Recycled HDPE reinforced Al₂O₃ and SiC three dimensional printed patterns for sandwich composite material

The recycling of thermoplastics is one of the major concerns in recent industrial environment [1, 2]. Yang et al (2012) outlined that one of the hot issues in re-using polymeric multi-material thermoplastics is the difficulty in isolating their sub-parts [3]. Therefore it is important to have an alternative solution which allows multi-material recycling either without the need of separation or some chemical/physical route which enables recycling without segregation of multi-materials [4]. Commercially various waste materials are created by service, manufacturing industries and municipal solid wastes [5]. The expanding awareness about the Earth has contributed to the concerns related with transfer of the disposal of the wastes [6]. One of the major concerns especially in developing nations like India is solid waste management. McDougall et al (2008) highlighted that with the shortage of space for land filling and because of its regularly expanding cost, waste utilization has turned into an appealing option in contrast to disposal [7]. Research is being carried out on the usage of waste items in various application areas [8]. Plastic wood is one example of the items manufactured by utilizing waste plastic gathered in civil recycling programs [9]. Panda et al (2010) reported that regularly, half or a greater amount of the feedstock utilized for plastic timber production is made out of HDPE, low density polyethylene (LDPE) and polypropylene (PP) [10]. Municipal solid waste is the principle source of waste plastics, around 66% of the aggregate created, while the second waste plastics stream originates from the conveyance and mechanical divisions [11, 12]. This issue should essentially incorporate a substantial scale utilization of the different reusing strategies for materials and additionally energy recovery [13].

Figure 1 depicts the flow chart of recycling and reusing the plastic solid waste. In recent past various studies have been reported on recycled HDPE for wood plastic composite (WPC) [14–20]. In addition, WPC from bio plastics have been examined, yet high costs of bio-plastics prevent its further usage [21]. Sommerhuber et al (2016) investigated mechanical characteristics of WPC produced using chrome copper arsenate (CCA) mixed with virgin and reused HDPE [22, 23]. Fillers are typically added to the polymer framework with the point of enhancing its thermal and mechanical properties [24, 25]. The conventional IC comprises of the mould and the typical ceramic shells for preparations of prototypes [26]. The chain of processes for the ceramic shells processes (see figure 2) comprises of the tooling, shell creation and casting stages. The shape for wax design generation is planned and machined from aluminum stocks. For intricate patterns, numerous separation lines and movable inserts are consolidated into the moulds [27]. Pattnaik et al (2012) outlined that the finished form is covered with agent, assembled and infused with liquid wax [28]. After lowering the temperature, the mould is broken to get the final patterns. Each pattern is joined onto a wax sprue framework to shape a group (see figure 3) in shell creation stage. The cluster is repeatedly dipped to put coat of investment slurry, which contain graded suspensions of refractory particles, followed by stucco application to prepare shell thickness and strength. When dried, the wax design is liquefied out by means of autoclaving to uncover the inside cavities of mould. Further the firing was performed to assemble quality and evacuate residues. In casting stage, liquid metal is filled in the heated shells to shape the castings, which are removed after cooling by splitting the shell amid the knockout procedure. Every casting is isolated, washed down and exposed to final procedures.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In case of conventional IC process, substantial investments are focused on model or generation of tooling advancement [26]. The committed resources increment significantly with mould intricacy or low volume fabrication. Tooling expenses for wax IC reach from a few thousands to a huge amount relying upon size and multi-complex nature, while lead-times go between half a month to months extendable upon machine floor routines as well as capacities. All things considered, a tool maker needs to assess individual mould configurations before focusing to manufacturing since design errors or iterations are typically costly and tedious to alter [29–32]. Recent innovative advances have upgraded the precision, execution and durability of final results, enabling some to serve as tooling for low-volume production and in some cases, high-volume production [33–35]. The rapid prototyping (RP) process chain as rapid investment casting (RIC) is shown in figure 4.

Zoom In Zoom Out Reset image size

In this proposed method of RP assisted IC, initially feed stock filament wire comprising of HDPE reinforced with Al₂O₃/SiC (of average diameter 50 μm) was prepared, which was used to print patterns with open source FDM. After this the printed components were used as SP and ceramic in the form of siladent powder (commercially used in clinical dentistry to prepare outer shell) has been applied. Further, steps for autoclaving and knockout remain same as traditional IC. The positive feature of this method is that unlike tradition methods which take 10–12 weeks for single part, costing from 100–500 US$, this has total processing time of 7–8 h, which not only saves time but also cost significantly. The steps for proposed novel route are shown in figure 5. Further based upon figures 5, 6 shows steps followed for preparation of SCM in this study. In this work ceramic particle reinforcement like, SiC, Al₂O₃ chosen as reinforcement to improve the surface characteristics. Further reinforcements as pilot test was performed for characterization of melt flow index (MFI) for different blends of SiC, Al₂O₃ with HDPE matrix.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

MFI is one of the materials properties which represent the function of flow ability of materials per unit time [33]. The MFI of recycled HDPE used in present study was 10 g/10 min as per DIN ISO 1133 [34]. After establishing MFI with different proportions, feedstock filaments have been extruded on single screw extruder (SSE). The 09 sets of feedstock filaments were extruded and filament with best mechanical properties was selected for 3D printing on FDM. After selection of feed stock filament, parametric optimization was performed on FDM setup for printing the cubical shaped prototypes as SP. Finally these SP were used for IC.

