details are provided in Table 1. The analysis of variance (ANOVA) results are shown in Tables 2-4 for H, ID, and DD, respectively.  The path to the optimized region for each parameter was found using the response surface method. Finally, all three responses were optimized simultaneously using desirability function technique (utility transfer function). The optimization results are demonstrated in Table 5. The levels of significant factors were selected so that DD was minimized, and H and ID approached the target values of 3 mm and 5 mm, respectively.
The DSC results for the silicone binder reveal that it gets cured almost immediately at a temperature in the range of 100-110°C. In order to make sure this polymerization temperature is reached in 60 s, the temperature of powder bed was measured using a thermocouple. The temperature increase is plotted in Fig. 1 .

Experimental design, materials, and methods
To measure the temperature of powder-bed, a thermocouple was fixed on the surface of the feeding chamber filled with silicone powder using a Kapton tape. The powder-bed temperature was increased by exposing it to the heat provided by a thermal lamp. The temperature values were transferred to a computer using a data acquisition device (NI USB-6009, National Instrμments, TX, USA), and recorded using an in-house developed program in LabView environment.