Initially flow behavior of selected proportions was established and subjected for filament wire preparation (after blending ceramic particles SiC/Al₂O₃ in Al matrix). Establishment of flow properties is necessary step as excess density of melting material may chock the FDM nozzle and may results in lower quality of product. So, MFI of product was established with the help of melt flow tester machine according to the ASTM-D1238 on different blends. Table 1 shows 'L9' based design of experiment (DOE) followed for preparation of wire feed stock filament and output as peak strength.

As observed from table 1, the best obtained wire in terms of peak strength is sample 9, which has been selected for further experimentation. Again by following DOE based upon Taguchi L9 orthogonal array, parts were printed in form of cubical shape on FDM. Table 2 shows various input parameters selected for printing on FDM.

At this stage, the wire, which was prepared with the help of the SSE, was made to run on the FDM machine and cubical patterns of 20 × 20 × 20 mm were printed. All the patterns, prepared with the help of the computer aided design (CAD) software, were held in the die for the preparation of the outer shell as shown in figure 6. Then after locating the part in die, the die stone powder mixed with siladent liquid was poured in to the die and placed for 10–20 min (for solidification). Die stone powder is already established for castings in clinical dentistry applications for development of partial dentures/implants. So, this proposed route has been tried for preparation of SCM. After this the ceramic shell mould was placed in the heating furnace and set at 700 °C. After 1 h all the material present in the ceramic shell was lost (except reinforced ceramic particles) and Al (alloy 6063) was poured into the cavity created after the evaporation (see figure 6). Since the ceramic particles were blended in the wire, and after evaporation of HDPE these ceramic particles are settled on the base due to gravity in the final cast. This is the major advantage of this method, as the outer surface comprises of ceramic particles and the inner core is still soft. So in order to prepare SCM, a particular surface/face of the cast can be easily modified in place of making whole part reinforced with ceramic particles. The final obtained casts were then subjected to the hardness, and microstructure testing. In addition grain size and surface porosity has been tested to check flaws on the surface of the cast. Finally, the results obtained are compiled in table 3 and analyzed by using the analysis of variance (ANOVA) technique.

Based upon table 3, table 4 shows signal to noise (SN) ratio for hardness (by considering maximum the better type case) and porosity (by considering minimum the better type case). To ascertain the statistical implication of developed experimental model for industrial utility, ANOVA at 95% confidence interval level has been considered.

Based upon table 4, figures 7(a) and (b) shows graphical representation of SN ratios for surface hardness and porosity.

Zoom In Zoom Out Reset image size

Further table 5 shows that probability (P) value for both cases (hardness, porosity) have been observed under 0.05, this ensures the legitimacy of the experimentation.

Based upon table 5, table 6 shows response table for ranking of input parameters.

As observed from figure 7(a), part density 60%, raster angle 60° and nozzle diameter (0.5 mm) are the best settings for hardness whereas for porosity part density 60%, raster angle 60° and nozzle diameter (0.3 mm) are the best settings. Since the grain size only depicts the structural stability and formulation of the structure the best settings for grain size were not explored. As observed from tables 7(a), (b) and 6 for surface hardness nozzle diameter (0.5 mm) with maximum value and for porosity, print part density (60%) minimum value are the input parameters ranked as number '1'. This is but obvious because larger nozzle diameter will lead to deposition of more ceramic particles per printed layer, where as less print density eases the evaporation of HDPE thus reduces the porosity. Based on table 6, a corollary has been solved to check the variation in results.

For hardness:

$$\begin{eqnarray}&&{\eta }_{{\rm{opt}}}={\rm{m}}+\left({{\rm{m}}}_{{\rm{A}}1}\,\mbox{--}\,{\rm{m}}\right)+\left({{\rm{m}}}_{{\rm{B}}2}\,\mbox{--}\,{\rm{m}}\right)+\left({{\rm{m}}}_{{\rm{C}}3}\,\mbox{--}\,{\rm{m}}\right)\end{eqnarray} \tag{ 1 }$$

Where: η_opt is optimum SN ratio, m is the overall mean of SN data, m_A1 is the mean of SN ratios for density at level 1 and m_B2 is the mean of SN ratios for angle at level 2, m_C3 is the mean of SN data for nozzle dia. at level 3.

Predicted property value:

$$\begin{eqnarray}&&{{{\rm{Y}}}_{{\rm{opt}}}}^{2}={\left(10\right)}^{\eta }{{\rm{opt}}}^{/10},{\rm{for}}\, \mbox{`} {\rm{larger}}\,{\rm{is}}\,{\rm{better}}\mbox{'}\,{\rm{case}}\end{eqnarray} \tag{ 2 }$$

Calculations:

Here m = 30.23 (mean of SN ratios)

From the response table of surface hardness, m_A1 = 30.50, m_B2 = 30.33, m_C3 = 30.65 (see table 7).

Further putting the values,

$$\begin{eqnarray}&&{\eta }_{{\rm{opt}}}=30.23+\left(30.50-30.23\right)+\left(30.33-30.23\right)+\left(30.65-30.23\right)\end{eqnarray} \tag{ 3 }$$

$$\begin{eqnarray}&&{\eta }_{{\rm{opt}}}=31.02\,{\rm{dB}}\end{eqnarray} \tag{ 4 }$$

$$\begin{eqnarray}&&{\rm{Now}},{{{\rm{y}}}_{{\rm{opt}}}}^{2}={\left(10\right)}^{\eta {\rm{opt}}/10}\end{eqnarray} \tag{ 5 }$$

$$\begin{eqnarray}&&{{{\rm{Y}}}_{{\rm{opt}}}}^{2}={\left(10\right)}^{31.02/10}\end{eqnarray} \tag{ 6 }$$

$$\begin{eqnarray}&&{{\rm{y}}}_{{\rm{opt}}}=35.56\,{\rm{HV}}\end{eqnarray} \tag{ 7 }$$

It has been calculated the predicted surface hardness is 35.56 HV. The confirmatory experimentations has been performed at the predicted setting (i.e. density at level 1, raster angle at level 2 and nozzle diameter at level 3) the mean of values of surface hardness was obtained 34.9 HV that is very close to the predicted value. Similarly, the other properties has been predicted and tabulated against the observed values for selection of optimum process parameters (see table 7).

Based upon table 2, photomicrographs for three different FDM printed part densities (S.No.1, 4 and 7) are shown in figure 8.

Zoom In Zoom Out Reset image size

As observed form figures 8(a), (b), (photomicrographs and porosity graphs) uniform grain structure and porosity at the surface of the Al matrix composites have been observed as per ASTM E562, ASTM E1245. Figure 8(c) shows the rendered images for better understanding of the microstructures of the samples which clearly shows uniform and equally spaced grains with no distortion. Further figure 8(d) shows surface roughness profile which clearly depicts that sample prepared with processing conditions as per S.No 7 is better in comparison to S.No.1 and 4 (see table 3). On the contrary the porosity (%) for sample 7 was also higher. So SEM based analysis has been performed on the sample at S.No.3 which has shown best surface properties (as per table 3).

The SEM images have been obtained at ×500 for experimental conditions as per S.No. 3 (see table 3) as it has maximum surface hardness and minimum porosity (see figures 9(a)–(e)). As observed from figures 9(a)–(c), ceramic particles are present on the bearing surface in Al matrix, thus justifying the novel route for preparation of SCM. Further as observed from rendered images of photomicrographs (figure 9(d)) and surface roughness profile (figure 9(e)) the sample no. 3 (as per table 3) has uniformly distributed grains with better Ra as compared to sample no. 1, 4 and 7 (see figure 8(d)) and posses better hardness 34.9 HV with acceptable porosity and grain size (see table 3).

Zoom In Zoom Out Reset image size

These SCM can be used as two phase material [36], reinforcing elements for cement composites with fractal architecture [37], shape optimization of the force networks of masonry structures for civil engineering applications, tensegrity architecture, meta-material printing [38–40]. The proposed route for recycling of plastic solid waste is in line with the application domain discussed by other investigators via 3D printing [41–46].

• In this case study a SCM has been developed with novel method by using the die stone powder. The result of the study suggests that SCM developed by proposed method resulted into good shell stability as well as ability to withstand high temperature (around 1100 °C). Hence the proposed route can be successfully used for lab scale development of SCM in very short time and one does not required heavy setup of conventional IC.
• Since the ceramic mould made by the die stone powder can withstand temperature up to the 1100 °C, this means not only Al, but also various other metals and alloys, which has melting range up to the 1100 °C can be used for IC.
• The maximum hardness value obtained in present case study was 34.9 HV and predicted maximum value comes to be 35.56 HV (which are very close). Also the predicted porosity values were also seen very close to actual value (actual as 6.16% and predicted value 5.53%). Hence the proposed macro model can be suitably used for controlling the surface hardness and porosity levels in final functional prototypes.

Authors are highly thankful to DST for financial support under project (DST/TSG/NTS/2014/104, Dated-2-12/2015).

The authors have no conflict of interest.